Hugging Face's logo

ggml-org
/
GLM-4.5-Air-GGUF

Text Generation
GGUF
quantized
conversational
Model card Files
xet
 Community
GLM-4.5-Air-GGUF / convert.log
ggerganov's picture
ggerganov HF Staff
Upload folder using huggingface_hub
23a7a81 verified
Raw
History Blame Contribute Delete
410 kB
+ OUTPUT_DIR=./upload-GLM_4.5_Air
+ LLAMA_CPP=./llama.cpp
+ DISPLAY_NAME=GLM-4.5-Air
+ QUANTIZE=./llama.cpp/build/bin/llama-quantize
+ python3 ./llama.cpp/convert_hf_to_gguf.py ./model-temp-GLM_4.5_Air-PRIMARY --no-tensor-first-split --outtype bf16 --outfile ./upload-GLM_4.5_Air/GLM-4.5-Air-BF16.gguf --no-mtp --model-name GLM-4.5-Air
INFO:hf-to-gguf:Loading model: model-temp-GLM_4.5_Air-PRIMARY
INFO:hf-to-gguf:Model architecture: Glm4MoeForCausalLM
INFO:hf-to-gguf:gguf: loading model weight map from 'model.safetensors.index.json'
INFO:hf-to-gguf:gguf: indexing model part 'model-00001-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00002-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00003-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00004-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00005-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00006-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00007-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00008-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00009-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00010-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00011-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00012-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00013-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00014-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00015-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00016-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00017-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00018-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00019-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00020-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00021-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00022-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00023-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00024-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00025-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00026-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00027-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00028-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00029-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00030-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00031-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00032-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00033-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00034-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00035-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00036-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00037-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00038-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00039-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00040-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00041-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00042-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00043-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00044-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00045-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00046-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00047-of-00047.safetensors'
INFO:gguf.gguf_writer:gguf: This GGUF file is for Little Endian only
INFO:hf-to-gguf:Exporting model...
INFO:hf-to-gguf:token_embd.weight, torch.bfloat16 --> BF16, shape = {4096, 151552}
INFO:hf-to-gguf:blk.0.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.0.ffn_down.weight, torch.bfloat16 --> BF16, shape = {10944, 4096}
INFO:hf-to-gguf:blk.0.ffn_gate.weight, torch.bfloat16 --> BF16, shape = {4096, 10944}
INFO:hf-to-gguf:blk.0.ffn_up.weight, torch.bfloat16 --> BF16, shape = {4096, 10944}
INFO:hf-to-gguf:blk.0.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.0.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.0.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.0.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.0.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.0.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.0.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.0.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.1.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.1.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.1.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.1.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.1.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.1.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.1.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.1.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.1.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.1.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.1.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.1.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.1.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.1.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.1.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.1.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.1.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.2.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.2.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.2.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.2.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.2.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.2.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.2.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.2.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.2.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.2.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.2.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.2.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.2.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.2.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.2.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.2.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.2.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.3.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.3.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.3.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.3.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.3.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.3.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.3.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.3.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.3.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.3.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.3.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.3.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.3.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.3.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.3.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.3.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.3.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.4.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.4.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.4.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.4.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.4.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.4.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.4.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.4.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.4.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.4.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.4.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.4.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.4.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.4.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.4.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.4.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.4.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.5.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.5.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.5.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.5.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.5.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.5.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.5.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.5.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.5.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.5.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.5.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.5.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.5.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.5.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.5.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.5.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.5.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.6.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.6.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.6.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.6.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.6.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.6.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.6.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.6.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.6.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.6.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.6.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.6.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.6.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.6.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.6.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.6.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.6.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.7.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.7.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.7.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.7.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.7.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.7.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.7.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.7.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.7.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.7.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.7.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.7.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.7.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.7.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.7.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.7.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.7.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.8.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.8.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.8.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.8.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.8.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.8.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.8.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.8.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.8.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.8.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.8.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.8.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.8.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.8.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.8.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.8.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.8.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.9.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.9.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.9.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.9.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.9.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.9.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.9.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.9.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.9.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.9.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.9.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.9.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.9.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.9.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.9.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.9.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.9.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.10.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.10.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.10.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.10.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.10.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.10.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.10.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.10.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.10.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.10.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.10.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.10.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.10.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.10.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.10.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.10.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.10.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.11.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.11.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.11.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.11.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.11.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.11.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.11.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.11.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.11.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.11.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.11.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.11.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.11.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.11.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.11.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.11.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.11.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.12.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.12.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.12.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.12.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.12.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.12.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.12.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.12.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.12.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.12.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.12.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.12.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.12.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.12.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.12.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.12.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.12.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.13.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.13.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.13.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.13.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.13.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.13.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.13.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.13.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.13.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.13.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.13.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.13.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.13.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.13.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.13.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.13.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.13.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.14.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.14.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.14.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.14.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.14.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.14.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.14.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.14.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.14.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.14.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.14.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.14.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.14.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.14.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.14.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.14.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.14.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.15.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.15.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.15.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.15.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.15.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.15.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.15.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.15.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.15.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.15.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.15.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.15.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.15.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.15.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.15.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.15.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.15.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.16.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.16.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.16.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.16.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.16.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.16.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.16.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.16.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.16.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.16.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.16.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.16.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.16.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.16.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.16.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.16.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.16.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.17.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.17.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.17.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.17.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.17.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.17.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.17.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.17.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.17.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.17.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.17.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.17.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.17.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.17.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.17.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.17.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.17.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.18.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.18.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.18.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.18.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.18.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.18.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.18.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.18.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.18.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.18.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.18.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.18.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.18.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.18.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.18.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.18.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.18.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.19.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.19.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.19.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.19.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.19.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.19.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.19.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.19.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.19.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.19.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.19.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.19.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.19.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.19.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.19.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.19.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.19.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.20.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.20.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.20.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.20.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.20.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.20.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.20.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.20.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.20.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.20.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.20.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.20.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.20.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.20.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.20.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.20.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.20.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.21.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.21.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.21.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.21.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.21.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.21.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.21.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.21.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.21.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.21.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.21.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.21.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.21.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.21.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.21.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.21.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.21.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.22.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.22.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.22.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.22.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.22.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.22.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.22.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.22.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.22.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.22.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.22.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.22.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.22.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.22.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.22.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.22.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.22.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.23.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.23.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.23.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.23.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.23.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.23.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.23.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.23.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.23.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.23.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.23.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.23.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.23.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.23.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.23.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.23.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.23.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.24.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.24.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.24.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.24.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.24.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.24.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.24.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.24.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.24.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.24.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.24.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.24.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.24.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.24.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.24.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.24.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.24.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.25.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.25.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.25.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.25.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.25.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.25.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.25.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.25.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.25.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.25.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.25.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.25.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.25.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.25.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.25.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.25.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.25.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.26.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.26.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.26.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.26.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.26.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.26.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.26.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.26.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.26.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.26.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.26.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.26.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.26.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.26.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.26.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.26.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.26.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.27.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.27.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.27.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.27.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.27.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.27.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.27.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.27.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.27.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.27.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.27.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.27.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.27.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.27.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.27.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.27.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.27.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.28.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.28.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.28.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.28.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.28.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.28.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.28.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.28.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.28.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.28.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.28.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.28.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.28.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.28.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.28.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.28.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.28.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.29.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.29.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.29.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.29.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.29.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.29.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.29.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.29.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.29.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.29.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.29.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.29.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.29.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.29.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.29.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.29.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.29.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.30.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.30.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.30.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.30.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.30.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.30.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.30.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.30.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.30.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.30.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.30.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.30.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.30.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.30.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.30.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.30.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.30.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.31.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.31.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.31.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.31.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.31.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.31.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.31.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.31.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.31.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.31.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.31.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.31.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.31.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.31.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.31.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.31.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.31.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.32.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.32.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.32.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.32.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.32.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.32.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.32.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.32.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.32.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.32.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.32.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.32.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.32.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.32.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.32.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.32.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.32.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.33.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.33.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.33.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.33.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.33.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.33.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.33.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.33.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.33.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.33.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.33.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.33.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.33.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.33.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.33.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.33.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.33.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.34.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.34.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.34.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.34.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.34.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.34.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.34.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.34.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.34.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.34.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.34.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.34.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.34.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.34.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.34.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.34.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.34.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.35.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.35.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.35.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.35.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.35.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.35.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.35.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.35.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.35.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.35.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.35.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.35.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.35.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.35.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.35.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.35.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.35.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.36.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.36.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.36.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.36.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.36.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.36.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.36.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.36.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.36.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.36.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.36.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.36.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.36.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.36.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.36.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.36.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.36.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.37.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.37.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.37.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.37.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.37.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.37.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.37.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.37.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.37.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.37.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.37.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.37.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.37.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.37.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.37.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.37.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.37.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.38.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.38.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.38.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.38.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.38.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.38.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.38.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.38.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.38.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.38.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.38.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.38.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.38.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.38.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.38.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.38.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.38.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.39.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.39.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.39.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.39.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.39.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.39.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.39.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.39.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.39.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.39.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.39.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.39.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.39.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.39.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.39.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.39.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.39.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.40.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.40.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.40.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.40.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.40.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.40.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.40.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.40.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.40.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.40.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.40.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.40.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.40.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.40.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.40.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.40.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.40.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.41.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.41.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.41.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.41.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.41.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.41.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.41.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.41.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.41.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.41.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.41.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.41.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.41.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.41.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.41.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.41.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.41.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.42.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.42.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.42.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.42.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.42.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.42.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.42.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.42.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.42.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.42.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.42.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.42.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.42.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.42.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.42.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.42.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.42.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.43.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.43.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.43.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.43.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.43.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.43.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.43.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.43.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.43.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.43.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.43.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.43.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.43.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.43.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.43.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.43.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.43.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.44.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.44.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.44.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.44.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.44.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.44.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.44.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.44.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.44.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.44.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.44.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.44.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.44.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.44.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.44.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.44.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.44.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:output.weight, torch.bfloat16 --> BF16, shape = {4096, 151552}
INFO:hf-to-gguf:blk.45.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.45.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.45.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.45.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.45.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.45.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.45.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.45.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.45.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.45.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.45.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.45.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.45.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.45.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.45.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.45.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.45.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:output_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:Set meta model
INFO:hf-to-gguf:Set model parameters
INFO:hf-to-gguf:gguf: context length = 131072
INFO:hf-to-gguf:gguf: embedding length = 4096
INFO:hf-to-gguf:gguf: feed forward length = 10944
INFO:hf-to-gguf:gguf: head count = 96
INFO:hf-to-gguf:gguf: key-value head count = 8
WARNING:hf-to-gguf:Unknown RoPE type: default
INFO:hf-to-gguf:gguf: rope scaling type = NONE
INFO:hf-to-gguf:gguf: rope theta = 1000000
INFO:hf-to-gguf:gguf: rms norm epsilon = 1e-05
INFO:hf-to-gguf:gguf: expert count = 128
INFO:hf-to-gguf:gguf: experts used count = 8
INFO:hf-to-gguf:gguf: expert groups count = 1
INFO:hf-to-gguf:gguf: expert groups used count = 1
INFO:hf-to-gguf:gguf: file type = 32
WARNING:gguf.gguf_writer:Duplicated key name 'glm4moe.expert_count', overwriting it with new value 128 of type UINT32
INFO:hf-to-gguf:Set model quantization version
INFO:hf-to-gguf:Set model tokenizer
INFO:gguf.vocab:Adding 318088 merge(s).
INFO:gguf.vocab:Setting special token type eos to 151329
INFO:gguf.vocab:Setting special token type pad to 151329
INFO:gguf.vocab:Setting special token type bos to 151331
INFO:gguf.vocab:Setting special token type eot to 151336
INFO:gguf.vocab:Setting special token type unk to 151329
INFO:gguf.vocab:Setting special token type eom to 151338
INFO:gguf.vocab:Setting chat_template to [gMASK]<sop>
{%- if tools -%}
<|system|>
# Tools
You may call one or more functions to assist with the user query.
You are provided with function signatures within <tools></tools> XML tags:
<tools>
{% for tool in tools %}
{{ tool | tojson(ensure_ascii=False) }}
{% endfor %}
</tools>
For each function call, output the function name and arguments within the following XML format:
<tool_call>{function-name}
<arg_key>{arg-key-1}</arg_key>
<arg_value>{arg-value-1}</arg_value>
<arg_key>{arg-key-2}</arg_key>
<arg_value>{arg-value-2}</arg_value>
...
</tool_call>{%- endif -%}
{%- macro visible_text(content) -%}
 {%- if content is string -%}
 {{- content }}
 {%- elif content is iterable and content is not mapping -%}
 {%- for item in content -%}
 {%- if item is mapping and item.type == 'text' -%}
 {{- item.text }}
 {%- elif item is string -%}
 {{- item }}
 {%- endif -%}
 {%- endfor -%}
 {%- else -%}
 {{- content }}
 {%- endif -%}
{%- endmacro -%}
{%- set ns = namespace(last_user_index=-1) %}
{%- for m in messages %}
 {%- if m.role == 'user' %}
 {% set ns.last_user_index = loop.index0 -%}
 {%- endif %}
{%- endfor %}
{% for m in messages %}
{%- if m.role == 'user' -%}<|user|>
{{ visible_text(m.content) }}
{{- '/nothink' if (enable_thinking is defined and not enable_thinking and not visible_text(m.content).endswith("/nothink")) else '' -}}
{%- elif m.role == 'assistant' -%}
<|assistant|>
{%- set reasoning_content = '' %}
{%- set content = visible_text(m.content) %}
{%- if m.reasoning_content is string %}
 {%- set reasoning_content = m.reasoning_content %}
{%- else %}
 {%- if '</think>' in content %}
 {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
 {%- set content = content.split('</think>')[-1].lstrip('\n') %}
 {%- endif %}
{%- endif %}
{%- if loop.index0 > ns.last_user_index and reasoning_content -%}
{{ '\n<think>' + reasoning_content.strip() + '</think>'}}
{%- else -%}
{{ '\n<think></think>' }}
{%- endif -%}
{%- if content.strip() -%}
{{ '\n' + content.strip() }}
{%- endif -%}
{% if m.tool_calls %}
{% for tc in m.tool_calls %}
{%- if tc.function %}
 {%- set tc = tc.function %}
{%- endif %}
{{ '\n<tool_call>' + tc.name }}
{% set _args = tc.arguments %}
{% for k, v in _args.items() %}
<arg_key>{{ k }}</arg_key>
<arg_value>{{ v | tojson(ensure_ascii=False) if v is not string else v }}</arg_value>
{% endfor %}
</tool_call>{% endfor %}
{% endif %}
{%- elif m.role == 'tool' -%}
{%- if m.content is string -%}
{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
 {{- '<|observation|>' }}
{%- endif %}
{{- '\n<tool_response>\n' }}
{{- m.content }}
{{- '\n</tool_response>' }}
{%- else -%}
<|observation|>{% for tr in m.content %}
<tool_response>
{{ tr.output if tr.output is defined else tr }}
</tool_response>{% endfor -%}
{% endif -%}
{%- elif m.role == 'system' -%}
<|system|>
{{ visible_text(m.content) }}
{%- endif -%}
{%- endfor -%}
{%- if add_generation_prompt -%}
 <|assistant|>{{- '\n<think></think>' if (enable_thinking is defined and not enable_thinking) else '' -}}
{%- endif -%}
INFO:gguf.gguf_writer:Writing the following files:
INFO:gguf.gguf_writer:upload-GLM_4.5_Air/GLM-4.5-Air-BF16-00001-of-00002.gguf: n_tensors = 0, total_size = negligible - metadata only
INFO:gguf.gguf_writer:upload-GLM_4.5_Air/GLM-4.5-Air-BF16-00002-of-00002.gguf: n_tensors = 780, total_size = 213.8G
 Shard (0/2): 0.00byte [00:00, ?byte/s]
 Writing: 0%| | 0.00/214G [00:00<?, ?byte/s]
 Shard (1/2): : 0.00byte [00:00, ?byte/s]
 Shard (1/2): : 0.00byte [00:00, ?byte/s]
 Shard (2/2): : 0.00byte [00:00, ?byte/s]
 Shard (2/2): 0%| | 0.00/214G [00:00<?, ?byte/s]
 Shard (2/2): 1%| | 1.24G/214G [00:02<08:14, 430Mbyte/s]
 Writing: 1%| | 1.24G/214G [00:02<08:14, 430Mbyte/s]
 Shard (2/2): 1%| | 1.72G/214G [00:04<08:18, 425Mbyte/s]
 Writing: 1%| | 1.72G/214G [00:04<08:18, 425Mbyte/s]
 Shard (2/2): 1%|▏ | 3.21G/214G [00:08<09:04, 386Mbyte/s]
 Writing: 1%|▏ | 3.21G/214G [00:08<09:05, 386Mbyte/s]
 Shard (2/2): 2%|▏ | 4.68G/214G [00:11<08:20, 418Mbyte/s]
 Writing: 2%|▏ | 4.68G/214G [00:11<08:20, 418Mbyte/s]
 Shard (2/2): 3%|▎ | 6.16G/214G [00:15<08:26, 410Mbyte/s]
 Writing: 3%|▎ | 6.16G/214G [00:15<08:26, 410Mbyte/s]
 Shard (2/2): 4%|▎ | 7.89G/214G [00:19<08:22, 410Mbyte/s]
 Writing: 4%|▎ | 7.89G/214G [00:19<08:22, 410Mbyte/s]
 Shard (2/2): 4%|▍ | 9.37G/214G [00:22<08:18, 410Mbyte/s]
 Writing: 4%|▍ | 9.37G/214G [00:22<08:18, 410Mbyte/s]
 Shard (2/2): 5%|▌ | 10.8G/214G [00:26<08:06, 417Mbyte/s]
 Writing: 5%|▌ | 10.8G/214G [00:26<08:06, 417Mbyte/s]
 Shard (2/2): 6%|▌ | 12.6G/214G [00:30<08:03, 416Mbyte/s]
 Writing: 6%|▌ | 12.6G/214G [00:30<08:03, 416Mbyte/s]
 Shard (2/2): 7%|▋ | 14.0G/214G [00:33<07:56, 419Mbyte/s]
 Writing: 7%|▋ | 14.0G/214G [00:33<07:56, 419Mbyte/s]
 Shard (2/2): 7%|▋ | 15.5G/214G [00:37<07:44, 427Mbyte/s]
 Writing: 7%|▋ | 15.5G/214G [00:37<07:44, 427Mbyte/s]
 Shard (2/2): 8%|▊ | 17.3G/214G [00:41<07:45, 423Mbyte/s]
 Writing: 8%|▊ | 17.3G/214G [00:41<07:45, 423Mbyte/s]
 Shard (2/2): 9%|▉ | 18.7G/214G [00:44<07:28, 435Mbyte/s]
 Writing: 9%|▉ | 18.7G/214G [00:44<07:28, 435Mbyte/s]
 Shard (2/2): 9%|▉ | 20.2G/214G [00:47<07:22, 437Mbyte/s]
 Writing: 9%|▉ | 20.2G/214G [00:47<07:22, 437Mbyte/s]
 Shard (2/2): 10%|█ | 21.9G/214G [00:51<07:24, 432Mbyte/s]
 Writing: 10%|█ | 21.9G/214G [00:51<07:24, 432Mbyte/s]
 Shard (2/2): 11%|█ | 23.4G/214G [00:55<07:12, 440Mbyte/s]
 Writing: 11%|█ | 23.4G/214G [00:55<07:12, 440Mbyte/s]
 Shard (2/2): 12%|█▏ | 24.9G/214G [00:58<07:17, 432Mbyte/s]
 Writing: 12%|█▏ | 24.9G/214G [00:58<07:17, 432Mbyte/s]
 Shard (2/2): 12%|█▏ | 26.6G/214G [01:03<07:30, 415Mbyte/s]
 Writing: 12%|█▏ | 26.6G/214G [01:03<07:30, 415Mbyte/s]
 Shard (2/2): 13%|█▎ | 28.1G/214G [01:07<07:38, 405Mbyte/s]
 Writing: 13%|█▎ | 28.1G/214G [01:07<07:38, 405Mbyte/s]
 Shard (2/2): 14%|█▍ | 29.6G/214G [01:11<07:45, 396Mbyte/s]
 Writing: 14%|█▍ | 29.6G/214G [01:11<07:45, 396Mbyte/s]
 Shard (2/2): 15%|█▍ | 31.3G/214G [01:16<08:17, 367Mbyte/s]
 Writing: 15%|█▍ | 31.3G/214G [01:16<08:17, 367Mbyte/s]
 Shard (2/2): 15%|█▌ | 32.8G/214G [01:20<08:17, 364Mbyte/s]
 Writing: 15%|█▌ | 32.8G/214G [01:20<08:17, 364Mbyte/s]
 Shard (2/2): 16%|█▌ | 34.3G/214G [01:25<08:38, 346Mbyte/s]
 Writing: 16%|█▌ | 34.3G/214G [01:25<08:38, 346Mbyte/s]
 Shard (2/2): 17%|█▋ | 36.0G/214G [01:29<08:16, 358Mbyte/s]
 Writing: 17%|█▋ | 36.0G/214G [01:29<08:16, 358Mbyte/s]
 Shard (2/2): 18%|█▊ | 37.5G/214G [01:33<08:09, 360Mbyte/s]
 Writing: 18%|█▊ | 37.5G/214G [01:33<08:09, 360Mbyte/s]
 Shard (2/2): 18%|█▊ | 38.9G/214G [01:38<08:05, 360Mbyte/s]
 Writing: 18%|█▊ | 38.9G/214G [01:38<08:05, 360Mbyte/s]
 Shard (2/2): 19%|█▉ | 40.7G/214G [01:43<08:23, 344Mbyte/s]
 Writing: 19%|█▉ | 40.7G/214G [01:43<08:23, 344Mbyte/s]
 Shard (2/2): 20%|█▉ | 42.2G/214G [01:47<08:13, 348Mbyte/s]
 Writing: 20%|█▉ | 42.2G/214G [01:47<08:13, 348Mbyte/s]
 Shard (2/2): 20%|██ | 43.6G/214G [01:51<08:03, 352Mbyte/s]
 Writing: 20%|██ | 43.6G/214G [01:51<08:03, 352Mbyte/s]
 Shard (2/2): 21%|██ | 45.4G/214G [01:56<08:03, 348Mbyte/s]
 Writing: 21%|██ | 45.4G/214G [01:56<08:03, 348Mbyte/s]
 Shard (2/2): 22%|██▏ | 46.8G/214G [02:00<07:46, 358Mbyte/s]
 Writing: 22%|██▏ | 46.8G/214G [02:00<07:46, 358Mbyte/s]
 Shard (2/2): 23%|██▎ | 48.3G/214G [02:04<07:36, 363Mbyte/s]
 Writing: 23%|██▎ | 48.3G/214G [02:04<07:36, 363Mbyte/s]
 Shard (2/2): 23%|██▎ | 50.0G/214G [02:09<07:42, 354Mbyte/s]
 Writing: 23%|██▎ | 50.0G/214G [02:09<07:42, 354Mbyte/s]
 Shard (2/2): 24%|██▍ | 51.5G/214G [02:13<07:22, 367Mbyte/s]
 Writing: 24%|██▍ | 51.5G/214G [02:13<07:22, 367Mbyte/s]
 Shard (2/2): 25%|██▍ | 53.0G/214G [02:17<07:22, 363Mbyte/s]
 Writing: 25%|██▍ | 53.0G/214G [02:17<07:22, 363Mbyte/s]
 Shard (2/2): 26%|██▌ | 54.7G/214G [02:22<07:18, 363Mbyte/s]
 Writing: 26%|██▌ | 54.7G/214G [02:22<07:18, 363Mbyte/s]
 Shard (2/2): 26%|██▋ | 56.2G/214G [02:26<07:14, 363Mbyte/s]
 Writing: 26%|██▋ | 56.2G/214G [02:26<07:14, 363Mbyte/s]
 Shard (2/2): 27%|██▋ | 57.7G/214G [02:30<07:06, 366Mbyte/s]
 Writing: 27%|██▋ | 57.7G/214G [02:30<07:06, 366Mbyte/s]
 Shard (2/2): 28%|██▊ | 59.4G/214G [02:35<07:04, 364Mbyte/s]
 Writing: 28%|██▊ | 59.4G/214G [02:35<07:04, 364Mbyte/s]
 Shard (2/2): 28%|██▊ | 60.9G/214G [02:40<07:49, 326Mbyte/s]
 Writing: 28%|██▊ | 60.9G/214G [02:40<07:49, 326Mbyte/s]
 Shard (2/2): 29%|██▉ | 62.4G/214G [02:46<08:26, 299Mbyte/s]
 Writing: 29%|██▉ | 62.4G/214G [02:46<08:26, 299Mbyte/s]
 Shard (2/2): 30%|██▉ | 64.1G/214G [02:54<09:24, 265Mbyte/s]
 Writing: 30%|██▉ | 64.1G/214G [02:54<09:24, 265Mbyte/s]
 Shard (2/2): 31%|███ | 65.6G/214G [02:59<08:55, 277Mbyte/s]
 Writing: 31%|███ | 65.6G/214G [02:59<08:55, 277Mbyte/s]
 Shard (2/2): 31%|███▏ | 67.1G/214G [03:05<09:11, 266Mbyte/s]
 Writing: 31%|███▏ | 67.1G/214G [03:05<09:11, 266Mbyte/s]
 Shard (2/2): 32%|███▏ | 68.8G/214G [03:10<08:15, 293Mbyte/s]
 Writing: 32%|███▏ | 68.8G/214G [03:10<08:15, 293Mbyte/s]
 Shard (2/2): 33%|███▎ | 70.3G/214G [03:14<07:34, 316Mbyte/s]
 Writing: 33%|███▎ | 70.3G/214G [03:14<07:34, 316Mbyte/s]
 Shard (2/2): 34%|███▎ | 71.7G/214G [03:18<07:22, 321Mbyte/s]
 Writing: 34%|███▎ | 71.7G/214G [03:18<07:22, 321Mbyte/s]
 Shard (2/2): 34%|███▍ | 73.5G/214G [03:23<07:00, 334Mbyte/s]
 Writing: 34%|███▍ | 73.5G/214G [03:23<07:00, 334Mbyte/s]
 Shard (2/2): 35%|███▌ | 74.9G/214G [03:27<06:41, 346Mbyte/s]
 Writing: 35%|███▌ | 74.9G/214G [03:27<06:41, 346Mbyte/s]
 Shard (2/2): 36%|███▌ | 76.4G/214G [03:31<06:34, 348Mbyte/s]
 Writing: 36%|███▌ | 76.4G/214G [03:31<06:34, 348Mbyte/s]
 Shard (2/2): 37%|███▋ | 78.2G/214G [03:36<06:30, 347Mbyte/s]
 Writing: 37%|███▋ | 78.2G/214G [03:36<06:30, 347Mbyte/s]
 Shard (2/2): 37%|███▋ | 79.6G/214G [03:40<06:17, 355Mbyte/s]
 Writing: 37%|███▋ | 79.6G/214G [03:40<06:17, 355Mbyte/s]
 Shard (2/2): 38%|███▊ | 81.1G/214G [03:44<06:20, 348Mbyte/s]
 Writing: 38%|███▊ | 81.1G/214G [03:44<06:20, 348Mbyte/s]
 Shard (2/2): 39%|███▉ | 82.8G/214G [03:50<06:34, 332Mbyte/s]
 Writing: 39%|███▉ | 82.8G/214G [03:50<06:34, 332Mbyte/s]
 Shard (2/2): 39%|███▉ | 84.3G/214G [03:55<06:55, 312Mbyte/s]
 Writing: 39%|███▉ | 84.3G/214G [03:55<06:55, 312Mbyte/s]
 Shard (2/2): 40%|████ | 85.8G/214G [04:01<07:22, 289Mbyte/s]
 Writing: 40%|████ | 85.8G/214G [04:01<07:22, 289Mbyte/s]
 Shard (2/2): 41%|████ | 87.5G/214G [04:07<07:07, 295Mbyte/s]
 Writing: 41%|████ | 87.5G/214G [04:07<07:07, 295Mbyte/s]
 Shard (2/2): 42%|████▏ | 89.0G/214G [04:11<06:41, 311Mbyte/s]
 Writing: 42%|████▏ | 89.0G/214G [04:11<06:41, 311Mbyte/s]
 Shard (2/2): 42%|████▏ | 90.5G/214G [04:17<07:11, 286Mbyte/s]
 Writing: 42%|████▏ | 90.5G/214G [04:17<07:11, 286Mbyte/s]
 Shard (2/2): 43%|████▎ | 92.2G/214G [04:22<06:33, 309Mbyte/s]
 Writing: 43%|████▎ | 92.2G/214G [04:22<06:33, 309Mbyte/s]
 Shard (2/2): 44%|████▍ | 93.7G/214G [04:26<06:15, 320Mbyte/s]
 Writing: 44%|████▍ | 93.7G/214G [04:26<06:15, 320Mbyte/s]
 Shard (2/2): 45%|████▍ | 95.2G/214G [04:31<06:15, 316Mbyte/s]
 Writing: 45%|████▍ | 95.2G/214G [04:31<06:15, 316Mbyte/s]
 Shard (2/2): 45%|████▌ | 96.9G/214G [04:36<06:02, 322Mbyte/s]
 Writing: 45%|████▌ | 96.9G/214G [04:36<06:02, 322Mbyte/s]
 Shard (2/2): 46%|████▌ | 98.4G/214G [04:40<05:41, 338Mbyte/s]
 Writing: 46%|████▌ | 98.4G/214G [04:40<05:41, 338Mbyte/s]
 Shard (2/2): 47%|████▋ | 99.8G/214G [04:44<05:31, 344Mbyte/s]
 Writing: 47%|████▋ | 99.8G/214G [04:44<05:31, 344Mbyte/s]
 Shard (2/2): 48%|████▊ | 102G/214G [04:49<05:19, 351Mbyte/s] 
 Writing: 48%|████▊ | 102G/214G [04:49<05:19, 351Mbyte/s] 
 Shard (2/2): 48%|████▊ | 103G/214G [04:52<05:03, 365Mbyte/s]
 Writing: 48%|████▊ | 103G/214G [04:52<05:03, 365Mbyte/s]
 Shard (2/2): 49%|████▉ | 105G/214G [04:57<05:04, 359Mbyte/s]
 Writing: 49%|████▉ | 105G/214G [04:57<05:04, 359Mbyte/s]
 Shard (2/2): 50%|████▉ | 106G/214G [05:01<04:55, 363Mbyte/s]
 Writing: 50%|████▉ | 106G/214G [05:01<04:55, 363Mbyte/s]
 Shard (2/2): 50%|█████ | 108G/214G [05:06<04:56, 358Mbyte/s]
 Writing: 50%|█████ | 108G/214G [05:06<04:56, 358Mbyte/s]
 Shard (2/2): 51%|█████ | 109G/214G [05:12<05:29, 317Mbyte/s]
 Writing: 51%|█████ | 109G/214G [05:12<05:29, 317Mbyte/s]
 Shard (2/2): 52%|█████▏ | 111G/214G [05:19<05:52, 292Mbyte/s]
 Writing: 52%|█████▏ | 111G/214G [05:19<05:52, 292Mbyte/s]
 Shard (2/2): 53%|█████▎ | 112G/214G [05:22<05:19, 317Mbyte/s]
 Writing: 53%|█████▎ | 112G/214G [05:22<05:19, 317Mbyte/s]
 Shard (2/2): 53%|█████▎ | 114G/214G [05:27<05:20, 311Mbyte/s]
 Writing: 53%|█████▎ | 114G/214G [05:27<05:20, 311Mbyte/s]
 Shard (2/2): 54%|█████▍ | 116G/214G [05:32<05:07, 319Mbyte/s]
 Writing: 54%|█████▍ | 116G/214G [05:32<05:07, 319Mbyte/s]
 Shard (2/2): 55%|█████▍ | 117G/214G [05:36<04:54, 328Mbyte/s]
 Writing: 55%|█████▍ | 117G/214G [05:36<04:54, 328Mbyte/s]
 Shard (2/2): 55%|█████▌ | 119G/214G [05:42<05:10, 307Mbyte/s]
 Writing: 55%|█████▌ | 119G/214G [05:42<05:10, 307Mbyte/s]
 Shard (2/2): 56%|█████▋ | 120G/214G [05:47<04:59, 312Mbyte/s]
 Writing: 56%|█████▋ | 120G/214G [05:47<04:59, 312Mbyte/s]
 Shard (2/2): 57%|█████▋ | 122G/214G [05:51<04:36, 333Mbyte/s]
 Writing: 57%|█████▋ | 122G/214G [05:51<04:36, 333Mbyte/s]
 Shard (2/2): 58%|█████▊ | 123G/214G [05:55<04:25, 341Mbyte/s]
 Writing: 58%|█████▊ | 123G/214G [05:55<04:25, 341Mbyte/s]
 Shard (2/2): 58%|█████▊ | 125G/214G [06:00<04:17, 345Mbyte/s]
 Writing: 58%|█████▊ | 125G/214G [06:00<04:17, 345Mbyte/s]
 Shard (2/2): 59%|█████▉ | 126G/214G [06:04<04:03, 358Mbyte/s]
 Writing: 59%|█████▉ | 126G/214G [06:04<04:03, 358Mbyte/s]
 Shard (2/2): 60%|█████▉ | 128G/214G [06:08<03:59, 358Mbyte/s]
 Writing: 60%|█████▉ | 128G/214G [06:08<03:59, 358Mbyte/s]
 Shard (2/2): 61%|██████ | 130G/214G [06:13<03:58, 352Mbyte/s]
 Writing: 61%|██████ | 130G/214G [06:13<03:58, 352Mbyte/s]
 Shard (2/2): 61%|██████▏ | 131G/214G [06:17<03:50, 358Mbyte/s]
 Writing: 61%|██████▏ | 131G/214G [06:17<03:50, 358Mbyte/s]
 Shard (2/2): 62%|██████▏ | 133G/214G [06:21<03:46, 357Mbyte/s]
 Writing: 62%|██████▏ | 133G/214G [06:21<03:46, 357Mbyte/s]
 Shard (2/2): 63%|██████▎ | 134G/214G [06:28<04:17, 309Mbyte/s]
 Writing: 63%|██████▎ | 134G/214G [06:28<04:17, 309Mbyte/s]
 Shard (2/2): 64%|██████▎ | 136G/214G [06:34<04:30, 288Mbyte/s]
 Writing: 64%|██████▎ | 136G/214G [06:34<04:30, 288Mbyte/s]
 Shard (2/2): 64%|██████▍ | 137G/214G [06:39<04:13, 301Mbyte/s]
 Writing: 64%|██████▍ | 137G/214G [06:39<04:13, 301Mbyte/s]
 Shard (2/2): 65%|██████▌ | 139G/214G [06:44<04:06, 303Mbyte/s]
 Writing: 65%|██████▌ | 139G/214G [06:44<04:06, 303Mbyte/s]
 Shard (2/2): 66%|██████▌ | 141G/214G [06:48<03:44, 327Mbyte/s]
 Writing: 66%|██████▌ | 141G/214G [06:48<03:44, 327Mbyte/s]
 Shard (2/2): 66%|██████▋ | 142G/214G [06:52<03:29, 343Mbyte/s]
 Writing: 66%|██████▋ | 142G/214G [06:52<03:29, 343Mbyte/s]
 Shard (2/2): 67%|██████▋ | 144G/214G [06:58<03:42, 315Mbyte/s]
 Writing: 67%|██████▋ | 144G/214G [06:58<03:42, 315Mbyte/s]
 Shard (2/2): 68%|██████▊ | 145G/214G [07:02<03:22, 339Mbyte/s]
 Writing: 68%|██████▊ | 145G/214G [07:02<03:22, 339Mbyte/s]
 Shard (2/2): 69%|██████▊ | 147G/214G [07:06<03:12, 348Mbyte/s]
 Writing: 69%|██████▊ | 147G/214G [07:06<03:12, 348Mbyte/s]
 Shard (2/2): 69%|██████▉ | 148G/214G [07:10<03:04, 354Mbyte/s]
 Writing: 69%|██████▉ | 148G/214G [07:10<03:04, 354Mbyte/s]
 Shard (2/2): 70%|███████ | 150G/214G [07:14<02:52, 371Mbyte/s]
 Writing: 70%|███████ | 150G/214G [07:14<02:52, 371Mbyte/s]
 Shard (2/2): 71%|███████ | 151G/214G [07:18<02:52, 362Mbyte/s]
 Writing: 71%|███████ | 151G/214G [07:18<02:52, 362Mbyte/s]
 Shard (2/2): 72%|███████▏ | 153G/214G [07:23<02:48, 359Mbyte/s]
 Writing: 72%|███████▏ | 153G/214G [07:23<02:48, 359Mbyte/s]
 Shard (2/2): 72%|███████▏ | 155G/214G [07:27<02:39, 370Mbyte/s]
 Writing: 72%|███████▏ | 155G/214G [07:27<02:39, 370Mbyte/s]
 Shard (2/2): 73%|███████▎ | 156G/214G [07:31<02:39, 362Mbyte/s]
 Writing: 73%|███████▎ | 156G/214G [07:31<02:39, 362Mbyte/s]
 Shard (2/2): 74%|███████▍ | 158G/214G [07:36<02:40, 349Mbyte/s]
 Writing: 74%|███████▍ | 158G/214G [07:36<02:40, 349Mbyte/s]
 Shard (2/2): 75%|███████▍ | 159G/214G [07:42<02:54, 313Mbyte/s]
 Writing: 75%|███████▍ | 159G/214G [07:42<02:54, 313Mbyte/s]
 Shard (2/2): 75%|███████▌ | 161G/214G [07:48<03:00, 294Mbyte/s]
 Writing: 75%|███████▌ | 161G/214G [07:48<03:00, 294Mbyte/s]
 Shard (2/2): 76%|███████▌ | 162G/214G [07:53<02:44, 313Mbyte/s]
 Writing: 76%|███████▌ | 162G/214G [07:53<02:44, 313Mbyte/s]
 Shard (2/2): 77%|███████▋ | 164G/214G [07:57<02:35, 320Mbyte/s]
 Writing: 77%|███████▋ | 164G/214G [07:57<02:35, 320Mbyte/s]
 Shard (2/2): 77%|███████▋ | 165G/214G [08:03<02:46, 291Mbyte/s]
 Writing: 77%|███████▋ | 165G/214G [08:03<02:46, 291Mbyte/s]
 Shard (2/2): 78%|███████▊ | 167G/214G [08:09<02:32, 306Mbyte/s]
 Writing: 78%|███████▊ | 167G/214G [08:09<02:32, 306Mbyte/s]
 Shard (2/2): 79%|███████▉ | 169G/214G [08:12<02:17, 327Mbyte/s]
 Writing: 79%|███████▉ | 169G/214G [08:12<02:17, 327Mbyte/s]
 Shard (2/2): 80%|███████▉ | 170G/214G [08:16<02:08, 340Mbyte/s]
 Writing: 80%|███████▉ | 170G/214G [08:16<02:08, 340Mbyte/s]
 Shard (2/2): 80%|████████ | 172G/214G [08:21<02:00, 348Mbyte/s]
 Writing: 80%|████████ | 172G/214G [08:21<02:00, 348Mbyte/s]
 Shard (2/2): 81%|████████ | 173G/214G [08:25<01:52, 359Mbyte/s]
 Writing: 81%|████████ | 173G/214G [08:25<01:52, 359Mbyte/s]
 Shard (2/2): 82%|████████▏ | 175G/214G [08:29<01:49, 357Mbyte/s]
 Writing: 82%|████████▏ | 175G/214G [08:29<01:49, 357Mbyte/s]
 Shard (2/2): 83%|████████▎ | 177G/214G [08:34<01:45, 354Mbyte/s]
 Writing: 83%|████████▎ | 177G/214G [08:34<01:45, 354Mbyte/s]
 Shard (2/2): 83%|████████▎ | 178G/214G [08:37<01:36, 370Mbyte/s]
 Writing: 83%|████████▎ | 178G/214G [08:37<01:36, 370Mbyte/s]
 Shard (2/2): 84%|████████▍ | 179G/214G [08:41<01:32, 372Mbyte/s]
 Writing: 84%|████████▍ | 179G/214G [08:41<01:32, 372Mbyte/s]
 Shard (2/2): 85%|████████▍ | 181G/214G [08:46<01:28, 368Mbyte/s]
 Writing: 85%|████████▍ | 181G/214G [08:46<01:28, 368Mbyte/s]
 Shard (2/2): 85%|████████▌ | 183G/214G [08:51<01:29, 348Mbyte/s]
 Writing: 85%|████████▌ | 183G/214G [08:51<01:29, 348Mbyte/s]
 Shard (2/2): 86%|████████▌ | 184G/214G [08:57<01:33, 317Mbyte/s]
 Writing: 86%|████████▌ | 184G/214G [08:57<01:33, 317Mbyte/s]
 Shard (2/2): 87%|████████▋ | 186G/214G [09:04<01:37, 287Mbyte/s]
 Writing: 87%|████████▋ | 186G/214G [09:04<01:37, 287Mbyte/s]
 Shard (2/2): 88%|████████▊ | 187G/214G [09:08<01:26, 305Mbyte/s]
 Writing: 88%|████████▊ | 187G/214G [09:08<01:26, 305Mbyte/s]
 Shard (2/2): 88%|████████▊ | 189G/214G [09:13<01:21, 307Mbyte/s]
 Writing: 88%|████████▊ | 189G/214G [09:13<01:21, 307Mbyte/s]
 Shard (2/2): 89%|████████▉ | 191G/214G [09:18<01:15, 308Mbyte/s]
 Writing: 89%|████████▉ | 191G/214G [09:18<01:15, 308Mbyte/s]
 Shard (2/2): 90%|████████▉ | 192G/214G [09:22<01:07, 320Mbyte/s]
 Writing: 90%|████████▉ | 192G/214G [09:22<01:07, 320Mbyte/s]
 Shard (2/2): 91%|█████████ | 194G/214G [09:27<01:01, 329Mbyte/s]
 Writing: 91%|█████████ | 194G/214G [09:27<01:01, 329Mbyte/s]
 Shard (2/2): 91%|█████████▏| 195G/214G [09:33<00:59, 313Mbyte/s]
 Writing: 91%|█████████▏| 195G/214G [09:33<00:59, 313Mbyte/s]
 Shard (2/2): 92%|█████████▏| 197G/214G [09:37<00:52, 324Mbyte/s]
 Writing: 92%|█████████▏| 197G/214G [09:37<00:52, 324Mbyte/s]
 Shard (2/2): 93%|█████████▎| 198G/214G [09:42<00:48, 319Mbyte/s]
 Writing: 93%|█████████▎| 198G/214G [09:42<00:48, 319Mbyte/s]
 Shard (2/2): 94%|█████████▎| 200G/214G [09:46<00:41, 330Mbyte/s]
 Writing: 94%|█████████▎| 200G/214G [09:46<00:41, 330Mbyte/s]
 Shard (2/2): 94%|█████████▍| 201G/214G [09:50<00:35, 348Mbyte/s]
 Writing: 94%|█████████▍| 201G/214G [09:50<00:35, 348Mbyte/s]
 Shard (2/2): 95%|█████████▍| 203G/214G [09:55<00:31, 343Mbyte/s]
 Writing: 95%|█████████▍| 203G/214G [09:55<00:31, 343Mbyte/s]
 Shard (2/2): 96%|█████████▌| 205G/214G [10:00<00:26, 343Mbyte/s]
 Writing: 96%|█████████▌| 205G/214G [10:00<00:26, 343Mbyte/s]
 Shard (2/2): 96%|█████████▋| 206G/214G [10:03<00:21, 356Mbyte/s]
 Writing: 96%|█████████▋| 206G/214G [10:03<00:21, 356Mbyte/s]
 Shard (2/2): 97%|█████████▋| 208G/214G [10:08<00:17, 356Mbyte/s]
 Writing: 97%|█████████▋| 208G/214G [10:08<00:17, 356Mbyte/s]
 Shard (2/2): 98%|█████████▊| 209G/214G [10:11<00:12, 374Mbyte/s]
 Writing: 98%|█████████▊| 209G/214G [10:11<00:12, 374Mbyte/s]
 Shard (2/2): 98%|█████████▊| 211G/214G [10:16<00:09, 346Mbyte/s]
 Writing: 98%|█████████▊| 211G/214G [10:16<00:09, 346Mbyte/s]
 Shard (2/2): 99%|█████████▉| 212G/214G [10:21<00:05, 342Mbyte/s]
 Writing: 99%|█████████▉| 212G/214G [10:21<00:05, 342Mbyte/s]
 Shard (2/2): 100%|█████████▉| 213G/214G [10:26<00:00, 311Mbyte/s]
 Writing: 100%|█████████▉| 213G/214G [10:26<00:00, 311Mbyte/s] Shard (2/2): 100%|██████████| 214G/214G [10:27<00:00, 341Mbyte/s]
 Writing: 100%|██████████| 214G/214G [10:27<00:00, 341Mbyte/s]
INFO:hf-to-gguf:Model successfully exported to upload-GLM_4.5_Air/GLM-4.5-Air-BF16.gguf
+ python3 ./llama.cpp/convert_hf_to_gguf.py ./model-temp-GLM_4.5_Air-PRIMARY --outtype bf16 --outfile ./upload-GLM_4.5_Air/mtp-GLM-4.5-Air-BF16.gguf --mtp --model-name GLM-4.5-Air
INFO:hf-to-gguf:Loading model: model-temp-GLM_4.5_Air-PRIMARY
INFO:hf-to-gguf:Model architecture: Glm4MoeForCausalLM
INFO:hf-to-gguf:gguf: loading model weight map from 'model.safetensors.index.json'
INFO:hf-to-gguf:gguf: indexing model part 'model-00001-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00002-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00003-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00004-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00005-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00006-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00007-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00008-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00009-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00010-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00011-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00012-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00013-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00014-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00015-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00016-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00017-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00018-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00019-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00020-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00021-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00022-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00023-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00024-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00025-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00026-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00027-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00028-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00029-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00030-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00031-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00032-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00033-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00034-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00035-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00036-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00037-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00038-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00039-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00040-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00041-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00042-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00043-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00044-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00045-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00046-of-00047.safetensors'
INFO:hf-to-gguf:gguf: indexing model part 'model-00047-of-00047.safetensors'
INFO:gguf.gguf_writer:gguf: This GGUF file is for Little Endian only
INFO:hf-to-gguf:Exporting model...
INFO:hf-to-gguf:token_embd.weight, torch.bfloat16 --> BF16, shape = {4096, 151552}
INFO:hf-to-gguf:blk.46.nextn.embed_tokens.weight, torch.bfloat16 --> BF16, shape = {4096, 151552}
INFO:hf-to-gguf:output.weight, torch.bfloat16 --> BF16, shape = {4096, 151552}
INFO:hf-to-gguf:blk.46.nextn.shared_head_head.weight, torch.bfloat16 --> BF16, shape = {4096, 151552}
INFO:hf-to-gguf:output_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.46.nextn.eh_proj.weight, torch.bfloat16 --> BF16, shape = {8192, 4096}
INFO:hf-to-gguf:blk.46.nextn.enorm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.46.nextn.hnorm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.46.attn_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.46.ffn_down_exps.weight, torch.bfloat16 --> BF16, shape = {1408, 4096, 128}
INFO:hf-to-gguf:blk.46.ffn_gate_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.46.ffn_up_exps.weight, torch.bfloat16 --> BF16, shape = {4096, 1408, 128}
INFO:hf-to-gguf:blk.46.exp_probs_b.bias, torch.float32 --> F32, shape = {128}
INFO:hf-to-gguf:blk.46.ffn_gate_inp.weight, torch.bfloat16 --> F32, shape = {4096, 128}
INFO:hf-to-gguf:blk.46.ffn_down_shexp.weight, torch.bfloat16 --> BF16, shape = {1408, 4096}
INFO:hf-to-gguf:blk.46.ffn_gate_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.46.ffn_up_shexp.weight, torch.bfloat16 --> BF16, shape = {4096, 1408}
INFO:hf-to-gguf:blk.46.post_attention_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:blk.46.attn_k.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.46.attn_k.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.46.attn_output.weight, torch.bfloat16 --> BF16, shape = {12288, 4096}
INFO:hf-to-gguf:blk.46.attn_q.bias, torch.bfloat16 --> F32, shape = {12288}
INFO:hf-to-gguf:blk.46.attn_q.weight, torch.bfloat16 --> BF16, shape = {4096, 12288}
INFO:hf-to-gguf:blk.46.attn_v.bias, torch.bfloat16 --> F32, shape = {1024}
INFO:hf-to-gguf:blk.46.attn_v.weight, torch.bfloat16 --> BF16, shape = {4096, 1024}
INFO:hf-to-gguf:blk.46.nextn.shared_head_norm.weight, torch.bfloat16 --> F32, shape = {4096}
INFO:hf-to-gguf:Set meta model
INFO:hf-to-gguf:Set model parameters
INFO:hf-to-gguf:gguf: context length = 131072
INFO:hf-to-gguf:gguf: embedding length = 4096
INFO:hf-to-gguf:gguf: feed forward length = 10944
INFO:hf-to-gguf:gguf: head count = 96
INFO:hf-to-gguf:gguf: key-value head count = 8
WARNING:hf-to-gguf:Unknown RoPE type: default
INFO:hf-to-gguf:gguf: rope scaling type = NONE
INFO:hf-to-gguf:gguf: rope theta = 1000000
INFO:hf-to-gguf:gguf: rms norm epsilon = 1e-05
INFO:hf-to-gguf:gguf: expert count = 128
INFO:hf-to-gguf:gguf: experts used count = 8
INFO:hf-to-gguf:gguf: expert groups count = 1
INFO:hf-to-gguf:gguf: expert groups used count = 1
INFO:hf-to-gguf:gguf: file type = 32
WARNING:gguf.gguf_writer:Duplicated key name 'glm4moe.expert_count', overwriting it with new value 128 of type UINT32
INFO:hf-to-gguf:Set model quantization version
INFO:hf-to-gguf:Set model tokenizer
INFO:gguf.vocab:Adding 318088 merge(s).
INFO:gguf.vocab:Setting special token type eos to 151329
INFO:gguf.vocab:Setting special token type pad to 151329
INFO:gguf.vocab:Setting special token type bos to 151331
INFO:gguf.vocab:Setting special token type eot to 151336
INFO:gguf.vocab:Setting special token type unk to 151329
INFO:gguf.vocab:Setting special token type eom to 151338
INFO:gguf.vocab:Setting chat_template to [gMASK]<sop>
{%- if tools -%}
<|system|>
# Tools
You may call one or more functions to assist with the user query.
You are provided with function signatures within <tools></tools> XML tags:
<tools>
{% for tool in tools %}
{{ tool | tojson(ensure_ascii=False) }}
{% endfor %}
</tools>
For each function call, output the function name and arguments within the following XML format:
<tool_call>{function-name}
<arg_key>{arg-key-1}</arg_key>
<arg_value>{arg-value-1}</arg_value>
<arg_key>{arg-key-2}</arg_key>
<arg_value>{arg-value-2}</arg_value>
...
</tool_call>{%- endif -%}
{%- macro visible_text(content) -%}
 {%- if content is string -%}
 {{- content }}
 {%- elif content is iterable and content is not mapping -%}
 {%- for item in content -%}
 {%- if item is mapping and item.type == 'text' -%}
 {{- item.text }}
 {%- elif item is string -%}
 {{- item }}
 {%- endif -%}
 {%- endfor -%}
 {%- else -%}
 {{- content }}
 {%- endif -%}
{%- endmacro -%}
{%- set ns = namespace(last_user_index=-1) %}
{%- for m in messages %}
 {%- if m.role == 'user' %}
 {% set ns.last_user_index = loop.index0 -%}
 {%- endif %}
{%- endfor %}
{% for m in messages %}
{%- if m.role == 'user' -%}<|user|>
{{ visible_text(m.content) }}
{{- '/nothink' if (enable_thinking is defined and not enable_thinking and not visible_text(m.content).endswith("/nothink")) else '' -}}
{%- elif m.role == 'assistant' -%}
<|assistant|>
{%- set reasoning_content = '' %}
{%- set content = visible_text(m.content) %}
{%- if m.reasoning_content is string %}
 {%- set reasoning_content = m.reasoning_content %}
{%- else %}
 {%- if '</think>' in content %}
 {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
 {%- set content = content.split('</think>')[-1].lstrip('\n') %}
 {%- endif %}
{%- endif %}
{%- if loop.index0 > ns.last_user_index and reasoning_content -%}
{{ '\n<think>' + reasoning_content.strip() + '</think>'}}
{%- else -%}
{{ '\n<think></think>' }}
{%- endif -%}
{%- if content.strip() -%}
{{ '\n' + content.strip() }}
{%- endif -%}
{% if m.tool_calls %}
{% for tc in m.tool_calls %}
{%- if tc.function %}
 {%- set tc = tc.function %}
{%- endif %}
{{ '\n<tool_call>' + tc.name }}
{% set _args = tc.arguments %}
{% for k, v in _args.items() %}
<arg_key>{{ k }}</arg_key>
<arg_value>{{ v | tojson(ensure_ascii=False) if v is not string else v }}</arg_value>
{% endfor %}
</tool_call>{% endfor %}
{% endif %}
{%- elif m.role == 'tool' -%}
{%- if m.content is string -%}
{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
 {{- '<|observation|>' }}
{%- endif %}
{{- '\n<tool_response>\n' }}
{{- m.content }}
{{- '\n</tool_response>' }}
{%- else -%}
<|observation|>{% for tr in m.content %}
<tool_response>
{{ tr.output if tr.output is defined else tr }}
</tool_response>{% endfor -%}
{% endif -%}
{%- elif m.role == 'system' -%}
<|system|>
{{ visible_text(m.content) }}
{%- endif -%}
{%- endfor -%}
{%- if add_generation_prompt -%}
 <|assistant|>{{- '\n<think></think>' if (enable_thinking is defined and not enable_thinking) else '' -}}
{%- endif -%}
INFO:gguf.gguf_writer:Writing the following files:
INFO:gguf.gguf_writer:upload-GLM_4.5_Air/mtp-GLM-4.5-Air-BF16.gguf: n_tensors = 26, total_size = 9.7G
 Writing: 0%| | 0.00/9.72G [00:00<?, ?byte/s]
 Writing: 13%|█▎ | 1.24G/9.72G [00:03<00:20, 413Mbyte/s]
 Writing: 26%|██▌ | 2.48G/9.72G [00:06<00:18, 382Mbyte/s]
 Writing: 38%|███▊ | 3.72G/9.72G [00:09<00:15, 394Mbyte/s]
 Writing: 51%|█████ | 4.97G/9.72G [00:12<00:12, 379Mbyte/s]
 Writing: 67%|██████▋ | 6.51G/9.72G [00:18<00:09, 329Mbyte/s]
 Writing: 82%|████████▏ | 7.99G/9.72G [00:22<00:05, 345Mbyte/s]
 Writing: 97%|█████████▋| 9.46G/9.72G [00:26<00:00, 354Mbyte/s] Writing: 100%|██████████| 9.72G/9.72G [00:27<00:00, 355Mbyte/s]
INFO:hf-to-gguf:Model successfully exported to upload-GLM_4.5_Air/mtp-GLM-4.5-Air-BF16.gguf
+ FLAGS_Q4_K_M='--pure --tensor-type token_embd.weight=q8_0 --tensor-type ^output.weight=q6_k --tensor-type ffn_down_exps.weight=q4_0 --tensor-type ffn_down.weight=q8_0 --tensor-type ffn_gate.weight=q8_0 --tensor-type ffn_up.weight=q8_0 --tensor-type shexp=q8_0 --tensor-type attn_=q8_0 '
+ ./llama.cpp/build/bin/llama-quantize --keep-split ./upload-GLM_4.5_Air/GLM-4.5-Air-BF16-00001-of-00002.gguf ./upload-GLM_4.5_Air/GLM-4.5-Air-Q8_0.gguf Q8_0
version: 0.3.0-dev (build 1, commit c1d0e7a)
built with GNU 14.2.0 for Linux x86_64
llama_quantize: quantizing './upload-GLM_4.5_Air/GLM-4.5-Air-BF16-00001-of-00002.gguf' to './upload-GLM_4.5_Air/GLM-4.5-Air-Q8_0' as Q8_0
llama_model_loader: additional 1 GGUFs metadata loaded.
llama_model_loader: loaded meta data with 45 key-value pairs and 780 tensors from ./upload-GLM_4.5_Air/GLM-4.5-Air-BF16-00001-of-00002.gguf (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = glm4moe
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = GLM-4.5-Air
llama_model_loader: - kv 3: general.size_label str = 128x8.0B
llama_model_loader: - kv 4: general.license str = mit
llama_model_loader: - kv 5: general.tags arr[str,1] = ["text-generation"]
llama_model_loader: - kv 6: general.languages arr[str,2] = ["en", "zh"]
llama_model_loader: - kv 7: glm4moe.block_count u32 = 46
llama_model_loader: - kv 8: glm4moe.context_length u32 = 131072
llama_model_loader: - kv 9: glm4moe.embedding_length u32 = 4096
llama_model_loader: - kv 10: glm4moe.feed_forward_length u32 = 10944
llama_model_loader: - kv 11: glm4moe.attention.head_count u32 = 96
llama_model_loader: - kv 12: glm4moe.attention.head_count_kv u32 = 8
llama_model_loader: - kv 13: glm4moe.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 14: glm4moe.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 15: glm4moe.expert_count u32 = 128
llama_model_loader: - kv 16: glm4moe.expert_used_count u32 = 8
llama_model_loader: - kv 17: glm4moe.expert_group_count u32 = 1
llama_model_loader: - kv 18: glm4moe.expert_group_used_count u32 = 1
llama_model_loader: - kv 19: glm4moe.attention.key_length u32 = 128
llama_model_loader: - kv 20: glm4moe.attention.value_length u32 = 128
llama_model_loader: - kv 21: general.file_type u32 = 32
llama_model_loader: - kv 22: glm4moe.rope.dimension_count u32 = 64
llama_model_loader: - kv 23: glm4moe.expert_feed_forward_length u32 = 1408
llama_model_loader: - kv 24: glm4moe.expert_shared_count u32 = 1
llama_model_loader: - kv 25: glm4moe.leading_dense_block_count u32 = 1
llama_model_loader: - kv 26: glm4moe.expert_gating_func u32 = 2
llama_model_loader: - kv 27: glm4moe.expert_weights_scale f32 = 1.000000
llama_model_loader: - kv 28: glm4moe.expert_weights_norm bool = true
llama_model_loader: - kv 29: general.quantization_version u32 = 2
llama_model_loader: - kv 30: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 31: tokenizer.ggml.pre str = glm4
llama_model_loader: - kv 32: tokenizer.ggml.tokens arr[str,151552] = ["!", "\"", "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 33: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 34: tokenizer.ggml.merges arr[str,318088] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "...
llama_model_loader: - kv 35: tokenizer.ggml.eos_token_id u32 = 151329
llama_model_loader: - kv 36: tokenizer.ggml.padding_token_id u32 = 151329
llama_model_loader: - kv 37: tokenizer.ggml.bos_token_id u32 = 151331
llama_model_loader: - kv 38: tokenizer.ggml.eot_token_id u32 = 151336
llama_model_loader: - kv 39: tokenizer.ggml.unknown_token_id u32 = 151329
llama_model_loader: - kv 40: tokenizer.ggml.eom_token_id u32 = 151338
llama_model_loader: - kv 41: tokenizer.chat_template str = [gMASK]<sop>\n{%- if tools -%}\n<|syste...
llama_model_loader: - kv 42: split.no u16 = 0
llama_model_loader: - kv 43: split.count u16 = 2
llama_model_loader: - kv 44: split.tensors.count i32 = 780
llama_model_loader: - type f32: 321 tensors
llama_model_loader: - type bf16: 459 tensors
[ 1/ 780] token_embd.weight - [ 4096, 151552, 1, 1], type = bf16, converting to q8_0 .. size = 1184.00 MiB -> 629.00 MiB
[ 2/ 780] blk.0.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 3/ 780] blk.0.ffn_down.weight - [ 10944, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 85.50 MiB -> 45.42 MiB
[ 4/ 780] blk.0.ffn_gate.weight - [ 4096, 10944, 1, 1], type = bf16, converting to q8_0 .. size = 85.50 MiB -> 45.42 MiB
[ 5/ 780] blk.0.ffn_up.weight - [ 4096, 10944, 1, 1], type = bf16, converting to q8_0 .. size = 85.50 MiB -> 45.42 MiB
[ 6/ 780] blk.0.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 7/ 780] blk.0.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 8/ 780] blk.0.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 9/ 780] blk.0.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 10/ 780] blk.0.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 11/ 780] blk.0.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 12/ 780] blk.0.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 13/ 780] blk.0.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 14/ 780] blk.1.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 15/ 780] blk.1.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 16/ 780] blk.1.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 17/ 780] blk.1.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 18/ 780] blk.1.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 19/ 780] blk.1.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 20/ 780] blk.1.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 21/ 780] blk.1.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 22/ 780] blk.1.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 23/ 780] blk.1.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 24/ 780] blk.1.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 25/ 780] blk.1.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 26/ 780] blk.1.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 27/ 780] blk.1.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 28/ 780] blk.1.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 29/ 780] blk.1.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 30/ 780] blk.1.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 31/ 780] blk.2.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 32/ 780] blk.2.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 33/ 780] blk.2.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 34/ 780] blk.2.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 35/ 780] blk.2.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 36/ 780] blk.2.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 37/ 780] blk.2.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 38/ 780] blk.2.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 39/ 780] blk.2.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 40/ 780] blk.2.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 41/ 780] blk.2.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 42/ 780] blk.2.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 43/ 780] blk.2.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 44/ 780] blk.2.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 45/ 780] blk.2.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 46/ 780] blk.2.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 47/ 780] blk.2.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 48/ 780] blk.3.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 49/ 780] blk.3.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 50/ 780] blk.3.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 51/ 780] blk.3.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 52/ 780] blk.3.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 53/ 780] blk.3.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 54/ 780] blk.3.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 55/ 780] blk.3.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 56/ 780] blk.3.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 57/ 780] blk.3.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 58/ 780] blk.3.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 59/ 780] blk.3.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 60/ 780] blk.3.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 61/ 780] blk.3.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 62/ 780] blk.3.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 63/ 780] blk.3.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 64/ 780] blk.3.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 65/ 780] blk.4.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 66/ 780] blk.4.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 67/ 780] blk.4.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 68/ 780] blk.4.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 69/ 780] blk.4.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 70/ 780] blk.4.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 71/ 780] blk.4.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 72/ 780] blk.4.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 73/ 780] blk.4.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 74/ 780] blk.4.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 75/ 780] blk.4.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 76/ 780] blk.4.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 77/ 780] blk.4.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 78/ 780] blk.4.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 79/ 780] blk.4.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 80/ 780] blk.4.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 81/ 780] blk.4.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 82/ 780] blk.5.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 83/ 780] blk.5.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 84/ 780] blk.5.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 85/ 780] blk.5.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 86/ 780] blk.5.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 87/ 780] blk.5.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 88/ 780] blk.5.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 89/ 780] blk.5.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 90/ 780] blk.5.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 91/ 780] blk.5.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 92/ 780] blk.5.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 93/ 780] blk.5.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 94/ 780] blk.5.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 95/ 780] blk.5.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 96/ 780] blk.5.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 97/ 780] blk.5.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 98/ 780] blk.5.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 99/ 780] blk.6.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 100/ 780] blk.6.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 101/ 780] blk.6.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 102/ 780] blk.6.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 103/ 780] blk.6.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 104/ 780] blk.6.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 105/ 780] blk.6.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 106/ 780] blk.6.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 107/ 780] blk.6.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 108/ 780] blk.6.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 109/ 780] blk.6.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 110/ 780] blk.6.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 111/ 780] blk.6.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 112/ 780] blk.6.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 113/ 780] blk.6.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 114/ 780] blk.6.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 115/ 780] blk.6.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 116/ 780] blk.7.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 117/ 780] blk.7.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 118/ 780] blk.7.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 119/ 780] blk.7.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 120/ 780] blk.7.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 121/ 780] blk.7.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 122/ 780] blk.7.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 123/ 780] blk.7.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 124/ 780] blk.7.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 125/ 780] blk.7.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 126/ 780] blk.7.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 127/ 780] blk.7.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 128/ 780] blk.7.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 129/ 780] blk.7.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 130/ 780] blk.7.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 131/ 780] blk.7.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 132/ 780] blk.7.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 133/ 780] blk.8.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 134/ 780] blk.8.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 135/ 780] blk.8.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 136/ 780] blk.8.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 137/ 780] blk.8.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 138/ 780] blk.8.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 139/ 780] blk.8.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 140/ 780] blk.8.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 141/ 780] blk.8.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 142/ 780] blk.8.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 143/ 780] blk.8.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 144/ 780] blk.8.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 145/ 780] blk.8.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 146/ 780] blk.8.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 147/ 780] blk.8.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 148/ 780] blk.8.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 149/ 780] blk.8.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 150/ 780] blk.9.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 151/ 780] blk.9.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 152/ 780] blk.9.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 153/ 780] blk.9.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 154/ 780] blk.9.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 155/ 780] blk.9.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 156/ 780] blk.9.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 157/ 780] blk.9.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 158/ 780] blk.9.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 159/ 780] blk.9.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 160/ 780] blk.9.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 161/ 780] blk.9.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 162/ 780] blk.9.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 163/ 780] blk.9.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 164/ 780] blk.9.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 165/ 780] blk.9.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 166/ 780] blk.9.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 167/ 780] blk.10.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 168/ 780] blk.10.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 169/ 780] blk.10.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 170/ 780] blk.10.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 171/ 780] blk.10.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 172/ 780] blk.10.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 173/ 780] blk.10.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 174/ 780] blk.10.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 175/ 780] blk.10.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 176/ 780] blk.10.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 177/ 780] blk.10.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 178/ 780] blk.10.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 179/ 780] blk.10.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 180/ 780] blk.10.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 181/ 780] blk.10.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 182/ 780] blk.10.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 183/ 780] blk.10.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 184/ 780] blk.11.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 185/ 780] blk.11.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 186/ 780] blk.11.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 187/ 780] blk.11.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 188/ 780] blk.11.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 189/ 780] blk.11.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 190/ 780] blk.11.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 191/ 780] blk.11.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 192/ 780] blk.11.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 193/ 780] blk.11.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 194/ 780] blk.11.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 195/ 780] blk.11.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 196/ 780] blk.11.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 197/ 780] blk.11.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 198/ 780] blk.11.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 199/ 780] blk.11.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 200/ 780] blk.11.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 201/ 780] blk.12.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 202/ 780] blk.12.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 203/ 780] blk.12.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 204/ 780] blk.12.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 205/ 780] blk.12.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 206/ 780] blk.12.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 207/ 780] blk.12.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 208/ 780] blk.12.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 209/ 780] blk.12.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 210/ 780] blk.12.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 211/ 780] blk.12.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 212/ 780] blk.12.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 213/ 780] blk.12.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 214/ 780] blk.12.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 215/ 780] blk.12.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 216/ 780] blk.12.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 217/ 780] blk.12.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 218/ 780] blk.13.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 219/ 780] blk.13.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 220/ 780] blk.13.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 221/ 780] blk.13.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 222/ 780] blk.13.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 223/ 780] blk.13.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 224/ 780] blk.13.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 225/ 780] blk.13.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 226/ 780] blk.13.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 227/ 780] blk.13.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 228/ 780] blk.13.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 229/ 780] blk.13.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 230/ 780] blk.13.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 231/ 780] blk.13.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 232/ 780] blk.13.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 233/ 780] blk.13.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 234/ 780] blk.13.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 235/ 780] blk.14.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 236/ 780] blk.14.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 237/ 780] blk.14.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 238/ 780] blk.14.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 239/ 780] blk.14.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 240/ 780] blk.14.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 241/ 780] blk.14.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 242/ 780] blk.14.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 243/ 780] blk.14.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 244/ 780] blk.14.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 245/ 780] blk.14.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 246/ 780] blk.14.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 247/ 780] blk.14.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 248/ 780] blk.14.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 249/ 780] blk.14.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 250/ 780] blk.14.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 251/ 780] blk.14.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 252/ 780] blk.15.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 253/ 780] blk.15.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 254/ 780] blk.15.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 255/ 780] blk.15.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 256/ 780] blk.15.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 257/ 780] blk.15.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 258/ 780] blk.15.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 259/ 780] blk.15.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 260/ 780] blk.15.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 261/ 780] blk.15.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 262/ 780] blk.15.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 263/ 780] blk.15.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 264/ 780] blk.15.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 265/ 780] blk.15.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 266/ 780] blk.15.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 267/ 780] blk.15.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 268/ 780] blk.15.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 269/ 780] blk.16.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 270/ 780] blk.16.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 271/ 780] blk.16.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 272/ 780] blk.16.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 273/ 780] blk.16.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 274/ 780] blk.16.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 275/ 780] blk.16.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 276/ 780] blk.16.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 277/ 780] blk.16.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 278/ 780] blk.16.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 279/ 780] blk.16.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 280/ 780] blk.16.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 281/ 780] blk.16.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 282/ 780] blk.16.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 283/ 780] blk.16.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 284/ 780] blk.16.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 285/ 780] blk.16.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 286/ 780] blk.17.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 287/ 780] blk.17.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 288/ 780] blk.17.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 289/ 780] blk.17.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 290/ 780] blk.17.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 291/ 780] blk.17.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 292/ 780] blk.17.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 293/ 780] blk.17.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 294/ 780] blk.17.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 295/ 780] blk.17.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 296/ 780] blk.17.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 297/ 780] blk.17.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 298/ 780] blk.17.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 299/ 780] blk.17.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 300/ 780] blk.17.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 301/ 780] blk.17.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 302/ 780] blk.17.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 303/ 780] blk.18.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 304/ 780] blk.18.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 305/ 780] blk.18.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 306/ 780] blk.18.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 307/ 780] blk.18.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 308/ 780] blk.18.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 309/ 780] blk.18.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 310/ 780] blk.18.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 311/ 780] blk.18.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 312/ 780] blk.18.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 313/ 780] blk.18.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 314/ 780] blk.18.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 315/ 780] blk.18.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 316/ 780] blk.18.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 317/ 780] blk.18.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 318/ 780] blk.18.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 319/ 780] blk.18.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 320/ 780] blk.19.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 321/ 780] blk.19.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 322/ 780] blk.19.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 323/ 780] blk.19.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 324/ 780] blk.19.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 325/ 780] blk.19.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 326/ 780] blk.19.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 327/ 780] blk.19.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 328/ 780] blk.19.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 329/ 780] blk.19.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 330/ 780] blk.19.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 331/ 780] blk.19.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 332/ 780] blk.19.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 333/ 780] blk.19.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 334/ 780] blk.19.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 335/ 780] blk.19.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 336/ 780] blk.19.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 337/ 780] blk.20.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 338/ 780] blk.20.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 339/ 780] blk.20.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 340/ 780] blk.20.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 341/ 780] blk.20.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 342/ 780] blk.20.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 343/ 780] blk.20.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 344/ 780] blk.20.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 345/ 780] blk.20.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 346/ 780] blk.20.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 347/ 780] blk.20.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 348/ 780] blk.20.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 349/ 780] blk.20.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 350/ 780] blk.20.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 351/ 780] blk.20.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 352/ 780] blk.20.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 353/ 780] blk.20.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 354/ 780] blk.21.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 355/ 780] blk.21.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 356/ 780] blk.21.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 357/ 780] blk.21.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 358/ 780] blk.21.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 359/ 780] blk.21.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 360/ 780] blk.21.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 361/ 780] blk.21.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 362/ 780] blk.21.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 363/ 780] blk.21.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 364/ 780] blk.21.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 365/ 780] blk.21.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 366/ 780] blk.21.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 367/ 780] blk.21.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 368/ 780] blk.21.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 369/ 780] blk.21.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 370/ 780] blk.21.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 371/ 780] blk.22.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 372/ 780] blk.22.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 373/ 780] blk.22.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 374/ 780] blk.22.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 375/ 780] blk.22.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 376/ 780] blk.22.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 377/ 780] blk.22.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 378/ 780] blk.22.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 379/ 780] blk.22.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 380/ 780] blk.22.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 381/ 780] blk.22.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 382/ 780] blk.22.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 383/ 780] blk.22.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 384/ 780] blk.22.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 385/ 780] blk.22.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 386/ 780] blk.22.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 387/ 780] blk.22.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 388/ 780] blk.23.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 389/ 780] blk.23.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 390/ 780] blk.23.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 391/ 780] blk.23.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 392/ 780] blk.23.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 393/ 780] blk.23.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 394/ 780] blk.23.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 395/ 780] blk.23.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 396/ 780] blk.23.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 397/ 780] blk.23.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 398/ 780] blk.23.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 399/ 780] blk.23.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 400/ 780] blk.23.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 401/ 780] blk.23.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 402/ 780] blk.23.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 403/ 780] blk.23.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 404/ 780] blk.23.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 405/ 780] blk.24.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 406/ 780] blk.24.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 407/ 780] blk.24.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 408/ 780] blk.24.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 409/ 780] blk.24.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 410/ 780] blk.24.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 411/ 780] blk.24.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 412/ 780] blk.24.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 413/ 780] blk.24.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 414/ 780] blk.24.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 415/ 780] blk.24.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 416/ 780] blk.24.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 417/ 780] blk.24.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 418/ 780] blk.24.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 419/ 780] blk.24.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 420/ 780] blk.24.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 421/ 780] blk.24.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 422/ 780] blk.25.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 423/ 780] blk.25.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 424/ 780] blk.25.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 425/ 780] blk.25.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 426/ 780] blk.25.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 427/ 780] blk.25.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 428/ 780] blk.25.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 429/ 780] blk.25.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 430/ 780] blk.25.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 431/ 780] blk.25.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 432/ 780] blk.25.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 433/ 780] blk.25.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 434/ 780] blk.25.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 435/ 780] blk.25.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 436/ 780] blk.25.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 437/ 780] blk.25.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 438/ 780] blk.25.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 439/ 780] blk.26.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 440/ 780] blk.26.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 441/ 780] blk.26.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 442/ 780] blk.26.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 443/ 780] blk.26.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 444/ 780] blk.26.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 445/ 780] blk.26.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 446/ 780] blk.26.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 447/ 780] blk.26.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 448/ 780] blk.26.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 449/ 780] blk.26.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 450/ 780] blk.26.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 451/ 780] blk.26.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 452/ 780] blk.26.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 453/ 780] blk.26.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 454/ 780] blk.26.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 455/ 780] blk.26.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 456/ 780] blk.27.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 457/ 780] blk.27.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 458/ 780] blk.27.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 459/ 780] blk.27.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 460/ 780] blk.27.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 461/ 780] blk.27.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 462/ 780] blk.27.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 463/ 780] blk.27.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 464/ 780] blk.27.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 465/ 780] blk.27.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 466/ 780] blk.27.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 467/ 780] blk.27.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 468/ 780] blk.27.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 469/ 780] blk.27.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 470/ 780] blk.27.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 471/ 780] blk.27.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 472/ 780] blk.27.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 473/ 780] blk.28.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 474/ 780] blk.28.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 475/ 780] blk.28.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 476/ 780] blk.28.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 477/ 780] blk.28.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 478/ 780] blk.28.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 479/ 780] blk.28.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 480/ 780] blk.28.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 481/ 780] blk.28.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 482/ 780] blk.28.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 483/ 780] blk.28.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 484/ 780] blk.28.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 485/ 780] blk.28.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 486/ 780] blk.28.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 487/ 780] blk.28.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 488/ 780] blk.28.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 489/ 780] blk.28.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 490/ 780] blk.29.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 491/ 780] blk.29.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 492/ 780] blk.29.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 493/ 780] blk.29.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 494/ 780] blk.29.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 495/ 780] blk.29.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 496/ 780] blk.29.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 497/ 780] blk.29.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 498/ 780] blk.29.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 499/ 780] blk.29.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 500/ 780] blk.29.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 501/ 780] blk.29.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 502/ 780] blk.29.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 503/ 780] blk.29.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 504/ 780] blk.29.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 505/ 780] blk.29.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 506/ 780] blk.29.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 507/ 780] blk.30.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 508/ 780] blk.30.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 509/ 780] blk.30.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 510/ 780] blk.30.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 511/ 780] blk.30.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 512/ 780] blk.30.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 513/ 780] blk.30.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 514/ 780] blk.30.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 515/ 780] blk.30.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 516/ 780] blk.30.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 517/ 780] blk.30.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 518/ 780] blk.30.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 519/ 780] blk.30.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 520/ 780] blk.30.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 521/ 780] blk.30.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 522/ 780] blk.30.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 523/ 780] blk.30.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 524/ 780] blk.31.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 525/ 780] blk.31.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 526/ 780] blk.31.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 527/ 780] blk.31.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 528/ 780] blk.31.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 529/ 780] blk.31.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 530/ 780] blk.31.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 531/ 780] blk.31.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 532/ 780] blk.31.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 533/ 780] blk.31.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 534/ 780] blk.31.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 535/ 780] blk.31.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 536/ 780] blk.31.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 537/ 780] blk.31.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 538/ 780] blk.31.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 539/ 780] blk.31.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 540/ 780] blk.31.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 541/ 780] blk.32.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 542/ 780] blk.32.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 543/ 780] blk.32.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 544/ 780] blk.32.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 545/ 780] blk.32.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 546/ 780] blk.32.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 547/ 780] blk.32.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 548/ 780] blk.32.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 549/ 780] blk.32.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 550/ 780] blk.32.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 551/ 780] blk.32.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 552/ 780] blk.32.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 553/ 780] blk.32.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 554/ 780] blk.32.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 555/ 780] blk.32.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 556/ 780] blk.32.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 557/ 780] blk.32.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 558/ 780] blk.33.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 559/ 780] blk.33.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 560/ 780] blk.33.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 561/ 780] blk.33.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 562/ 780] blk.33.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 563/ 780] blk.33.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 564/ 780] blk.33.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 565/ 780] blk.33.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 566/ 780] blk.33.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 567/ 780] blk.33.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 568/ 780] blk.33.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 569/ 780] blk.33.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 570/ 780] blk.33.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 571/ 780] blk.33.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 572/ 780] blk.33.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 573/ 780] blk.33.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 574/ 780] blk.33.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 575/ 780] blk.34.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 576/ 780] blk.34.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 577/ 780] blk.34.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 578/ 780] blk.34.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 579/ 780] blk.34.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 580/ 780] blk.34.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 581/ 780] blk.34.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 582/ 780] blk.34.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 583/ 780] blk.34.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 584/ 780] blk.34.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 585/ 780] blk.34.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 586/ 780] blk.34.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 587/ 780] blk.34.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 588/ 780] blk.34.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 589/ 780] blk.34.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 590/ 780] blk.34.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 591/ 780] blk.34.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 592/ 780] blk.35.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 593/ 780] blk.35.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 594/ 780] blk.35.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 595/ 780] blk.35.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 596/ 780] blk.35.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 597/ 780] blk.35.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 598/ 780] blk.35.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 599/ 780] blk.35.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 600/ 780] blk.35.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 601/ 780] blk.35.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 602/ 780] blk.35.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 603/ 780] blk.35.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 604/ 780] blk.35.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 605/ 780] blk.35.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 606/ 780] blk.35.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 607/ 780] blk.35.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 608/ 780] blk.35.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 609/ 780] blk.36.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 610/ 780] blk.36.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 611/ 780] blk.36.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 612/ 780] blk.36.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 613/ 780] blk.36.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 614/ 780] blk.36.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 615/ 780] blk.36.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 616/ 780] blk.36.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 617/ 780] blk.36.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 618/ 780] blk.36.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 619/ 780] blk.36.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 620/ 780] blk.36.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 621/ 780] blk.36.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 622/ 780] blk.36.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 623/ 780] blk.36.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 624/ 780] blk.36.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 625/ 780] blk.36.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 626/ 780] blk.37.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 627/ 780] blk.37.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 628/ 780] blk.37.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 629/ 780] blk.37.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 630/ 780] blk.37.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 631/ 780] blk.37.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 632/ 780] blk.37.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 633/ 780] blk.37.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 634/ 780] blk.37.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 635/ 780] blk.37.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 636/ 780] blk.37.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 637/ 780] blk.37.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 638/ 780] blk.37.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 639/ 780] blk.37.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 640/ 780] blk.37.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 641/ 780] blk.37.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 642/ 780] blk.37.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 643/ 780] blk.38.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 644/ 780] blk.38.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 645/ 780] blk.38.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 646/ 780] blk.38.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 647/ 780] blk.38.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 648/ 780] blk.38.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 649/ 780] blk.38.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 650/ 780] blk.38.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 651/ 780] blk.38.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 652/ 780] blk.38.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 653/ 780] blk.38.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 654/ 780] blk.38.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 655/ 780] blk.38.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 656/ 780] blk.38.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 657/ 780] blk.38.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 658/ 780] blk.38.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 659/ 780] blk.38.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 660/ 780] blk.39.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 661/ 780] blk.39.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 662/ 780] blk.39.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 663/ 780] blk.39.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 664/ 780] blk.39.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 665/ 780] blk.39.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 666/ 780] blk.39.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 667/ 780] blk.39.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 668/ 780] blk.39.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 669/ 780] blk.39.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 670/ 780] blk.39.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 671/ 780] blk.39.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 672/ 780] blk.39.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 673/ 780] blk.39.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 674/ 780] blk.39.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 675/ 780] blk.39.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 676/ 780] blk.39.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 677/ 780] blk.40.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 678/ 780] blk.40.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 679/ 780] blk.40.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 680/ 780] blk.40.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 681/ 780] blk.40.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 682/ 780] blk.40.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 683/ 780] blk.40.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 684/ 780] blk.40.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 685/ 780] blk.40.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 686/ 780] blk.40.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 687/ 780] blk.40.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 688/ 780] blk.40.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 689/ 780] blk.40.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 690/ 780] blk.40.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 691/ 780] blk.40.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 692/ 780] blk.40.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 693/ 780] blk.40.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 694/ 780] blk.41.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 695/ 780] blk.41.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 696/ 780] blk.41.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 697/ 780] blk.41.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 698/ 780] blk.41.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 699/ 780] blk.41.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 700/ 780] blk.41.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 701/ 780] blk.41.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 702/ 780] blk.41.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 703/ 780] blk.41.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 704/ 780] blk.41.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 705/ 780] blk.41.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 706/ 780] blk.41.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 707/ 780] blk.41.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 708/ 780] blk.41.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 709/ 780] blk.41.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 710/ 780] blk.41.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 711/ 780] blk.42.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 712/ 780] blk.42.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 713/ 780] blk.42.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 714/ 780] blk.42.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 715/ 780] blk.42.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 716/ 780] blk.42.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 717/ 780] blk.42.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 718/ 780] blk.42.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 719/ 780] blk.42.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 720/ 780] blk.42.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 721/ 780] blk.42.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 722/ 780] blk.42.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 723/ 780] blk.42.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 724/ 780] blk.42.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 725/ 780] blk.42.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 726/ 780] blk.42.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 727/ 780] blk.42.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 728/ 780] blk.43.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 729/ 780] blk.43.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 730/ 780] blk.43.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 731/ 780] blk.43.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 732/ 780] blk.43.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 733/ 780] blk.43.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 734/ 780] blk.43.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 735/ 780] blk.43.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 736/ 780] blk.43.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 737/ 780] blk.43.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 738/ 780] blk.43.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 739/ 780] blk.43.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 740/ 780] blk.43.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 741/ 780] blk.43.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 742/ 780] blk.43.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 743/ 780] blk.43.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 744/ 780] blk.43.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 745/ 780] blk.44.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 746/ 780] blk.44.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 747/ 780] blk.44.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 748/ 780] blk.44.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 749/ 780] blk.44.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 750/ 780] blk.44.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 751/ 780] blk.44.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 752/ 780] blk.44.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 753/ 780] blk.44.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 754/ 780] blk.44.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 755/ 780] blk.44.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 756/ 780] blk.44.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 757/ 780] blk.44.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 758/ 780] blk.44.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 759/ 780] blk.44.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 760/ 780] blk.44.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 761/ 780] blk.44.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 762/ 780] output.weight - [ 4096, 151552, 1, 1], type = bf16, converting to q8_0 .. size = 1184.00 MiB -> 629.00 MiB
[ 763/ 780] blk.45.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 764/ 780] blk.45.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 765/ 780] blk.45.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 766/ 780] blk.45.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 767/ 780] blk.45.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 768/ 780] blk.45.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 769/ 780] blk.45.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 770/ 780] blk.45.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 771/ 780] blk.45.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 772/ 780] blk.45.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 773/ 780] blk.45.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 774/ 780] blk.45.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 775/ 780] blk.45.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 776/ 780] blk.45.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 777/ 780] blk.45.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 778/ 780] blk.45.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 779/ 780] blk.45.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 780/ 780] output_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
llama_model_quantize_impl: model size = 203851.49 MiB (16.00 BPW)
llama_model_quantize_impl: quant size = 108340.16 MiB (8.51 BPW)
llama_quantize: quantize time = 759559.41 ms
llama_quantize: total time = 759559.41 ms
+ ./llama.cpp/build/bin/llama-quantize --keep-split --pure --tensor-type token_embd.weight=q8_0 --tensor-type '^output.weight=q6_k' --tensor-type ffn_down_exps.weight=q4_0 --tensor-type ffn_down.weight=q8_0 --tensor-type ffn_gate.weight=q8_0 --tensor-type ffn_up.weight=q8_0 --tensor-type shexp=q8_0 --tensor-type attn_=q8_0 ./upload-GLM_4.5_Air/GLM-4.5-Air-BF16-00001-of-00002.gguf ./upload-GLM_4.5_Air/GLM-4.5-Air-Q4_K_M.gguf Q4_K_M
version: 0.3.0-dev (build 1, commit c1d0e7a)
built with GNU 14.2.0 for Linux x86_64
llama_quantize: quantizing './upload-GLM_4.5_Air/GLM-4.5-Air-BF16-00001-of-00002.gguf' to './upload-GLM_4.5_Air/GLM-4.5-Air-Q4_K_M' as Q4_K_M
llama_model_loader: additional 1 GGUFs metadata loaded.
llama_model_loader: loaded meta data with 45 key-value pairs and 780 tensors from ./upload-GLM_4.5_Air/GLM-4.5-Air-BF16-00001-of-00002.gguf (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = glm4moe
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = GLM-4.5-Air
llama_model_loader: - kv 3: general.size_label str = 128x8.0B
llama_model_loader: - kv 4: general.license str = mit
llama_model_loader: - kv 5: general.tags arr[str,1] = ["text-generation"]
llama_model_loader: - kv 6: general.languages arr[str,2] = ["en", "zh"]
llama_model_loader: - kv 7: glm4moe.block_count u32 = 46
llama_model_loader: - kv 8: glm4moe.context_length u32 = 131072
llama_model_loader: - kv 9: glm4moe.embedding_length u32 = 4096
llama_model_loader: - kv 10: glm4moe.feed_forward_length u32 = 10944
llama_model_loader: - kv 11: glm4moe.attention.head_count u32 = 96
llama_model_loader: - kv 12: glm4moe.attention.head_count_kv u32 = 8
llama_model_loader: - kv 13: glm4moe.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 14: glm4moe.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 15: glm4moe.expert_count u32 = 128
llama_model_loader: - kv 16: glm4moe.expert_used_count u32 = 8
llama_model_loader: - kv 17: glm4moe.expert_group_count u32 = 1
llama_model_loader: - kv 18: glm4moe.expert_group_used_count u32 = 1
llama_model_loader: - kv 19: glm4moe.attention.key_length u32 = 128
llama_model_loader: - kv 20: glm4moe.attention.value_length u32 = 128
llama_model_loader: - kv 21: general.file_type u32 = 32
llama_model_loader: - kv 22: glm4moe.rope.dimension_count u32 = 64
llama_model_loader: - kv 23: glm4moe.expert_feed_forward_length u32 = 1408
llama_model_loader: - kv 24: glm4moe.expert_shared_count u32 = 1
llama_model_loader: - kv 25: glm4moe.leading_dense_block_count u32 = 1
llama_model_loader: - kv 26: glm4moe.expert_gating_func u32 = 2
llama_model_loader: - kv 27: glm4moe.expert_weights_scale f32 = 1.000000
llama_model_loader: - kv 28: glm4moe.expert_weights_norm bool = true
llama_model_loader: - kv 29: general.quantization_version u32 = 2
llama_model_loader: - kv 30: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 31: tokenizer.ggml.pre str = glm4
llama_model_loader: - kv 32: tokenizer.ggml.tokens arr[str,151552] = ["!", "\"", "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 33: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 34: tokenizer.ggml.merges arr[str,318088] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "...
llama_model_loader: - kv 35: tokenizer.ggml.eos_token_id u32 = 151329
llama_model_loader: - kv 36: tokenizer.ggml.padding_token_id u32 = 151329
llama_model_loader: - kv 37: tokenizer.ggml.bos_token_id u32 = 151331
llama_model_loader: - kv 38: tokenizer.ggml.eot_token_id u32 = 151336
llama_model_loader: - kv 39: tokenizer.ggml.unknown_token_id u32 = 151329
llama_model_loader: - kv 40: tokenizer.ggml.eom_token_id u32 = 151338
llama_model_loader: - kv 41: tokenizer.chat_template str = [gMASK]<sop>\n{%- if tools -%}\n<|syste...
llama_model_loader: - kv 42: split.no u16 = 0
llama_model_loader: - kv 43: split.count u16 = 2
llama_model_loader: - kv 44: split.tensors.count i32 = 780
llama_model_loader: - type f32: 321 tensors
llama_model_loader: - type bf16: 459 tensors
llama_tensor_get_type: token_embd.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.0.ffn_down.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.0.ffn_gate.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.0.ffn_up.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.0.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.0.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.0.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.0.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.1.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.1.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.1.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.1.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.1.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.1.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.1.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.1.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.2.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.2.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.2.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.2.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.2.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.2.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.2.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.2.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.3.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.3.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.3.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.3.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.3.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.3.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.3.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.3.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.4.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.4.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.4.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.4.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.4.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.4.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.4.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.4.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.5.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.5.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.5.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.5.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.5.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.5.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.5.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.5.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.6.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.6.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.6.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.6.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.6.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.6.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.6.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.6.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.7.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.7.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.7.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.7.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.7.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.7.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.7.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.7.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.8.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.8.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.8.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.8.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.8.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.8.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.8.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.8.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.9.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.9.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.9.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.9.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.9.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.9.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.9.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.9.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.10.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.10.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.10.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.10.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.10.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.10.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.10.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.10.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.11.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.11.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.11.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.11.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.11.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.11.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.11.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.11.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.12.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.12.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.12.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.12.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.12.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.12.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.12.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.12.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.13.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.13.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.13.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.13.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.13.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.13.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.13.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.13.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.14.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.14.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.14.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.14.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.14.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.14.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.14.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.14.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.15.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.15.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.15.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.15.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.15.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.15.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.15.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.15.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.16.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.16.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.16.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.16.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.16.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.16.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.16.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.16.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.17.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.17.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.17.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.17.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.17.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.17.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.17.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.17.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.18.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.18.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.18.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.18.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.18.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.18.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.18.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.18.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.19.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.19.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.19.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.19.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.19.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.19.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.19.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.19.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.20.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.20.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.20.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.20.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.20.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.20.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.20.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.20.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.21.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.21.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.21.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.21.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.21.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.21.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.21.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.21.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.22.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.22.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.22.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.22.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.22.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.22.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.22.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.22.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.23.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.23.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.23.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.23.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.23.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.23.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.23.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.23.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.24.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.24.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.24.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.24.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.24.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.24.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.24.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.24.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.25.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.25.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.25.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.25.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.25.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.25.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.25.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.25.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.26.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.26.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.26.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.26.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.26.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.26.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.26.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.26.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.27.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.27.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.27.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.27.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.27.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.27.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.27.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.27.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.28.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.28.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.28.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.28.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.28.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.28.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.28.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.28.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.29.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.29.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.29.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.29.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.29.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.29.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.29.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.29.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.30.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.30.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.30.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.30.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.30.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.30.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.30.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.30.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.31.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.31.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.31.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.31.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.31.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.31.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.31.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.31.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.32.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.32.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.32.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.32.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.32.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.32.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.32.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.32.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.33.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.33.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.33.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.33.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.33.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.33.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.33.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.33.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.34.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.34.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.34.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.34.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.34.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.34.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.34.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.34.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.35.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.35.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.35.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.35.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.35.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.35.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.35.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.35.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.36.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.36.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.36.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.36.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.36.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.36.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.36.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.36.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.37.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.37.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.37.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.37.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.37.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.37.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.37.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.37.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.38.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.38.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.38.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.38.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.38.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.38.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.38.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.38.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.39.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.39.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.39.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.39.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.39.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.39.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.39.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.39.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.40.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.40.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.40.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.40.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.40.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.40.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.40.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.40.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.41.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.41.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.41.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.41.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.41.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.41.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.41.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.41.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.42.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.42.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.42.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.42.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.42.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.42.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.42.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.42.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.43.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.43.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.43.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.43.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.43.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.43.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.43.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.43.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.44.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.44.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.44.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.44.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.44.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.44.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.44.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.44.attn_v.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: output.weight - applying manual override: q4_K -> q6_K
llama_tensor_get_type: blk.45.ffn_down_exps.weight - applying manual override: q4_K -> q4_0
llama_tensor_get_type: blk.45.ffn_down_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.45.ffn_gate_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.45.ffn_up_shexp.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.45.attn_k.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.45.attn_output.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.45.attn_q.weight - applying manual override: q4_K -> q8_0
llama_tensor_get_type: blk.45.attn_v.weight - applying manual override: q4_K -> q8_0
[ 1/ 780] token_embd.weight - [ 4096, 151552, 1, 1], type = bf16, converting to q8_0 .. size = 1184.00 MiB -> 629.00 MiB
[ 2/ 780] blk.0.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 3/ 780] blk.0.ffn_down.weight - [ 10944, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 85.50 MiB -> 45.42 MiB
[ 4/ 780] blk.0.ffn_gate.weight - [ 4096, 10944, 1, 1], type = bf16, converting to q8_0 .. size = 85.50 MiB -> 45.42 MiB
[ 5/ 780] blk.0.ffn_up.weight - [ 4096, 10944, 1, 1], type = bf16, converting to q8_0 .. size = 85.50 MiB -> 45.42 MiB
[ 6/ 780] blk.0.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 7/ 780] blk.0.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 8/ 780] blk.0.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 9/ 780] blk.0.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 10/ 780] blk.0.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 11/ 780] blk.0.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 12/ 780] blk.0.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 13/ 780] blk.0.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 14/ 780] blk.1.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 15/ 780] blk.1.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 16/ 780] blk.1.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 17/ 780] blk.1.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 18/ 780] blk.1.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 19/ 780] blk.1.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 20/ 780] blk.1.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 21/ 780] blk.1.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 22/ 780] blk.1.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 23/ 780] blk.1.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 24/ 780] blk.1.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 25/ 780] blk.1.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 26/ 780] blk.1.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 27/ 780] blk.1.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 28/ 780] blk.1.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 29/ 780] blk.1.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 30/ 780] blk.1.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 31/ 780] blk.2.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 32/ 780] blk.2.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 33/ 780] blk.2.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 34/ 780] blk.2.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 35/ 780] blk.2.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 36/ 780] blk.2.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 37/ 780] blk.2.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 38/ 780] blk.2.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 39/ 780] blk.2.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 40/ 780] blk.2.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 41/ 780] blk.2.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 42/ 780] blk.2.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 43/ 780] blk.2.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 44/ 780] blk.2.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 45/ 780] blk.2.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 46/ 780] blk.2.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 47/ 780] blk.2.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 48/ 780] blk.3.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 49/ 780] blk.3.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 50/ 780] blk.3.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 51/ 780] blk.3.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 52/ 780] blk.3.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 53/ 780] blk.3.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 54/ 780] blk.3.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 55/ 780] blk.3.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 56/ 780] blk.3.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 57/ 780] blk.3.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 58/ 780] blk.3.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 59/ 780] blk.3.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 60/ 780] blk.3.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 61/ 780] blk.3.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 62/ 780] blk.3.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 63/ 780] blk.3.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 64/ 780] blk.3.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 65/ 780] blk.4.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 66/ 780] blk.4.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 67/ 780] blk.4.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 68/ 780] blk.4.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 69/ 780] blk.4.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 70/ 780] blk.4.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 71/ 780] blk.4.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 72/ 780] blk.4.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 73/ 780] blk.4.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 74/ 780] blk.4.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 75/ 780] blk.4.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 76/ 780] blk.4.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 77/ 780] blk.4.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 78/ 780] blk.4.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 79/ 780] blk.4.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 80/ 780] blk.4.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 81/ 780] blk.4.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 82/ 780] blk.5.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 83/ 780] blk.5.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 84/ 780] blk.5.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 85/ 780] blk.5.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 86/ 780] blk.5.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 87/ 780] blk.5.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 88/ 780] blk.5.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 89/ 780] blk.5.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 90/ 780] blk.5.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 91/ 780] blk.5.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 92/ 780] blk.5.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 93/ 780] blk.5.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 94/ 780] blk.5.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 95/ 780] blk.5.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 96/ 780] blk.5.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 97/ 780] blk.5.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 98/ 780] blk.5.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 99/ 780] blk.6.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 100/ 780] blk.6.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 101/ 780] blk.6.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 102/ 780] blk.6.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 103/ 780] blk.6.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 104/ 780] blk.6.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 105/ 780] blk.6.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 106/ 780] blk.6.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 107/ 780] blk.6.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 108/ 780] blk.6.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 109/ 780] blk.6.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 110/ 780] blk.6.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 111/ 780] blk.6.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 112/ 780] blk.6.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 113/ 780] blk.6.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 114/ 780] blk.6.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 115/ 780] blk.6.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 116/ 780] blk.7.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 117/ 780] blk.7.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 118/ 780] blk.7.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 119/ 780] blk.7.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 120/ 780] blk.7.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 121/ 780] blk.7.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 122/ 780] blk.7.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 123/ 780] blk.7.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 124/ 780] blk.7.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 125/ 780] blk.7.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 126/ 780] blk.7.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 127/ 780] blk.7.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 128/ 780] blk.7.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 129/ 780] blk.7.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 130/ 780] blk.7.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 131/ 780] blk.7.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 132/ 780] blk.7.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 133/ 780] blk.8.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 134/ 780] blk.8.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 135/ 780] blk.8.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 136/ 780] blk.8.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 137/ 780] blk.8.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 138/ 780] blk.8.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 139/ 780] blk.8.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 140/ 780] blk.8.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 141/ 780] blk.8.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 142/ 780] blk.8.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 143/ 780] blk.8.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 144/ 780] blk.8.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 145/ 780] blk.8.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 146/ 780] blk.8.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 147/ 780] blk.8.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 148/ 780] blk.8.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 149/ 780] blk.8.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 150/ 780] blk.9.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 151/ 780] blk.9.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 152/ 780] blk.9.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 153/ 780] blk.9.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 154/ 780] blk.9.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 155/ 780] blk.9.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 156/ 780] blk.9.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 157/ 780] blk.9.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 158/ 780] blk.9.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 159/ 780] blk.9.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 160/ 780] blk.9.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 161/ 780] blk.9.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 162/ 780] blk.9.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 163/ 780] blk.9.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 164/ 780] blk.9.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 165/ 780] blk.9.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 166/ 780] blk.9.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 167/ 780] blk.10.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 168/ 780] blk.10.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 169/ 780] blk.10.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 170/ 780] blk.10.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 171/ 780] blk.10.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 172/ 780] blk.10.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 173/ 780] blk.10.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 174/ 780] blk.10.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 175/ 780] blk.10.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 176/ 780] blk.10.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 177/ 780] blk.10.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 178/ 780] blk.10.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 179/ 780] blk.10.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 180/ 780] blk.10.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 181/ 780] blk.10.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 182/ 780] blk.10.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 183/ 780] blk.10.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 184/ 780] blk.11.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 185/ 780] blk.11.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 186/ 780] blk.11.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 187/ 780] blk.11.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 188/ 780] blk.11.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 189/ 780] blk.11.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 190/ 780] blk.11.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 191/ 780] blk.11.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 192/ 780] blk.11.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 193/ 780] blk.11.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 194/ 780] blk.11.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 195/ 780] blk.11.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 196/ 780] blk.11.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 197/ 780] blk.11.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 198/ 780] blk.11.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 199/ 780] blk.11.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 200/ 780] blk.11.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 201/ 780] blk.12.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 202/ 780] blk.12.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 203/ 780] blk.12.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 204/ 780] blk.12.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 205/ 780] blk.12.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 206/ 780] blk.12.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 207/ 780] blk.12.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 208/ 780] blk.12.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 209/ 780] blk.12.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 210/ 780] blk.12.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 211/ 780] blk.12.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 212/ 780] blk.12.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 213/ 780] blk.12.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 214/ 780] blk.12.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 215/ 780] blk.12.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 216/ 780] blk.12.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 217/ 780] blk.12.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 218/ 780] blk.13.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 219/ 780] blk.13.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 220/ 780] blk.13.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 221/ 780] blk.13.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 222/ 780] blk.13.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 223/ 780] blk.13.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 224/ 780] blk.13.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 225/ 780] blk.13.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 226/ 780] blk.13.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 227/ 780] blk.13.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 228/ 780] blk.13.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 229/ 780] blk.13.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 230/ 780] blk.13.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 231/ 780] blk.13.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 232/ 780] blk.13.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 233/ 780] blk.13.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 234/ 780] blk.13.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 235/ 780] blk.14.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 236/ 780] blk.14.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 237/ 780] blk.14.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 238/ 780] blk.14.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 239/ 780] blk.14.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 240/ 780] blk.14.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 241/ 780] blk.14.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 242/ 780] blk.14.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 243/ 780] blk.14.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 244/ 780] blk.14.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 245/ 780] blk.14.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 246/ 780] blk.14.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 247/ 780] blk.14.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 248/ 780] blk.14.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 249/ 780] blk.14.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 250/ 780] blk.14.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 251/ 780] blk.14.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 252/ 780] blk.15.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 253/ 780] blk.15.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 254/ 780] blk.15.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 255/ 780] blk.15.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 256/ 780] blk.15.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 257/ 780] blk.15.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 258/ 780] blk.15.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 259/ 780] blk.15.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 260/ 780] blk.15.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 261/ 780] blk.15.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 262/ 780] blk.15.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 263/ 780] blk.15.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 264/ 780] blk.15.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 265/ 780] blk.15.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 266/ 780] blk.15.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 267/ 780] blk.15.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 268/ 780] blk.15.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 269/ 780] blk.16.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 270/ 780] blk.16.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 271/ 780] blk.16.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 272/ 780] blk.16.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 273/ 780] blk.16.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 274/ 780] blk.16.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 275/ 780] blk.16.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 276/ 780] blk.16.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 277/ 780] blk.16.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 278/ 780] blk.16.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 279/ 780] blk.16.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 280/ 780] blk.16.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 281/ 780] blk.16.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 282/ 780] blk.16.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 283/ 780] blk.16.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 284/ 780] blk.16.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 285/ 780] blk.16.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 286/ 780] blk.17.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 287/ 780] blk.17.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 288/ 780] blk.17.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 289/ 780] blk.17.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 290/ 780] blk.17.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 291/ 780] blk.17.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 292/ 780] blk.17.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 293/ 780] blk.17.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 294/ 780] blk.17.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 295/ 780] blk.17.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 296/ 780] blk.17.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 297/ 780] blk.17.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 298/ 780] blk.17.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 299/ 780] blk.17.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 300/ 780] blk.17.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 301/ 780] blk.17.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 302/ 780] blk.17.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 303/ 780] blk.18.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 304/ 780] blk.18.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 305/ 780] blk.18.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 306/ 780] blk.18.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 307/ 780] blk.18.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 308/ 780] blk.18.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 309/ 780] blk.18.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 310/ 780] blk.18.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 311/ 780] blk.18.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 312/ 780] blk.18.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 313/ 780] blk.18.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 314/ 780] blk.18.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 315/ 780] blk.18.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 316/ 780] blk.18.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 317/ 780] blk.18.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 318/ 780] blk.18.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 319/ 780] blk.18.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 320/ 780] blk.19.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 321/ 780] blk.19.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 322/ 780] blk.19.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 323/ 780] blk.19.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 324/ 780] blk.19.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 325/ 780] blk.19.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 326/ 780] blk.19.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 327/ 780] blk.19.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 328/ 780] blk.19.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 329/ 780] blk.19.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 330/ 780] blk.19.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 331/ 780] blk.19.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 332/ 780] blk.19.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 333/ 780] blk.19.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 334/ 780] blk.19.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 335/ 780] blk.19.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 336/ 780] blk.19.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 337/ 780] blk.20.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 338/ 780] blk.20.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 339/ 780] blk.20.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 340/ 780] blk.20.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 341/ 780] blk.20.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 342/ 780] blk.20.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 343/ 780] blk.20.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 344/ 780] blk.20.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 345/ 780] blk.20.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 346/ 780] blk.20.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 347/ 780] blk.20.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 348/ 780] blk.20.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 349/ 780] blk.20.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 350/ 780] blk.20.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 351/ 780] blk.20.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 352/ 780] blk.20.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 353/ 780] blk.20.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 354/ 780] blk.21.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 355/ 780] blk.21.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 356/ 780] blk.21.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 357/ 780] blk.21.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 358/ 780] blk.21.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 359/ 780] blk.21.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 360/ 780] blk.21.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 361/ 780] blk.21.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 362/ 780] blk.21.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 363/ 780] blk.21.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 364/ 780] blk.21.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 365/ 780] blk.21.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 366/ 780] blk.21.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 367/ 780] blk.21.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 368/ 780] blk.21.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 369/ 780] blk.21.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 370/ 780] blk.21.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 371/ 780] blk.22.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 372/ 780] blk.22.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 373/ 780] blk.22.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 374/ 780] blk.22.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 375/ 780] blk.22.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 376/ 780] blk.22.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 377/ 780] blk.22.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 378/ 780] blk.22.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 379/ 780] blk.22.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 380/ 780] blk.22.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 381/ 780] blk.22.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 382/ 780] blk.22.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 383/ 780] blk.22.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 384/ 780] blk.22.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 385/ 780] blk.22.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 386/ 780] blk.22.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 387/ 780] blk.22.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 388/ 780] blk.23.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 389/ 780] blk.23.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 390/ 780] blk.23.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 391/ 780] blk.23.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 392/ 780] blk.23.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 393/ 780] blk.23.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 394/ 780] blk.23.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 395/ 780] blk.23.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 396/ 780] blk.23.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 397/ 780] blk.23.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 398/ 780] blk.23.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 399/ 780] blk.23.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 400/ 780] blk.23.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 401/ 780] blk.23.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 402/ 780] blk.23.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 403/ 780] blk.23.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 404/ 780] blk.23.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 405/ 780] blk.24.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 406/ 780] blk.24.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 407/ 780] blk.24.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 408/ 780] blk.24.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 409/ 780] blk.24.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 410/ 780] blk.24.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 411/ 780] blk.24.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 412/ 780] blk.24.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 413/ 780] blk.24.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 414/ 780] blk.24.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 415/ 780] blk.24.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 416/ 780] blk.24.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 417/ 780] blk.24.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 418/ 780] blk.24.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 419/ 780] blk.24.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 420/ 780] blk.24.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 421/ 780] blk.24.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 422/ 780] blk.25.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 423/ 780] blk.25.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 424/ 780] blk.25.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 425/ 780] blk.25.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 426/ 780] blk.25.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 427/ 780] blk.25.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 428/ 780] blk.25.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 429/ 780] blk.25.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 430/ 780] blk.25.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 431/ 780] blk.25.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 432/ 780] blk.25.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 433/ 780] blk.25.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 434/ 780] blk.25.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 435/ 780] blk.25.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 436/ 780] blk.25.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 437/ 780] blk.25.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 438/ 780] blk.25.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 439/ 780] blk.26.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 440/ 780] blk.26.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 441/ 780] blk.26.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 442/ 780] blk.26.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 443/ 780] blk.26.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 444/ 780] blk.26.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 445/ 780] blk.26.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 446/ 780] blk.26.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 447/ 780] blk.26.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 448/ 780] blk.26.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 449/ 780] blk.26.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 450/ 780] blk.26.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 451/ 780] blk.26.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 452/ 780] blk.26.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 453/ 780] blk.26.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 454/ 780] blk.26.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 455/ 780] blk.26.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 456/ 780] blk.27.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 457/ 780] blk.27.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 458/ 780] blk.27.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 459/ 780] blk.27.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 460/ 780] blk.27.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 461/ 780] blk.27.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 462/ 780] blk.27.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 463/ 780] blk.27.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 464/ 780] blk.27.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 465/ 780] blk.27.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 466/ 780] blk.27.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 467/ 780] blk.27.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 468/ 780] blk.27.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 469/ 780] blk.27.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 470/ 780] blk.27.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 471/ 780] blk.27.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 472/ 780] blk.27.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 473/ 780] blk.28.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 474/ 780] blk.28.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 475/ 780] blk.28.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 476/ 780] blk.28.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 477/ 780] blk.28.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 478/ 780] blk.28.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 479/ 780] blk.28.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 480/ 780] blk.28.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 481/ 780] blk.28.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 482/ 780] blk.28.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 483/ 780] blk.28.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 484/ 780] blk.28.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 485/ 780] blk.28.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 486/ 780] blk.28.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 487/ 780] blk.28.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 488/ 780] blk.28.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 489/ 780] blk.28.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 490/ 780] blk.29.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 491/ 780] blk.29.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 492/ 780] blk.29.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 493/ 780] blk.29.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 494/ 780] blk.29.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 495/ 780] blk.29.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 496/ 780] blk.29.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 497/ 780] blk.29.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 498/ 780] blk.29.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 499/ 780] blk.29.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 500/ 780] blk.29.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 501/ 780] blk.29.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 502/ 780] blk.29.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 503/ 780] blk.29.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 504/ 780] blk.29.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 505/ 780] blk.29.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 506/ 780] blk.29.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 507/ 780] blk.30.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 508/ 780] blk.30.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 509/ 780] blk.30.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 510/ 780] blk.30.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 511/ 780] blk.30.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 512/ 780] blk.30.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 513/ 780] blk.30.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 514/ 780] blk.30.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 515/ 780] blk.30.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 516/ 780] blk.30.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 517/ 780] blk.30.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 518/ 780] blk.30.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 519/ 780] blk.30.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 520/ 780] blk.30.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 521/ 780] blk.30.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 522/ 780] blk.30.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 523/ 780] blk.30.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 524/ 780] blk.31.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 525/ 780] blk.31.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 526/ 780] blk.31.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 527/ 780] blk.31.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 528/ 780] blk.31.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 529/ 780] blk.31.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 530/ 780] blk.31.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 531/ 780] blk.31.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 532/ 780] blk.31.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 533/ 780] blk.31.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 534/ 780] blk.31.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 535/ 780] blk.31.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 536/ 780] blk.31.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 537/ 780] blk.31.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 538/ 780] blk.31.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 539/ 780] blk.31.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 540/ 780] blk.31.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 541/ 780] blk.32.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 542/ 780] blk.32.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 543/ 780] blk.32.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 544/ 780] blk.32.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 545/ 780] blk.32.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 546/ 780] blk.32.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 547/ 780] blk.32.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 548/ 780] blk.32.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 549/ 780] blk.32.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 550/ 780] blk.32.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 551/ 780] blk.32.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 552/ 780] blk.32.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 553/ 780] blk.32.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 554/ 780] blk.32.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 555/ 780] blk.32.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 556/ 780] blk.32.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 557/ 780] blk.32.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 558/ 780] blk.33.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 559/ 780] blk.33.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 560/ 780] blk.33.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 561/ 780] blk.33.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 562/ 780] blk.33.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 563/ 780] blk.33.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 564/ 780] blk.33.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 565/ 780] blk.33.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 566/ 780] blk.33.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 567/ 780] blk.33.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 568/ 780] blk.33.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 569/ 780] blk.33.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 570/ 780] blk.33.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 571/ 780] blk.33.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 572/ 780] blk.33.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 573/ 780] blk.33.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 574/ 780] blk.33.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 575/ 780] blk.34.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 576/ 780] blk.34.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 577/ 780] blk.34.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 578/ 780] blk.34.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 579/ 780] blk.34.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 580/ 780] blk.34.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 581/ 780] blk.34.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 582/ 780] blk.34.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 583/ 780] blk.34.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 584/ 780] blk.34.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 585/ 780] blk.34.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 586/ 780] blk.34.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 587/ 780] blk.34.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 588/ 780] blk.34.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 589/ 780] blk.34.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 590/ 780] blk.34.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 591/ 780] blk.34.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 592/ 780] blk.35.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 593/ 780] blk.35.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 594/ 780] blk.35.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 595/ 780] blk.35.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 596/ 780] blk.35.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 597/ 780] blk.35.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 598/ 780] blk.35.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 599/ 780] blk.35.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 600/ 780] blk.35.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 601/ 780] blk.35.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 602/ 780] blk.35.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 603/ 780] blk.35.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 604/ 780] blk.35.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 605/ 780] blk.35.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 606/ 780] blk.35.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 607/ 780] blk.35.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 608/ 780] blk.35.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 609/ 780] blk.36.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 610/ 780] blk.36.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 611/ 780] blk.36.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 612/ 780] blk.36.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 613/ 780] blk.36.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 614/ 780] blk.36.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 615/ 780] blk.36.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 616/ 780] blk.36.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 617/ 780] blk.36.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 618/ 780] blk.36.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 619/ 780] blk.36.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 620/ 780] blk.36.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 621/ 780] blk.36.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 622/ 780] blk.36.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 623/ 780] blk.36.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 624/ 780] blk.36.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 625/ 780] blk.36.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 626/ 780] blk.37.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 627/ 780] blk.37.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 628/ 780] blk.37.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 629/ 780] blk.37.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 630/ 780] blk.37.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 631/ 780] blk.37.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 632/ 780] blk.37.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 633/ 780] blk.37.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 634/ 780] blk.37.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 635/ 780] blk.37.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 636/ 780] blk.37.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 637/ 780] blk.37.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 638/ 780] blk.37.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 639/ 780] blk.37.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 640/ 780] blk.37.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 641/ 780] blk.37.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 642/ 780] blk.37.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 643/ 780] blk.38.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 644/ 780] blk.38.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 645/ 780] blk.38.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 646/ 780] blk.38.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 647/ 780] blk.38.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 648/ 780] blk.38.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 649/ 780] blk.38.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 650/ 780] blk.38.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 651/ 780] blk.38.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 652/ 780] blk.38.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 653/ 780] blk.38.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 654/ 780] blk.38.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 655/ 780] blk.38.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 656/ 780] blk.38.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 657/ 780] blk.38.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 658/ 780] blk.38.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 659/ 780] blk.38.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 660/ 780] blk.39.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 661/ 780] blk.39.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 662/ 780] blk.39.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 663/ 780] blk.39.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 664/ 780] blk.39.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 665/ 780] blk.39.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 666/ 780] blk.39.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 667/ 780] blk.39.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 668/ 780] blk.39.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 669/ 780] blk.39.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 670/ 780] blk.39.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 671/ 780] blk.39.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 672/ 780] blk.39.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 673/ 780] blk.39.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 674/ 780] blk.39.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 675/ 780] blk.39.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 676/ 780] blk.39.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 677/ 780] blk.40.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 678/ 780] blk.40.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 679/ 780] blk.40.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 680/ 780] blk.40.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 681/ 780] blk.40.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 682/ 780] blk.40.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 683/ 780] blk.40.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 684/ 780] blk.40.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 685/ 780] blk.40.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 686/ 780] blk.40.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 687/ 780] blk.40.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 688/ 780] blk.40.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 689/ 780] blk.40.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 690/ 780] blk.40.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 691/ 780] blk.40.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 692/ 780] blk.40.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 693/ 780] blk.40.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 694/ 780] blk.41.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 695/ 780] blk.41.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 696/ 780] blk.41.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 697/ 780] blk.41.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 698/ 780] blk.41.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 699/ 780] blk.41.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 700/ 780] blk.41.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 701/ 780] blk.41.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 702/ 780] blk.41.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 703/ 780] blk.41.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 704/ 780] blk.41.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 705/ 780] blk.41.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 706/ 780] blk.41.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 707/ 780] blk.41.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 708/ 780] blk.41.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 709/ 780] blk.41.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 710/ 780] blk.41.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 711/ 780] blk.42.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 712/ 780] blk.42.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 713/ 780] blk.42.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 714/ 780] blk.42.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 715/ 780] blk.42.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 716/ 780] blk.42.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 717/ 780] blk.42.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 718/ 780] blk.42.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 719/ 780] blk.42.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 720/ 780] blk.42.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 721/ 780] blk.42.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 722/ 780] blk.42.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 723/ 780] blk.42.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 724/ 780] blk.42.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 725/ 780] blk.42.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 726/ 780] blk.42.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 727/ 780] blk.42.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 728/ 780] blk.43.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 729/ 780] blk.43.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 730/ 780] blk.43.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 731/ 780] blk.43.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 732/ 780] blk.43.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 733/ 780] blk.43.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 734/ 780] blk.43.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 735/ 780] blk.43.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 736/ 780] blk.43.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 737/ 780] blk.43.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 738/ 780] blk.43.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 739/ 780] blk.43.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 740/ 780] blk.43.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 741/ 780] blk.43.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 742/ 780] blk.43.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 743/ 780] blk.43.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 744/ 780] blk.43.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 745/ 780] blk.44.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 746/ 780] blk.44.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 747/ 780] blk.44.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 748/ 780] blk.44.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 749/ 780] blk.44.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 750/ 780] blk.44.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 751/ 780] blk.44.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 752/ 780] blk.44.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 753/ 780] blk.44.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 754/ 780] blk.44.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 755/ 780] blk.44.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 756/ 780] blk.44.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 757/ 780] blk.44.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 758/ 780] blk.44.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 759/ 780] blk.44.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 760/ 780] blk.44.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 761/ 780] blk.44.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 762/ 780] output.weight - [ 4096, 151552, 1, 1], type = bf16, converting to q6_K .. size = 1184.00 MiB -> 485.62 MiB
[ 763/ 780] blk.45.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 764/ 780] blk.45.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 765/ 780] blk.45.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 766/ 780] blk.45.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_K .. size = 1408.00 MiB -> 396.00 MiB
[ 767/ 780] blk.45.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 768/ 780] blk.45.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 769/ 780] blk.45.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 770/ 780] blk.45.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 771/ 780] blk.45.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 772/ 780] blk.45.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 773/ 780] blk.45.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 774/ 780] blk.45.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 775/ 780] blk.45.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 776/ 780] blk.45.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 777/ 780] blk.45.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 778/ 780] blk.45.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 779/ 780] blk.45.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 780/ 780] output_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
llama_model_quantize_impl: model size = 203851.49 MiB (16.00 BPW)
llama_model_quantize_impl: quant size = 60676.79 MiB (4.76 BPW)
llama_quantize: quantize time = 1076918.57 ms
llama_quantize: total time = 1076918.57 ms
+ ./llama.cpp/build/bin/llama-quantize ./upload-GLM_4.5_Air/mtp-GLM-4.5-Air-BF16.gguf ./upload-GLM_4.5_Air/mtp-GLM-4.5-Air-Q8_0.gguf Q8_0
version: 0.3.0-dev (build 1, commit c1d0e7a)
built with GNU 14.2.0 for Linux x86_64
llama_quantize: quantizing './upload-GLM_4.5_Air/mtp-GLM-4.5-Air-BF16.gguf' to './upload-GLM_4.5_Air/mtp-GLM-4.5-Air-Q8_0.gguf' as Q8_0
llama_model_loader: loaded meta data with 43 key-value pairs and 26 tensors from ./upload-GLM_4.5_Air/mtp-GLM-4.5-Air-BF16.gguf (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = glm4moe
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = GLM-4.5-Air
llama_model_loader: - kv 3: general.size_label str = 128x2.7B
llama_model_loader: - kv 4: general.license str = mit
llama_model_loader: - kv 5: general.tags arr[str,1] = ["text-generation"]
llama_model_loader: - kv 6: general.languages arr[str,2] = ["en", "zh"]
llama_model_loader: - kv 7: glm4moe.block_count u32 = 47
llama_model_loader: - kv 8: glm4moe.context_length u32 = 131072
llama_model_loader: - kv 9: glm4moe.embedding_length u32 = 4096
llama_model_loader: - kv 10: glm4moe.feed_forward_length u32 = 10944
llama_model_loader: - kv 11: glm4moe.attention.head_count u32 = 96
llama_model_loader: - kv 12: glm4moe.attention.head_count_kv u32 = 8
llama_model_loader: - kv 13: glm4moe.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 14: glm4moe.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 15: glm4moe.expert_count u32 = 128
llama_model_loader: - kv 16: glm4moe.expert_used_count u32 = 8
llama_model_loader: - kv 17: glm4moe.expert_group_count u32 = 1
llama_model_loader: - kv 18: glm4moe.expert_group_used_count u32 = 1
llama_model_loader: - kv 19: glm4moe.attention.key_length u32 = 128
llama_model_loader: - kv 20: glm4moe.attention.value_length u32 = 128
llama_model_loader: - kv 21: general.file_type u32 = 32
llama_model_loader: - kv 22: glm4moe.rope.dimension_count u32 = 64
llama_model_loader: - kv 23: glm4moe.expert_feed_forward_length u32 = 1408
llama_model_loader: - kv 24: glm4moe.expert_shared_count u32 = 1
llama_model_loader: - kv 25: glm4moe.leading_dense_block_count u32 = 1
llama_model_loader: - kv 26: glm4moe.expert_gating_func u32 = 2
llama_model_loader: - kv 27: glm4moe.expert_weights_scale f32 = 1.000000
llama_model_loader: - kv 28: glm4moe.expert_weights_norm bool = true
llama_model_loader: - kv 29: glm4moe.nextn_predict_layers u32 = 1
llama_model_loader: - kv 30: general.quantization_version u32 = 2
llama_model_loader: - kv 31: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 32: tokenizer.ggml.pre str = glm4
llama_model_loader: - kv 33: tokenizer.ggml.tokens arr[str,151552] = ["!", "\"", "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 34: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 35: tokenizer.ggml.merges arr[str,318088] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "...
llama_model_loader: - kv 36: tokenizer.ggml.eos_token_id u32 = 151329
llama_model_loader: - kv 37: tokenizer.ggml.padding_token_id u32 = 151329
llama_model_loader: - kv 38: tokenizer.ggml.bos_token_id u32 = 151331
llama_model_loader: - kv 39: tokenizer.ggml.eot_token_id u32 = 151336
llama_model_loader: - kv 40: tokenizer.ggml.unknown_token_id u32 = 151329
llama_model_loader: - kv 41: tokenizer.ggml.eom_token_id u32 = 151338
llama_model_loader: - kv 42: tokenizer.chat_template str = [gMASK]<sop>\n{%- if tools -%}\n<|syste...
llama_model_loader: - type f32: 11 tensors
llama_model_loader: - type bf16: 15 tensors
[ 1/ 26] output.weight - [ 4096, 151552, 1, 1], type = bf16, converting to q8_0 .. size = 1184.00 MiB -> 629.00 MiB
[ 2/ 26] output_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 3/ 26] token_embd.weight - [ 4096, 151552, 1, 1], type = bf16, converting to q8_0 .. size = 1184.00 MiB -> 629.00 MiB
[ 4/ 26] blk.46.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 5/ 26] blk.46.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 6/ 26] blk.46.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 7/ 26] blk.46.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 8/ 26] blk.46.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 9/ 26] blk.46.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q8_0 .. size = 96.00 MiB -> 51.00 MiB
[ 10/ 26] blk.46.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 11/ 26] blk.46.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q8_0 .. size = 8.00 MiB -> 4.25 MiB
[ 12/ 26] blk.46.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 13/ 26] blk.46.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 14/ 26] blk.46.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 15/ 26] blk.46.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 16/ 26] blk.46.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 17/ 26] blk.46.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 18/ 26] blk.46.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q8_0 .. size = 1408.00 MiB -> 748.00 MiB
[ 19/ 26] blk.46.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q8_0 .. size = 11.00 MiB -> 5.84 MiB
[ 20/ 26] blk.46.nextn.eh_proj.weight - [ 8192, 4096, 1, 1], type = bf16, converting to q8_0 .. size = 64.00 MiB -> 34.00 MiB
[ 21/ 26] blk.46.nextn.embed_tokens.weight - [ 4096, 151552, 1, 1], type = bf16, converting to q8_0 .. size = 1184.00 MiB -> 629.00 MiB
[ 22/ 26] blk.46.nextn.enorm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 23/ 26] blk.46.nextn.hnorm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 24/ 26] blk.46.nextn.shared_head_head.weight - [ 4096, 151552, 1, 1], type = bf16, converting to q8_0 .. size = 1184.00 MiB -> 629.00 MiB
[ 25/ 26] blk.46.nextn.shared_head_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 26/ 26] blk.46.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
llama_model_quantize_impl: model size = 9267.15 MiB (16.00 BPW)
llama_model_quantize_impl: quant size = 4924.18 MiB (8.50 BPW)
llama_quantize: quantize time = 41421.57 ms
llama_quantize: total time = 41421.57 ms
+ ./llama.cpp/build/bin/llama-quantize --pure ./upload-GLM_4.5_Air/mtp-GLM-4.5-Air-BF16.gguf ./upload-GLM_4.5_Air/mtp-GLM-4.5-Air-Q4_0.gguf Q4_0
version: 0.3.0-dev (build 1, commit c1d0e7a)
built with GNU 14.2.0 for Linux x86_64
llama_quantize: quantizing './upload-GLM_4.5_Air/mtp-GLM-4.5-Air-BF16.gguf' to './upload-GLM_4.5_Air/mtp-GLM-4.5-Air-Q4_0.gguf' as Q4_0
llama_model_loader: loaded meta data with 43 key-value pairs and 26 tensors from ./upload-GLM_4.5_Air/mtp-GLM-4.5-Air-BF16.gguf (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = glm4moe
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = GLM-4.5-Air
llama_model_loader: - kv 3: general.size_label str = 128x2.7B
llama_model_loader: - kv 4: general.license str = mit
llama_model_loader: - kv 5: general.tags arr[str,1] = ["text-generation"]
llama_model_loader: - kv 6: general.languages arr[str,2] = ["en", "zh"]
llama_model_loader: - kv 7: glm4moe.block_count u32 = 47
llama_model_loader: - kv 8: glm4moe.context_length u32 = 131072
llama_model_loader: - kv 9: glm4moe.embedding_length u32 = 4096
llama_model_loader: - kv 10: glm4moe.feed_forward_length u32 = 10944
llama_model_loader: - kv 11: glm4moe.attention.head_count u32 = 96
llama_model_loader: - kv 12: glm4moe.attention.head_count_kv u32 = 8
llama_model_loader: - kv 13: glm4moe.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 14: glm4moe.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 15: glm4moe.expert_count u32 = 128
llama_model_loader: - kv 16: glm4moe.expert_used_count u32 = 8
llama_model_loader: - kv 17: glm4moe.expert_group_count u32 = 1
llama_model_loader: - kv 18: glm4moe.expert_group_used_count u32 = 1
llama_model_loader: - kv 19: glm4moe.attention.key_length u32 = 128
llama_model_loader: - kv 20: glm4moe.attention.value_length u32 = 128
llama_model_loader: - kv 21: general.file_type u32 = 32
llama_model_loader: - kv 22: glm4moe.rope.dimension_count u32 = 64
llama_model_loader: - kv 23: glm4moe.expert_feed_forward_length u32 = 1408
llama_model_loader: - kv 24: glm4moe.expert_shared_count u32 = 1
llama_model_loader: - kv 25: glm4moe.leading_dense_block_count u32 = 1
llama_model_loader: - kv 26: glm4moe.expert_gating_func u32 = 2
llama_model_loader: - kv 27: glm4moe.expert_weights_scale f32 = 1.000000
llama_model_loader: - kv 28: glm4moe.expert_weights_norm bool = true
llama_model_loader: - kv 29: glm4moe.nextn_predict_layers u32 = 1
llama_model_loader: - kv 30: general.quantization_version u32 = 2
llama_model_loader: - kv 31: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 32: tokenizer.ggml.pre str = glm4
llama_model_loader: - kv 33: tokenizer.ggml.tokens arr[str,151552] = ["!", "\"", "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 34: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 35: tokenizer.ggml.merges arr[str,318088] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "...
llama_model_loader: - kv 36: tokenizer.ggml.eos_token_id u32 = 151329
llama_model_loader: - kv 37: tokenizer.ggml.padding_token_id u32 = 151329
llama_model_loader: - kv 38: tokenizer.ggml.bos_token_id u32 = 151331
llama_model_loader: - kv 39: tokenizer.ggml.eot_token_id u32 = 151336
llama_model_loader: - kv 40: tokenizer.ggml.unknown_token_id u32 = 151329
llama_model_loader: - kv 41: tokenizer.ggml.eom_token_id u32 = 151338
llama_model_loader: - kv 42: tokenizer.chat_template str = [gMASK]<sop>\n{%- if tools -%}\n<|syste...
llama_model_loader: - type f32: 11 tensors
llama_model_loader: - type bf16: 15 tensors
[ 1/ 26] output.weight - [ 4096, 151552, 1, 1], type = bf16, converting to q4_0 .. size = 1184.00 MiB -> 333.00 MiB
[ 2/ 26] output_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 3/ 26] token_embd.weight - [ 4096, 151552, 1, 1], type = bf16, converting to q4_0 .. size = 1184.00 MiB -> 333.00 MiB
[ 4/ 26] blk.46.attn_k.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 5/ 26] blk.46.attn_k.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q4_0 .. size = 8.00 MiB -> 2.25 MiB
[ 6/ 26] blk.46.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 7/ 26] blk.46.attn_output.weight - [ 12288, 4096, 1, 1], type = bf16, converting to q4_0 .. size = 96.00 MiB -> 27.00 MiB
[ 8/ 26] blk.46.attn_q.bias - [ 12288, 1, 1, 1], type = f32, size = 0.047 MiB
[ 9/ 26] blk.46.attn_q.weight - [ 4096, 12288, 1, 1], type = bf16, converting to q4_0 .. size = 96.00 MiB -> 27.00 MiB
[ 10/ 26] blk.46.attn_v.bias - [ 1024, 1, 1, 1], type = f32, size = 0.004 MiB
[ 11/ 26] blk.46.attn_v.weight - [ 4096, 1024, 1, 1], type = bf16, converting to q4_0 .. size = 8.00 MiB -> 2.25 MiB
[ 12/ 26] blk.46.exp_probs_b.bias - [ 128, 1, 1, 1], type = f32, size = 0.000 MiB
[ 13/ 26] blk.46.ffn_down_exps.weight - [ 1408, 4096, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 14/ 26] blk.46.ffn_down_shexp.weight - [ 1408, 4096, 1, 1], type = bf16, converting to q4_0 .. size = 11.00 MiB -> 3.09 MiB
[ 15/ 26] blk.46.ffn_gate_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 16/ 26] blk.46.ffn_gate_inp.weight - [ 4096, 128, 1, 1], type = f32, size = 2.000 MiB
[ 17/ 26] blk.46.ffn_gate_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q4_0 .. size = 11.00 MiB -> 3.09 MiB
[ 18/ 26] blk.46.ffn_up_exps.weight - [ 4096, 1408, 128, 1], type = bf16, converting to q4_0 .. size = 1408.00 MiB -> 396.00 MiB
[ 19/ 26] blk.46.ffn_up_shexp.weight - [ 4096, 1408, 1, 1], type = bf16, converting to q4_0 .. size = 11.00 MiB -> 3.09 MiB
[ 20/ 26] blk.46.nextn.eh_proj.weight - [ 8192, 4096, 1, 1], type = bf16, converting to q4_0 .. size = 64.00 MiB -> 18.00 MiB
[ 21/ 26] blk.46.nextn.embed_tokens.weight - [ 4096, 151552, 1, 1], type = bf16, converting to q4_0 .. size = 1184.00 MiB -> 333.00 MiB
[ 22/ 26] blk.46.nextn.enorm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 23/ 26] blk.46.nextn.hnorm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 24/ 26] blk.46.nextn.shared_head_head.weight - [ 4096, 151552, 1, 1], type = bf16, converting to q4_0 .. size = 1184.00 MiB -> 333.00 MiB
[ 25/ 26] blk.46.nextn.shared_head_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
[ 26/ 26] blk.46.post_attention_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MiB
llama_model_quantize_impl: model size = 9267.15 MiB (16.00 BPW)
llama_model_quantize_impl: quant size = 2607.93 MiB (4.50 BPW)
llama_quantize: quantize time = 12801.47 ms
llama_quantize: total time = 12801.47 ms
+ echo GLM-4.5-Air-Q8_0-00001-of-00002.gguf
+ echo GLM-4.5-Air-Q8_0-00002-of-00002.gguf
+ echo GLM-4.5-Air-Q4_K_M-00001-of-00002.gguf
+ echo GLM-4.5-Air-Q4_K_M-00002-of-00002.gguf
+ echo mtp-GLM-4.5-Air-Q8_0.gguf
+ echo mtp-GLM-4.5-Air-Q4_0.gguf