(APIServer pid=2525559) INFO 01-20 15:00:14 [api_server.py:1772] vLLM API server version 0.12.0
(APIServer pid=2525559) INFO 01-20 15:00:14 [utils.py:253] non-default args: {'model_tag': '/mnt/tidal-alsh01/usr/dawo/dawo/model_ori/ZhipuAI/GLM-4.6V', 'host': '0.0.0.0', 'port': 1054, 'api_key': ['token-abc123'], 'model': '/mnt/tidal-alsh01/usr/dawo/dawo/model_ori/ZhipuAI/GLM-4.6V', 'trust_remote_code': True, 'max_model_len': 16384, 'tensor_parallel_size': 4}
(APIServer pid=2525559) INFO 01-20 15:00:14 [model.py:637] Resolved architecture: Glm4vMoeForConditionalGeneration
(APIServer pid=2525559) INFO 01-20 15:00:14 [model.py:1750] Using max model len 16384
(APIServer pid=2525559) INFO 01-20 15:00:14 [scheduler.py:228] Chunked prefill is enabled with max_num_batched_tokens=8192.
(EngineCore_DP0 pid=2525786) INFO 01-20 15:00:23 [core.py:93] Initializing a V1 LLM engine (v0.12.0) with config: model='/mnt/tidal-alsh01/usr/dawo/dawo/model_ori/ZhipuAI/GLM-4.6V', speculative_config=None, tokenizer='/mnt/tidal-alsh01/usr/dawo/dawo/model_ori/ZhipuAI/GLM-4.6V', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=True, dtype=torch.bfloat16, max_seq_len=16384, download_dir=None, load_format=auto, tensor_parallel_size=4, pipeline_parallel_size=1, data_parallel_size=1, disable_custom_all_reduce=False, quantization=None, enforce_eager=False, kv_cache_dtype=auto, device_config=cuda, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_fallback=False, disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser='', reasoning_parser_plugin='', enable_in_reasoning=False), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None, kv_cache_metrics=False, kv_cache_metrics_sample=0.01), seed=0, served_model_name=/mnt/tidal-alsh01/usr/dawo/dawo/model_ori/ZhipuAI/GLM-4.6V, enable_prefix_caching=True, enable_chunked_prefill=True, pooler_config=None, compilation_config={'level': None, 'mode': <CompilationMode.VLLM_COMPILE: 3>, 'debug_dump_path': None, 'cache_dir': '', 'compile_cache_save_format': 'binary', 'backend': 'inductor', 'custom_ops': ['none'], 'splitting_ops': ['vllm::unified_attention', 'vllm::unified_attention_with_output', 'vllm::unified_mla_attention', 'vllm::unified_mla_attention_with_output', 'vllm::mamba_mixer2', 'vllm::mamba_mixer', 'vllm::short_conv', 'vllm::linear_attention', 'vllm::plamo2_mamba_mixer', 'vllm::gdn_attention_core', 'vllm::kda_attention', 'vllm::sparse_attn_indexer'], 'compile_mm_encoder': False, 'compile_sizes': [], 'inductor_compile_config': {'enable_auto_functionalized_v2': False, 'combo_kernels': True, 'benchmark_combo_kernel': True}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.FULL_AND_PIECEWISE: (2, 1)>, 'cudagraph_num_of_warmups': 1, 'cudagraph_capture_sizes': [1, 2, 4, 8, 16, 24, 32, 40, 48, 56, 64, 72, 80, 88, 96, 104, 112, 120, 128, 136, 144, 152, 160, 168, 176, 184, 192, 200, 208, 216, 224, 232, 240, 248, 256, 272, 288, 304, 320, 336, 352, 368, 384, 400, 416, 432, 448, 464, 480, 496, 512], 'cudagraph_copy_inputs': False, 'cudagraph_specialize_lora': True, 'use_inductor_graph_partition': False, 'pass_config': {'fuse_norm_quant': False, 'fuse_act_quant': False, 'fuse_attn_quant': False, 'eliminate_noops': True, 'enable_sp': False, 'fuse_gemm_comms': False, 'fuse_allreduce_rms': False}, 'max_cudagraph_capture_size': 512, 'dynamic_shapes_config': {'type': <DynamicShapesType.BACKED: 'backed'>}, 'local_cache_dir': None}
(EngineCore_DP0 pid=2525786) WARNING 01-20 15:00:23 [multiproc_executor.py:880] Reducing Torch parallelism from 96 threads to 1 to avoid unnecessary CPU contention. Set OMP_NUM_THREADS in the external environment to tune this value as needed.
INFO 01-20 15:00:37 [parallel_state.py:1200] world_size=4 rank=3 local_rank=3 distributed_init_method=tcp://127.0.0.1:38903 backend=nccl
INFO 01-20 15:00:37 [parallel_state.py:1200] world_size=4 rank=1 local_rank=1 distributed_init_method=tcp://127.0.0.1:38903 backend=nccl
INFO 01-20 15:00:37 [parallel_state.py:1200] world_size=4 rank=0 local_rank=0 distributed_init_method=tcp://127.0.0.1:38903 backend=nccl
INFO 01-20 15:00:37 [parallel_state.py:1200] world_size=4 rank=2 local_rank=2 distributed_init_method=tcp://127.0.0.1:38903 backend=nccl
INFO 01-20 15:00:37 [pynccl.py:111] vLLM is using nccl==2.27.5
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO cudaDriverVersion 12080
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Bootstrap: Using veth1:10.148.9.243<0>
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NCCL version 2.27.5+cuda12.9
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/Plugin: Plugin name set by env to libnccl-net.so
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v10 symbol.
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v9 symbol.
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v8 symbol.
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v7 symbol.
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/Plugin: Loaded net plugin NCCL RDMA Plugin v6 (v6)
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v10 symbol.
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v9 symbol.
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v8 symbol.
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v7 symbol.
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v6 symbol.
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Successfully loaded external plugin libnccl-net.so
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Plugin Path : /usr/local/lib/libnccl-net.so
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Plugin version : 1.4@Tencent Cloud
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO P2P plugin IBext
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/IB : No device found.
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/IB : No device found.
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/IB : Using [RO]; OOB veth1:10.148.9.243<0>
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NET/Socket : Using [0]veth1:10.148.9.243<0> [1]eth0:10.148.9.243<0>
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Initialized NET plugin Socket
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Assigned NET plugin Socket to comm
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Using network Socket
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO ncclCommInitRank comm 0x1dc951b0 rank 3 nranks 4 cudaDev 3 nvmlDev 3 busId 67000 commId 0x8d6e43a37d913efb - Init START
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO RAS client listening socket at ::1<28028>
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Bootstrap timings total 0.003037 (create 0.000021, send 0.000073, recv 0.000068, ring 0.000040, delay 0.000000)
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO MNNVL busId 0x67000 fabric UUID 0.0 cliqueId 0x0 state 3 healthMask 0x0
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Setting affinity for GPU 3 to 0-47,96-143
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NVLS multicast support is available on dev 3 (NVLS_NCHANNELS 16)
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO comm 0x1dc951b0 rank 3 nRanks 4 nNodes 1 localRanks 4 localRank 3 MNNVL 0
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Trees [0] -1/-1/-1->3->2 [1] -1/-1/-1->3->2 [2] 0/-1/-1->3->2 [3] 0/-1/-1->3->2 [4] 0/-1/-1->3->2 [5] 0/-1/-1->3->2 [6] 0/-1/-1->3->-1 [7] 0/-1/-1->3->-1 [8] -1/-1/-1->3->2 [9] -1/-1/-1->3->2 [10] 0/-1/-1->3->2 [11] 0/-1/-1->3->2 [12] qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO cudaDriverVersion 12080
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Bootstrap: Using veth1:10.148.9.243<0>
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NCCL version 2.27.5+cuda12.9
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/Plugin: Plugin name set by env to libnccl-net.so
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v10 symbol.
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v9 symbol.
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v8 symbol.
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v7 symbol.
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/Plugin: Loaded net plugin NCCL RDMA Plugin v6 (v6)
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v10 symbol.
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v9 symbol.
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v8 symbol.
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v7 symbol.
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v6 symbol.
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Successfully loaded external plugin libnccl-net.so
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Plugin Path : /usr/local/lib/libnccl-net.so
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Plugin version : 1.4@Tencent Cloud
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO P2P plugin IBext
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/IB : No device found.
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/IB : No device found.
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/IB : Using [RO]; OOB veth1:10.148.9.243<0>
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NET/Socket : Using [0]veth1:10.148.9.243<0> [1]eth0:10.148.9.243<0>
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Initialized NET plugin Socket
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Assigned NET plugin Socket to comm
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Using network Socket
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO ncclCommInitRank comm 0x1e5c1d90 rank 1 nranks 4 cudaDev 1 nvmlDev 1 busId 15000 commId 0x8d6e43a37d913efb - Init START
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO RAS client listening socket at ::1<28028>
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Bootstrap timings total 0.012950 (create 0.000021, send 0.000067, recv 0.000093, ring 0.012439, delay 0.000000)
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO MNNVL busId 0x15000 fabric UUID 0.0 cliqueId 0x0 state 3 healthMask 0x0
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Setting affinity for GPU 1 to 0-47,96-143
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NVLS multicast support is available on dev 1 (NVLS_NCHANNELS 16)
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO comm 0x1e5c1d90 rank 1 nRanks 4 nNodes 1 localRanks 4 localRank 1 MNNVL 0
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Trees [0] 2/-1/-1->1->0 [1] 2/-1/-1->1->0 [2] 2/-1/-1->1->-1 [3] 2/-1/-1->1->-1 [4] -1/-1/-1->1->0 [5] -1/-1/-1->1->0 [6] 2/-1/-1->1->0 [7] 2/-1/-1->1->0 [8] 2/-1/-1->1->0 [9] 2/-1/-1->1->0 [10] 2/-1/-1->1->-1 [11] 2/-1/-1->1->-1 [12] qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Bootstrap: Using veth1:10.148.9.243<0>
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO cudaDriverVersion 12080
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NCCL version 2.27.5+cuda12.9
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/Plugin: Plugin name set by env to libnccl-net.so
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v10 symbol.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v9 symbol.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v8 symbol.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v7 symbol.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/Plugin: Loaded net plugin NCCL RDMA Plugin v6 (v6)
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v10 symbol.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v9 symbol.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v8 symbol.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v7 symbol.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v6 symbol.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Successfully loaded external plugin libnccl-net.so
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Plugin Path : /usr/local/lib/libnccl-net.so
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Plugin version : 1.4@Tencent Cloud
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO P2P plugin IBext
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/IB : No device found.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/IB : No device found.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/IB : Using [RO]; OOB veth1:10.148.9.243<0>
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NET/Socket : Using [0]veth1:10.148.9.243<0> [1]eth0:10.148.9.243<0>
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Initialized NET plugin Socket
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Assigned NET plugin Socket to comm
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Using network Socket
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO ncclCommInitRank comm 0x1e11aa90 rank 0 nranks 4 cudaDev 0 nvmlDev 0 busId e000 commId 0x8d6e43a37d913efb - Init START
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO RAS client listening socket at ::1<28028>
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Bootstrap timings total 0.092276 (create 0.000024, send 0.000070, recv 0.079392, ring 0.000038, delay 0.000000)
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO MNNVL busId 0xe000 fabric UUID 0.0 cliqueId 0x0 state 3 healthMask 0x0
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Setting affinity for GPU 0 to 0-47,96-143
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NVLS multicast support is available on dev 0 (NVLS_NCHANNELS 16)
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO comm 0x1e11aa90 rank 0 nRanks 4 nNodes 1 localRanks 4 localRank 0 MNNVL 0
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 00/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 01/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 02/16 : 0 1 2 3
qs-48753-1076858-ai-437773-deqs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO cudaDriverVersion 12080
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Bootstrap: Using veth1:10.148.9.243<0>
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NCCL version 2.27.5+cuda12.9
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/Plugin: Plugin name set by env to libnccl-net.so
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v10 symbol.
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v9 symbol.
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v8 symbol.
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/Plugin: Failed to find ncclNetPlugin_v7 symbol.
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/Plugin: Loaded net plugin NCCL RDMA Plugin v6 (v6)
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v10 symbol.
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v9 symbol.
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v8 symbol.
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v7 symbol.
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/Plugin: Failed to find ncclCollNetPlugin_v6 symbol.
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Successfully loaded external plugin libnccl-net.so
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Plugin Path : /usr/local/lib/libnccl-net.so
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Plugin version : 1.4@Tencent Cloud
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO P2P plugin IBext
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/IB : No device found.
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/IB : No device found.
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/IB : Using [RO]; OOB veth1:10.148.9.243<0>
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NET/Socket : Using [0]veth1:10.148.9.243<0> [1]eth0:10.148.9.243<0>
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Initialized NET plugin Socket
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Assigned NET plugin Socket to comm
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Using network Socket
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO ncclCommInitRank comm 0x132b12e0 rank 2 nranks 4 cudaDev 2 nvmlDev 2 busId 1c000 commId 0x8d6e43a37d913efb - Init START
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO RAS client listening socket at ::1<28028>
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Bootstrap timings total 0.037508 (create 0.000023, send 0.000069, recv 0.034522, ring 0.000037, delay 0.000000)
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO MNNVL busId 0x1c000 fabric UUID 0.0 cliqueId 0x0 state 3 healthMask 0x0
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Setting affinity for GPU 2 to 0-47,96-143
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NVLS multicast support is available on dev 2 (NVLS_NCHANNELS 16)
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO comm 0x132b12e0 rank 2 nRanks 4 nNodes 1 localRanks 4 localRank 2 MNNVL 0
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Trees [0] 3/-1/-1->2->1 [1] 3/-1/-1->2->1 [2] 3/-1/-1->2->1 [3] 3/-1/-1->2->1 [4] 3/-1/-1->2->-1 [5] 3/-1/-1->2->-1 [6] -1/-1/-1->2->1 [7] -1/-1/-1->2->1 [8] 3/-1/-1->2->1 [9] 3/-1/-1->2->1 [10] 3/-1/-1->2->1 [11] 3/-1/-1->2->1 [12] 3/fault0-0:2525924:2525924 [0] NCCL INFO Channel 03/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 04/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 05/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 06/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 07/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 08/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 09/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 10/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 11/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 12/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 13/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 14/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 15/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Trees [0] 1/-1/-1->0->-1 [1] 1/-1/-1->0->-1 [2] -1/-1/-1->0->3 [3] -1/-1/-1->0->3 [4] 1/-1/-1->0->3 [5] 1/-1/-1->0->3 [6] 1/-1/-1->0->3 [7] 1/-1/-1->0->3 [8] 1/-1/-1->0->-1 [9] 1/-1/-1->0->-1 [10] -1/-1/-1->0->3 [11] -1/-1/-1->0->3 [12] 1/-1/-1->0->3 [13] 1/-1/-1->0->3 [14] 1/-1/-1->0->3 [15] 1/-1/-1->0->3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO P2P Chunksize set to 524288
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO PROFILER/Plugin: Could not find: libnccl-profiler.so. 
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0
qs-48753-1076858-ai-437773-default0-0:2525924:2526010 [0] NCCL INFO [Proxy Service] Device 0 CPU core 5
qs-48753-1076858-ai-437773-default0-0:2525924:2526011 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 43
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO threadThresholds 8/8/64 | 32/8/64 | 512 | 512
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO 16 coll channels, 16 collnet channels, 16 nvls channels, 16 p2p channels, 16 p2p channels per peer
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO CC Off, workFifoBytes 1048576
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO TUNER/Plugin: Could not find: libnccl-tuner.so. Using internal tuner plugin.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO TUNER/Plugin: Failed to find ncclTunerPlugin_v4 symbol.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO TUNER/Plugin: Failed to find ncclTunerPlugin_v3 symbol.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO TUNER/Plugin: Failed to find ncclTunerPlugin_v2 symbol, using internal tuner instead.
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO ncclCommInitRank comm 0x1e11aa90 rank 0 nranks 4 cudaDev 0 nvmlDev 0 busId e000 commId 0x8d6e43a37d913efb - Init COMPLETE
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Init timings - ncclCommInitRank: rank 0 nranks 4 total 1.24 (kernels 0.21, alloc 0.48, bootstrap 0.09, allgathers 0.22, topo 0.03, graphs 0.01, connections 0.20, rest 0.00)
qs-48753-1076858-ai-437773-default0-0:2525924:2526016 [0] NCCL INFO Channel 00/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526016 [0] NCCL INFO Channel 01/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526016 [0] NCCL INFO Channel 02/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526016 [0] NCCL INFO Channel 03/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526016 [0] NCCL INFO Channel 04/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526016 [0] NCCL INFO Channel 05/0 : 0[0] -> 1[1] via Pqs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Assigned NET plugin Socket to comm
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Using network Socket
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO ncclCommInitRank comm 0x11dd2220 rank 0 nranks 4 cudaDev 0 nvmlDev 0 busId e000 commId 0x7ea95cdacdf94659 - Init START
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Bootstrap timings total 0.000269 (create 0.000015, send 0.000076, recv 0.000068, ring 0.000027, delay 0.000000)
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO MNNVL busId 0xe000 fabric UUID 0.0 cliqueId 0x0 state 3 healthMask 0x0
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Setting affinity for GPU 0 to 0-47,96-143
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO NVLS multicast support is available on dev 0 (NVLS_NCHANNELS 16)
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO comm 0x11dd2220 rank 0 nRanks 4 nNodes 1 localRanks 4 localRank 0 MNNVL 0
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 00/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 01/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 02/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 03/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 04/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 05/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 06/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 07/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 08/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 09/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 10/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 11/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 12/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 13/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 14/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Channel 15/16 : 0 1 2 3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Trees [0] 1/-1/-1->0->-1 [1] 1/-1/-1->0->-1 [2] -1/-1/-1->0->3 [3] -1/-1/-1->0->3 [4] 1/-1/-1->0->3 [5] 1/-1/-1->0->3 [6] 1/-1/-1->0->3 [7] 1/-1/-1->0->3 [8] 1/-1/-1->0->-1 [9] 1/-1/-1->0->-1 [10] -1/-1/-1->0->3 [11] -1/-1/-1->0->3 [12] 1/-1/-1->0->3 [13] 1/-1/-1->0->3 [14] 1/-1/-1->0->3 [15] 1/-1/-1->0->3
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO P2P Chunksize set to 524288
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Check P2P Type isAllDirectP2p 1 directMode 0
qs-48753-1076858-ai-437773-default0-0:2525924:2526132 [0] NCCL INFO [Proxy Service] Device 0 CPU core 11
qs-48753-1076858-ai-437773-default0-0:2525924:2526134 [0] NCCL INFO [Proxy Service UDS] Device 0 CPU core 119
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO threadThresholds 8/8/64 | 32/8/64 | 512 | 512
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO 16 coll channels, 16 collnet channels, 16 nvls channels, 16 p2p channels, 16 p2p channels per peer
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO CC Off, workFifoBytes 1048576
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO ncclCommInitRank comm 0x11dd2220 rank 0 nranks 4 cudaDev 0 nvmlDev 0 busId e000 commId 0x7ea95cdacdf94659 - Init COMPLETE
qs-48753-1076858-ai-437773-default0-0:2525924:2525924 [0] NCCL INFO Init timings - ncclCommInitRank: rank 0 nranks 4 total 0.16 (kernels 0.00, alloc 0.00, bootstrap 0.00, aqs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Assigned NET plugin Socket to comm
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Using network Socket
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO ncclCommInitRank comm 0x10a6be10 rank 1 nranks 4 cudaDev 1 nvmlDev 1 busId 15000 commId 0x7ea95cdacdf94659 - Init START
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Bootstrap timings total 0.000559 (create 0.000018, send 0.000062, recv 0.000191, ring 0.000028, delay 0.000000)
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO MNNVL busId 0x15000 fabric UUID 0.0 cliqueId 0x0 state 3 healthMask 0x0
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Setting affinity for GPU 1 to 0-47,96-143
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO NVLS multicast support is available on dev 1 (NVLS_NCHANNELS 16)
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO comm 0x10a6be10 rank 1 nRanks 4 nNodes 1 localRanks 4 localRank 1 MNNVL 0
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Trees [0] 2/-1/-1->1->0 [1] 2/-1/-1->1->0 [2] 2/-1/-1->1->-1 [3] 2/-1/-1->1->-1 [4] -1/-1/-1->1->0 [5] -1/-1/-1->1->0 [6] 2/-1/-1->1->0 [7] 2/-1/-1->1->0 [8] 2/-1/-1->1->0 [9] 2/-1/-1->1->0 [10] 2/-1/-1->1->-1 [11] 2/-1/-1->1->-1 [12] -1/-1/-1->1->0 [13] -1/-1/-1->1->0 [14] 2/-1/-1->1->0 [15] 2/-1/-1->1->0
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO P2P Chunksize set to 524288
qs-48753-1076858-ai-437773-default0-0:2525925:2526127 [1] NCCL INFO [Proxy Service] Device 1 CPU core 106
qs-48753-1076858-ai-437773-default0-0:2525925:2526129 [1] NCCL INFO [Proxy Service UDS] Device 1 CPU core 22
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO threadThresholds 8/8/64 | 32/8/64 | 512 | 512
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO 16 coll channels, 16 collnet channels, 16 nvls channels, 16 p2p channels, 16 p2p channels per peer
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO ncclCommInitRank comm 0x10a6be10 rank 1 nranks 4 cudaDev 1 nvmlDev 1 busId 15000 commId 0x7ea95cdacdf94659 - Init COMPLETE
qs-48753-1076858-ai-437773-default0-0:2525925:2525925 [1] NCCL INFO Init timings - ncclCommInitRank: rank 1 nranks 4 total 0.16 (kernels 0.00, alloc 0.00, bootstrap 0.00, allgathers 0.00, topo 0.02, graphs 0.01, connections 0.13, rest 0.00)
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 00/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 01/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 02/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 03/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 04/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 05/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 06/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 07/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 08/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 09/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 10/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 11/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 12/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 13/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Chaqs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Assigned NET plugin Socket to comm
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Using network Socket
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO ncclCommInitRank comm 0x184f33d0 rank 3 nranks 4 cudaDev 3 nvmlDev 3 busId 67000 commId 0x7ea95cdacdf94659 - Init START
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Bootstrap timings total 0.000519 (create 0.000018, send 0.000056, recv 0.000304, ring 0.000057, delay 0.000000)
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO MNNVL busId 0x67000 fabric UUID 0.0 cliqueId 0x0 state 3 healthMask 0x0
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Setting affinity for GPU 3 to 0-47,96-143
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO NVLS multicast support is available on dev 3 (NVLS_NCHANNELS 16)
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO comm 0x184f33d0 rank 3 nRanks 4 nNodes 1 localRanks 4 localRank 3 MNNVL 0
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Trees [0] -1/-1/-1->3->2 [1] -1/-1/-1->3->2 [2] 0/-1/-1->3->2 [3] 0/-1/-1->3->2 [4] 0/-1/-1->3->2 [5] 0/-1/-1->3->2 [6] 0/-1/-1->3->-1 [7] 0/-1/-1->3->-1 [8] -1/-1/-1->3->2 [9] -1/-1/-1->3->2 [10] 0/-1/-1->3->2 [11] 0/-1/-1->3->2 [12] 0/-1/-1->3->2 [13] 0/-1/-1->3->2 [14] 0/-1/-1->3->-1 [15] 0/-1/-1->3->-1
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO P2P Chunksize set to 524288
qs-48753-1076858-ai-437773-default0-0:2525927:2526128 [3] NCCL INFO [Proxy Service] Device 3 CPU core 101
qs-48753-1076858-ai-437773-default0-0:2525927:2526130 [3] NCCL INFO [Proxy Service UDS] Device 3 CPU core 33
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO threadThresholds 8/8/64 | 32/8/64 | 512 | 512
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO 16 coll channels, 16 collnet channels, 16 nvls channels, 16 p2p channels, 16 p2p channels per peer
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO ncclCommInitRank comm 0x184f33d0 rank 3 nranks 4 cudaDev 3 nvmlDev 3 busId 67000 commId 0x7ea95cdacdf94659 - Init COMPLETE
qs-48753-1076858-ai-437773-default0-0:2525927:2525927 [3] NCCL INFO Init timings - ncclCommInitRank: rank 3 nranks 4 total 0.16 (kernels 0.00, alloc 0.00, bootstrap 0.00, allgathers 0.00, topo 0.02, graphs 0.01, connections 0.13, rest 0.00)
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 00/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 01/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 02/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 03/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 04/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 05/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 06/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 07/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 08/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 09/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 10/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 11/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 12/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 13/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Chaqs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Assigned NET plugin Socket to comm
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Using network Socket
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO ncclCommInitRank comm 0x17e213b0 rank 2 nranks 4 cudaDev 2 nvmlDev 2 busId 1c000 commId 0x7ea95cdacdf94659 - Init START
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Bootstrap timings total 0.000424 (create 0.000017, send 0.000065, recv 0.000087, ring 0.000163, delay 0.000000)
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO MNNVL busId 0x1c000 fabric UUID 0.0 cliqueId 0x0 state 3 healthMask 0x0
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Setting affinity for GPU 2 to 0-47,96-143
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO NVLS multicast support is available on dev 2 (NVLS_NCHANNELS 16)
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO comm 0x17e213b0 rank 2 nRanks 4 nNodes 1 localRanks 4 localRank 2 MNNVL 0
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Trees [0] 3/-1/-1->2->1 [1] 3/-1/-1->2->1 [2] 3/-1/-1->2->1 [3] 3/-1/-1->2->1 [4] 3/-1/-1->2->-1 [5] 3/-1/-1->2->-1 [6] -1/-1/-1->2->1 [7] -1/-1/-1->2->1 [8] 3/-1/-1->2->1 [9] 3/-1/-1->2->1 [10] 3/-1/-1->2->1 [11] 3/-1/-1->2->1 [12] 3/-1/-1->2->-1 [13] 3/-1/-1->2->-1 [14] -1/-1/-1->2->1 [15] -1/-1/-1->2->1
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO P2P Chunksize set to 524288
qs-48753-1076858-ai-437773-default0-0:2525926:2526131 [2] NCCL INFO [Proxy Service] Device 2 CPU core 43
qs-48753-1076858-ai-437773-default0-0:2525926:2526133 [2] NCCL INFO [Proxy Service UDS] Device 2 CPU core 99
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO threadThresholds 8/8/64 | 32/8/64 | 512 | 512
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO 16 coll channels, 16 collnet channels, 16 nvls channels, 16 p2p channels, 16 p2p channels per peer
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO ncclCommInitRank comm 0x17e213b0 rank 2 nranks 4 cudaDev 2 nvmlDev 2 busId 1c000 commId 0x7ea95cdacdf94659 - Init COMPLETE
qs-48753-1076858-ai-437773-default0-0:2525926:2525926 [2] NCCL INFO Init timings - ncclCommInitRank: rank 2 nranks 4 total 0.16 (kernels 0.00, alloc 0.00, bootstrap 0.00, allgathers 0.00, topo 0.02, graphs 0.01, connections 0.13, rest 0.00)
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 00/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 01/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 02/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 03/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 04/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 05/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 06/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 07/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 08/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 09/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 10/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 11/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 12/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 13/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO ChanINFO 01-20 15:00:39 [parallel_state.py:1408] rank 1 in world size 4 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 1, EP rank 1
INFO 01-20 15:00:39 [parallel_state.py:1408] rank 0 in world size 4 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 0, EP rank 0
INFO 01-20 15:00:39 [parallel_state.py:1408] rank 3 in world size 4 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 3, EP rank 3
INFO 01-20 15:00:39 [parallel_state.py:1408] rank 2 in world size 4 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 2, EP rank 2
(Worker_TP0 pid=2525924) INFO 01-20 15:00:47 [gpu_model_runner.py:3467] Starting to load model /mnt/tidal-alsh01/usr/dawo/dawo/model_ori/ZhipuAI/GLM-4.6V...
(Worker_TP0 pid=2525924) INFO 01-20 15:00:47 [cuda.py:411] Using FLASH_ATTN attention backend out of potential backends: ['FLASH_ATTN', 'FLASHINFER', 'TRITON_ATTN', 'FLEX_ATTENTION']
(Worker_TP0 pid=2525924) INFO 01-20 15:00:47 [layer.py:379] Enabled separate cuda stream for MoE shared_experts
(Worker_TP2 pid=2525926) INFO 01-20 15:00:47 [cuda.py:411] Using FLASH_ATTN attention backend out of potential backends: ['FLASH_ATTN', 'FLASHINFER', 'TRITON_ATTN', 'FLEX_ATTENTION']
(Worker_TP2 pid=2525926) INFO 01-20 15:00:47 [layer.py:379] Enabled separate cuda stream for MoE shared_experts
(Worker_TP3 pid=2525927) INFO 01-20 15:00:47 [cuda.py:411] Using FLASH_ATTN attention backend out of potential backends: ['FLASH_ATTN', 'FLASHINFER', 'TRITON_ATTN', 'FLEX_ATTENTION']
(Worker_TP3 pid=2525927) INFO 01-20 15:00:47 [layer.py:379] Enabled separate cuda stream for MoE shared_experts
(Worker_TP1 pid=2525925) INFO 01-20 15:00:47 [cuda.py:411] Using FLASH_ATTN attention backend out of potential backends: ['FLASH_ATTN', 'FLASHINFER', 'TRITON_ATTN', 'FLEX_ATTENTION']
(Worker_TP1 pid=2525925) INFO 01-20 15:00:47 [layer.py:379] Enabled separate cuda stream for MoE shared_experts
(Worker_TP0 pid=2525924) INFO 01-20 15:10:06 [default_loader.py:308] Loading weights took 557.62 seconds
(Worker_TP0 pid=2525924) INFO 01-20 15:10:07 [gpu_model_runner.py:3549] Model loading took 50.3468 GiB memory and 558.793287 seconds
(Worker_TP2 pid=2525926) INFO 01-20 15:10:07 [gpu_model_runner.py:4306] Encoder cache will be initialized with a budget of 8192 tokens, and profiled with 1 video items of the maximum feature size.
(Worker_TP1 pid=2525925) INFO 01-20 15:10:07 [gpu_model_runner.py:4306] Encoder cache will be initialized with a budget of 8192 tokens, and profiled with 1 video items of the maximum feature size.
(Worker_TP0 pid=2525924) INFO 01-20 15:10:07 [gpu_model_runner.py:4306] Encoder cache will be initialized with a budget of 8192 tokens, and profiled with 1 video items of the maximum feature size.
(Worker_TP3 pid=2525927) INFO 01-20 15:10:07 [gpu_model_runner.py:4306] Encoder cache will be initialized with a budget of 8192 tokens, and profiled with 1 video items of the maximum feature size.
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] WorkerProc hit an exception.
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] Traceback (most recent call last):
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/executor/multiproc_executor.py", line 817, in worker_busy_loop
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     output = func(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return func(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/worker/gpu_worker.py", line 324, in determine_available_memory
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     self.model_runner.profile_run()
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/worker/gpu_model_runner.py", line 4322, in profile_run
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     dummy_encoder_outputs = self.model.embed_multimodal(
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         **batched_dummy_mm_inputs
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 1646, in embed_multimodal
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     video_embeddings = self._process_video_input(multimodal_input)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 1593, in _process_video_input
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     video_embeds = self.visual(
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         pixel_values_videos, grid_thw=grid_thw.tolist()
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._call_impl(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return forward_call(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 820, in forward
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     x = blk(
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         x,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<3 lines>...
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         max_seqlen=max_seqlen,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._call_impl(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return forward_call(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 444, in forward
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     x_attn = self.attn(
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         self.norm1(x),
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<3 lines>...
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         max_seqlen=max_seqlen,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._call_impl(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return forward_call(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 362, in forward
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     output = self.flash_attn_varlen_func(
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         q,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<7 lines>...
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         causal=False,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/vllm_flash_attn/flash_attn_interface.py", line 253, in flash_attn_varlen_func
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     out, softmax_lse = torch.ops._vllm_fa2_C.varlen_fwd(
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]                        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         q, k, v,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         ^^^^^^^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<19 lines>...
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         None,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         ^^^^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/_ops.py", line 1255, in __call__
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._op(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] torch.AcceleratorError: CUDA error: the provided PTX was compiled with an unsupported toolchain.
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] Search for `cudaErrorUnsupportedPtxVersion' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information.
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] For debugging consider passing CUDA_LAUNCH_BLOCKING=1
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] 
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] Traceback (most recent call last):
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/executor/multiproc_executor.py", line 817, in worker_busy_loop
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     output = func(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return func(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/worker/gpu_worker.py", line 324, in determine_available_memory
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     self.model_runner.profile_run()
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/worker/gpu_model_runner.py", line 4322, in profile_run
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     dummy_encoder_outputs = self.model.embed_multimodal(
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         **batched_dummy_mm_inputs
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 1646, in embed_multimodal
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     video_embeddings = self._process_video_input(multimodal_input)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 1593, in _process_video_input
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     video_embeds = self.visual(
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         pixel_values_videos, grid_thw=grid_thw.tolist()
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._call_impl(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return forward_call(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 820, in forward
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     x = blk(
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         x,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<3 lines>...
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         max_seqlen=max_seqlen,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._call_impl(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return forward_call(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 444, in forward
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     x_attn = self.attn(
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         self.norm1(x),
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<3 lines>...
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         max_seqlen=max_seqlen,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._call_impl(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return forward_call(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 362, in forward
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     output = self.flash_attn_varlen_func(
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         q,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<7 lines>...
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         causal=False,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/vllm_flash_attn/flash_attn_interface.py", line 253, in flash_attn_varlen_func
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     out, softmax_lse = torch.ops._vllm_fa2_C.varlen_fwd(
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]                        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         q, k, v,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         ^^^^^^^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<19 lines>...
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         None,
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         ^^^^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/_ops.py", line 1255, in __call__
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._op(*args, **kwargs)
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] torch.AcceleratorError: CUDA error: the provided PTX was compiled with an unsupported toolchain.
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] Search for `cudaErrorUnsupportedPtxVersion' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information.
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] For debugging consider passing CUDA_LAUNCH_BLOCKING=1
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] 
(Worker_TP3 pid=2525927) ERROR 01-20 15:10:10 [multiproc_executor.py:822] 
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] WorkerProc hit an exception.
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] Traceback (most recent call last):
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/executor/multiproc_executor.py", line 817, in worker_busy_loop
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     output = func(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return func(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/worker/gpu_worker.py", line 324, in determine_available_memory
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     self.model_runner.profile_run()
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/worker/gpu_model_runner.py", line 4322, in profile_run
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     dummy_encoder_outputs = self.model.embed_multimodal(
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         **batched_dummy_mm_inputs
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 1646, in embed_multimodal
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     video_embeddings = self._process_video_input(multimodal_input)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 1593, in _process_video_input
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     video_embeds = self.visual(
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         pixel_values_videos, grid_thw=grid_thw.tolist()
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._call_impl(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return forward_call(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 820, in forward
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     x = blk(
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         x,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<3 lines>...
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         max_seqlen=max_seqlen,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._call_impl(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return forward_call(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 444, in forward
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     x_attn = self.attn(
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         self.norm1(x),
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<3 lines>...
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         max_seqlen=max_seqlen,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._call_impl(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return forward_call(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 362, in forward
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     output = self.flash_attn_varlen_func(
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         q,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<7 lines>...
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         causal=False,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/vllm_flash_attn/flash_attn_interface.py", line 253, in flash_attn_varlen_func
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     out, softmax_lse = torch.ops._vllm_fa2_C.varlen_fwd(
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]                        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         q, k, v,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         ^^^^^^^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<19 lines>...
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         None,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         ^^^^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/_ops.py", line 1255, in __call__
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._op(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] torch.AcceleratorError: CUDA error: the provided PTX was compiled with an unsupported toolchain.
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] Search for `cudaErrorUnsupportedPtxVersion' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information.
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] For debugging consider passing CUDA_LAUNCH_BLOCKING=1
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] 
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] Traceback (most recent call last):
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/executor/multiproc_executor.py", line 817, in worker_busy_loop
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     output = func(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return func(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/worker/gpu_worker.py", line 324, in determine_available_memory
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     self.model_runner.profile_run()
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/worker/gpu_model_runner.py", line 4322, in profile_run
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     dummy_encoder_outputs = self.model.embed_multimodal(
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         **batched_dummy_mm_inputs
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 1646, in embed_multimodal
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     video_embeddings = self._process_video_input(multimodal_input)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 1593, in _process_video_input
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     video_embeds = self.visual(
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         pixel_values_videos, grid_thw=grid_thw.tolist()
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._call_impl(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return forward_call(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 820, in forward
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     x = blk(
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         x,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<3 lines>...
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         max_seqlen=max_seqlen,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._call_impl(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return forward_call(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 444, in forward
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     x_attn = self.attn(
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         self.norm1(x),
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<3 lines>...
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         max_seqlen=max_seqlen,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1775, in _wrapped_call_impl
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._call_impl(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return forward_call(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/model_executor/models/glm4_1v.py", line 362, in forward
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     output = self.flash_attn_varlen_func(
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         q,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<7 lines>...
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         causal=False,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/vllm_flash_attn/flash_attn_interface.py", line 253, in flash_attn_varlen_func
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     out, softmax_lse = torch.ops._vllm_fa2_C.varlen_fwd(
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]                        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         q, k, v,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         ^^^^^^^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ...<19 lines>...
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         None,
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]         ^^^^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     )
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     ^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/torch/_ops.py", line 1255, in __call__
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]     return self._op(*args, **kwargs)
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822]            ~~~~~~~~^^^^^^^^^^^^^^^^^
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] torch.AcceleratorError: CUDA error: the provided PTX was compiled with an unsupported toolchain.
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] Search for `cudaErrorUnsupportedPtxVersion' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information.
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] For debugging consider passing CUDA_LAUNCH_BLOCKING=1
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] 
(Worker_TP0 pid=2525924) ERROR 01-20 15:10:10 [multiproc_executor.py:822] 
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843] EngineCore failed to start.
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843] Traceback (most recent call last):
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/engine/core.py", line 834, in run_engine_core
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     engine_core = EngineCoreProc(*args, **kwargs)
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/engine/core.py", line 610, in __init__
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     super().__init__(
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     ~~~~~~~~~~~~~~~~^
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]         vllm_config, executor_class, log_stats, executor_fail_callback
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     )
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     ^
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/engine/core.py", line 109, in __init__
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     num_gpu_blocks, num_cpu_blocks, kv_cache_config = self._initialize_kv_caches(
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]                                                       ~~~~~~~~~~~~~~~~~~~~~~~~~~^
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]         vllm_config
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]         ^^^^^^^^^^^
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     )
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     ^
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/engine/core.py", line 235, in _initialize_kv_caches
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     available_gpu_memory = self.model_executor.determine_available_memory()
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/executor/abstract.py", line 126, in determine_available_memory
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     return self.collective_rpc("determine_available_memory")
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]            ~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/executor/multiproc_executor.py", line 361, in collective_rpc
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     return aggregate(get_response())
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]                      ~~~~~~~~~~~~^^
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]   File "/root/anaconda3/envs/glm46v/lib/python3.13/site-packages/vllm/v1/executor/multiproc_executor.py", line 344, in get_response
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     raise RuntimeError(
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     ...<2 lines>...
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843]     )
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843] RuntimeError: Worker failed with error 'CUDA error: the provided PTX was compiled with an unsupported toolchain.
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843] Search for `cudaErrorUnsupportedPtxVersion' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information.
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843] CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843] For debugging consider passing CUDA_LAUNCH_BLOCKING=1
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843] Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
(EngineCore_DP0 pid=2525786) ERROR 01-20 15:10:10 [core.py:843] ', please check the stack trace above for the root cause
(Worker_TP0 pid=2525924) INFO 01-20 15:10:10 [multiproc_executor.py:709] Parent process exited, terminating worker
(Worker_TP2 pid=2525926) INFO 01-20 15:10:10 [multiproc_executor.py:709] Parent process exited, terminating worker
(Worker_TP3 pid=2525927) INFO 01-20 15:10:10 [multiproc_executor.py:709] Parent process exited, terminating worker
(Worker_TP1 pid=2525925) INFO 01-20 15:10:10 [multiproc_executor.py:709] Parent process exited, terminating worker
nnel 14/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Channel 15/0 : 1[1] -> 2[2] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525925:2526136 [1] NCCL INFO Connected all rings, use ring PXN 0 GDR 1
llgathers 0.00, topo 0.02, graphs 0.01, connections 0.13, rest 0.00)
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 00/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 01/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 02/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 03/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 04/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 05/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 06/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 07/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 08/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 09/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 10/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 11/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 12/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 13/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 14/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Channel 15/0 : 0[0] -> 1[1] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525924:2526137 [0] NCCL INFO Connected all rings, use ring PXN 0 GDR 1
nel 14/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Channel 15/0 : 2[2] -> 3[3] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525926:2526138 [2] NCCL INFO Connected all rings, use ring PXN 0 GDR 1
nnel 14/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Channel 15/0 : 3[3] -> 0[0] via P2P/CUMEM
qs-48753-1076858-ai-437773-default0-0:2525927:2526135 [3] NCCL INFO Connected all rings, use ring PXN 0 GDR 1
qs-48753-1076858-ai-437773-default0-0:2525926:2526131 [0] NCCL INFO [Service thread] Connection closed by localRank 3
qs-48753-1076858-ai-437773-default0-0:2525926:2526012 [0] NCCL INFO [Service thread] Connection closed by localRank 3
qs-48753-1076858-ai-437773-default0-0:2525924:2526132 [0] NCCL INFO [Service thread] Connection closed by localRank 3
qs-48753-1076858-ai-437773-default0-0:2525924:2526010 [0] NCCL INFO [Service thread] Connection closed by localRank 3