vllm: [Bug]: Failed to run model Qwen3-30B-A3B on DGX V100x4

Your current environment

Environment in the container:

INFO 04-29 06:34:01 [init.py:239] Automatically detected platform cuda. Collecting environment information… PyTorch version: 2.6.0+cu124 Is debug build: False CUDA used to build PyTorch: 12.4 ROCM used to build PyTorch: N/A

OS: Ubuntu 22.04.4 LTS (x86_64) GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 Clang version: Could not collect CMake version: version 4.0.0 Libc version: glibc-2.35

Python version: 3.12.10 (main, Apr 9 2025, 08:55:05) [GCC 11.4.0] (64-bit runtime) Python platform: Linux-4.15.0-213-generic-x86_64-with-glibc2.35 Is CUDA available: True CUDA runtime version: 12.4.131 CUDA_MODULE_LOADING set to: LAZY GPU models and configuration: GPU 0: Tesla V100-DGXS-32GB GPU 1: Tesla V100-DGXS-32GB GPU 2: Tesla V100-DGXS-32GB GPU 3: Tesla V100-DGXS-32GB

Nvidia driver version: 560.35.05 cuDNN version: Could not collect HIP runtime version: N/A MIOpen runtime version: N/A Is XNNPACK available: True

CPU: Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Address sizes: 46 bits physical, 48 bits virtual Byte Order: Little Endian CPU(s): 40 On-line CPU(s) list: 0-39 Vendor ID: GenuineIntel Model name: Intel® Xeon® CPU E5-2698 v4 @ 2.20GHz CPU family: 6 Model: 79 Thread(s) per core: 2 Core(s) per socket: 20 Socket(s): 1 Stepping: 1 CPU max MHz: 3600.0000 CPU min MHz: 1200.0000 BogoMIPS: 4397.45 Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 cdp_l3 invpcid_single pti intel_ppin ssbd ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm cqm rdt_a rdseed adx smap intel_pt xsaveopt cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local dtherm ida arat pln pts md_clear flush_l1d Virtualization: VT-x L1d cache: 640 KiB (20 instances) L1i cache: 640 KiB (20 instances) L2 cache: 5 MiB (20 instances) L3 cache: 50 MiB (1 instance) NUMA node(s): 1 NUMA node0 CPU(s): 0-39 Vulnerability Itlb multihit: KVM: Mitigation: Split huge pages Vulnerability L1tf: Mitigation; PTE Inversion; VMX conditional cache flushes, SMT vulnerable Vulnerability Mds: Mitigation; Clear CPU buffers; SMT vulnerable Vulnerability Meltdown: Mitigation; PTI Vulnerability Mmio stale data: Mitigation; Clear CPU buffers; SMT vulnerable Vulnerability Retbleed: Not affected Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization Vulnerability Spectre v2: Mitigation; Retpolines, IBPB conditional, IBRS_FW, STIBP conditional, RSB filling, PBRSB-eIBRS Not affected Vulnerability Srbds: Not affected Vulnerability Tsx async abort: Mitigation; Clear CPU buffers; SMT vulnerable

Versions of relevant libraries: [pip3] flashinfer-python==0.2.1.post2+cu124torch2.6 [pip3] numpy==2.2.5 [pip3] nvidia-cublas-cu12==12.4.5.8 [pip3] nvidia-cuda-cupti-cu12==12.4.127 [pip3] nvidia-cuda-nvrtc-cu12==12.4.127 [pip3] nvidia-cuda-runtime-cu12==12.4.127 [pip3] nvidia-cudnn-cu12==9.1.0.70 [pip3] nvidia-cufft-cu12==11.2.1.3 [pip3] nvidia-curand-cu12==10.3.5.147 [pip3] nvidia-cusolver-cu12==11.6.1.9 [pip3] nvidia-cusparse-cu12==12.3.1.170 [pip3] nvidia-cusparselt-cu12==0.6.2 [pip3] nvidia-nccl-cu12==2.21.5 [pip3] nvidia-nvjitlink-cu12==12.4.127 [pip3] nvidia-nvtx-cu12==12.4.127 [pip3] pyzmq==26.4.0 [pip3] torch==2.6.0 [pip3] torchaudio==2.6.0 [pip3] torchvision==0.21.0 [pip3] transformers==4.51.3 [pip3] triton==3.2.0 [conda] Could not collect ROCM Version: Could not collect Neuron SDK Version: N/A vLLM Version: 0.8.5 vLLM Build Flags: CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled GPU Topology: GPU0 GPU1 GPU2 GPU3 CPU Affinity NUMA Affinity GPU NUMA ID GPU0 X NV1 NV1 NV2 0-39 0 N/A GPU1 NV1 X NV2 NV1 0-39 0 N/A GPU2 NV1 NV2 X NV1 0-39 0 N/A GPU3 NV2 NV1 NV1 X 0-39 0 N/A

Legend:

X = Self SYS = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI) NODE = Connection traversing PCIe as well as the interconnect between PCIe Host Bridges within a NUMA node PHB = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU) PXB = Connection traversing multiple PCIe bridges (without traversing the PCIe Host Bridge) PIX = Connection traversing at most a single PCIe bridge NV# = Connection traversing a bonded set of # NVLinks

NVIDIA_VISIBLE_DEVICES=all NVIDIA_REQUIRE_CUDA=cuda>=12.4 brand=tesla,driver>=470,driver<471 brand=unknown,driver>=470,driver<471 brand=nvidia,driver>=470,driver<471 brand=nvidiartx,driver>=470,driver<471 brand=geforce,driver>=470,driver<471 brand=geforcertx,driver>=470,driver<471 brand=quadro,driver>=470,driver<471 brand=quadrortx,driver>=470,driver<471 brand=titan,driver>=470,driver<471 brand=titanrtx,driver>=470,driver<471 brand=tesla,driver>=525,driver<526 brand=unknown,driver>=525,driver<526 brand=nvidia,driver>=525,driver<526 brand=nvidiartx,driver>=525,driver<526 brand=geforce,driver>=525,driver<526 brand=geforcertx,driver>=525,driver<526 brand=quadro,driver>=525,driver<526 brand=quadrortx,driver>=525,driver<526 brand=titan,driver>=525,driver<526 brand=titanrtx,driver>=525,driver<526 brand=tesla,driver>=535,driver<536 brand=unknown,driver>=535,driver<536 brand=nvidia,driver>=535,driver<536 brand=nvidiartx,driver>=535,driver<536 brand=geforce,driver>=535,driver<536 brand=geforcertx,driver>=535,driver<536 brand=quadro,driver>=535,driver<536 brand=quadrortx,driver>=535,driver<536 brand=titan,driver>=535,driver<536 brand=titanrtx,driver>=535,driver<536 NCCL_VERSION=2.20.5-1 NVIDIA_DRIVER_CAPABILITIES=compute,utility NVIDIA_PRODUCT_NAME=CUDA VLLM_USAGE_SOURCE=production-docker-image CUDA_VERSION=12.4.0 LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64 NCCL_CUMEM_ENABLE=0 PYTORCH_NVML_BASED_CUDA_CHECK=1 TORCHINDUCTOR_COMPILE_THREADS=1 CUDA_MODULE_LOADING=LAZY

🐛 Describe the bug

command

docker run --gpus all -v ~/.cache/huggingface:/root/.cache/huggingface -p 8000:8000 --ipc=host vllm/vllm-openai:v0.8.5 --model Qwen/Qwen3-30B-A3B --tensor-parallel-size 4 --dtype=half --enable-reasoning --reasoning-parser deepseek_r1 --max-model-len 32768 --enforce-eager

Output:

NFO 04-29 06:17:33 [__init__.py:239] Automatically detected platform cuda.
INFO 04-29 06:17:38 [api_server.py:1043] vLLM API server version 0.8.5
INFO 04-29 06:17:38 [api_server.py:1044] args: Namespace(host=None, port=8000, uvicorn_log_level='info', disable_uvicorn_access_log=False, allow_credentials=False, allowed_origins=['*'], allowed_methods=['*'], allowed_headers=['*'], api_key=None, lora_modules=None, prompt_adapters=None, chat_template=None, chat_template_content_format='auto', response_role='assistant', ssl_keyfile=None, ssl_certfile=None, ssl_ca_certs=None, enable_ssl_refresh=False, ssl_cert_reqs=0, root_path=None, middleware=[], return_tokens_as_token_ids=False, disable_frontend_multiprocessing=False, enable_request_id_headers=False, enable_auto_tool_choice=False, tool_call_parser=None, tool_parser_plugin='', model='Qwen/Qwen3-30B-A3B', task='auto', tokenizer=None, hf_config_path=None, skip_tokenizer_init=False, revision=None, code_revision=None, tokenizer_revision=None, tokenizer_mode='auto', trust_remote_code=False, allowed_local_media_path=None, load_format='auto', download_dir=None, model_loader_extra_config={}, use_tqdm_on_load=True, config_format=<ConfigFormat.AUTO: 'auto'>, dtype='half', max_model_len=32768, guided_decoding_backend='auto', reasoning_parser='deepseek_r1', logits_processor_pattern=None, model_impl='auto', distributed_executor_backend=None, pipeline_parallel_size=1, tensor_parallel_size=4, data_parallel_size=1, enable_expert_parallel=False, max_parallel_loading_workers=None, ray_workers_use_nsight=False, disable_custom_all_reduce=False, block_size=None, gpu_memory_utilization=0.9, swap_space=4, kv_cache_dtype='auto', num_gpu_blocks_override=None, enable_prefix_caching=None, prefix_caching_hash_algo='builtin', cpu_offload_gb=0, calculate_kv_scales=False, disable_sliding_window=False, use_v2_block_manager=True, seed=None, max_logprobs=20, disable_log_stats=False, quantization=None, rope_scaling=None, rope_theta=None, hf_token=None, hf_overrides=None, enforce_eager=True, max_seq_len_to_capture=8192, tokenizer_pool_size=0, tokenizer_pool_type='ray', tokenizer_pool_extra_config={}, limit_mm_per_prompt={}, mm_processor_kwargs=None, disable_mm_preprocessor_cache=False, enable_lora=None, enable_lora_bias=False, max_loras=1, max_lora_rank=16, lora_extra_vocab_size=256, lora_dtype='auto', long_lora_scaling_factors=None, max_cpu_loras=None, fully_sharded_loras=False, enable_prompt_adapter=None, max_prompt_adapters=1, max_prompt_adapter_token=0, device='auto', speculative_config=None, ignore_patterns=[], served_model_name=None, qlora_adapter_name_or_path=None, show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None, disable_async_output_proc=False, max_num_batched_tokens=None, max_num_seqs=None, max_num_partial_prefills=1, max_long_partial_prefills=1, long_prefill_token_threshold=0, num_lookahead_slots=0, scheduler_delay_factor=0.0, preemption_mode=None, num_scheduler_steps=1, multi_step_stream_outputs=True, scheduling_policy='fcfs', enable_chunked_prefill=None, disable_chunked_mm_input=False, scheduler_cls='vllm.core.scheduler.Scheduler', override_neuron_config=None, override_pooler_config=None, compilation_config=None, kv_transfer_config=None, worker_cls='auto', worker_extension_cls='', generation_config='auto', override_generation_config=None, enable_sleep_mode=False, additional_config=None, enable_reasoning=True, disable_cascade_attn=False, disable_log_requests=False, max_log_len=None, disable_fastapi_docs=False, enable_prompt_tokens_details=False, enable_server_load_tracking=False)
WARNING 04-29 06:17:39 [config.py:2972] Casting torch.bfloat16 to torch.float16.
INFO 04-29 06:17:50 [config.py:717] This model supports multiple tasks: {'classify', 'embed', 'generate', 'score', 'reward'}. Defaulting to 'generate'.
WARNING 04-29 06:17:50 [arg_utils.py:1658] Compute Capability < 8.0 is not supported by the V1 Engine. Falling back to V0. 
INFO 04-29 06:17:50 [config.py:1770] Defaulting to use mp for distributed inference
WARNING 04-29 06:17:50 [cuda.py:93] To see benefits of async output processing, enable CUDA graph. Since, enforce-eager is enabled, async output processor cannot be used
INFO 04-29 06:17:50 [api_server.py:246] Started engine process with PID 97
INFO 04-29 06:17:55 [__init__.py:239] Automatically detected platform cuda.
INFO 04-29 06:17:57 [llm_engine.py:240] Initializing a V0 LLM engine (v0.8.5) with config: model='Qwen/Qwen3-30B-A3B', speculative_config=None, tokenizer='Qwen/Qwen3-30B-A3B', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, override_neuron_config=None, tokenizer_revision=None, trust_remote_code=False, dtype=torch.float16, max_seq_len=32768, download_dir=None, load_format=LoadFormat.AUTO, tensor_parallel_size=4, pipeline_parallel_size=1, disable_custom_all_reduce=False, quantization=None, enforce_eager=True, kv_cache_dtype=auto,  device_config=cuda, decoding_config=DecodingConfig(guided_decoding_backend='auto', reasoning_backend='deepseek_r1'), observability_config=ObservabilityConfig(show_hidden_metrics=False, otlp_traces_endpoint=None, collect_model_forward_time=False, collect_model_execute_time=False), seed=None, served_model_name=Qwen/Qwen3-30B-A3B, num_scheduler_steps=1, multi_step_stream_outputs=True, enable_prefix_caching=None, chunked_prefill_enabled=False, use_async_output_proc=False, disable_mm_preprocessor_cache=False, mm_processor_kwargs=None, pooler_config=None, compilation_config={"splitting_ops":[],"compile_sizes":[],"cudagraph_capture_sizes":[],"max_capture_size":0}, use_cached_outputs=True, 
WARNING 04-29 06:17:58 [multiproc_worker_utils.py:306] Reducing Torch parallelism from 20 threads to 1 to avoid unnecessary CPU contention. Set OMP_NUM_THREADS in the external environment to tune this value as needed.
INFO 04-29 06:18:00 [cuda.py:240] Cannot use FlashAttention-2 backend for Volta and Turing GPUs.
INFO 04-29 06:18:00 [cuda.py:289] Using XFormers backend.
INFO 04-29 06:18:03 [__init__.py:239] Automatically detected platform cuda.
INFO 04-29 06:18:03 [__init__.py:239] Automatically detected platform cuda.
INFO 04-29 06:18:03 [__init__.py:239] Automatically detected platform cuda.
(VllmWorkerProcess pid=146) INFO 04-29 06:18:05 [multiproc_worker_utils.py:225] Worker ready; awaiting tasks
(VllmWorkerProcess pid=147) INFO 04-29 06:18:05 [multiproc_worker_utils.py:225] Worker ready; awaiting tasks
(VllmWorkerProcess pid=145) INFO 04-29 06:18:05 [multiproc_worker_utils.py:225] Worker ready; awaiting tasks
(VllmWorkerProcess pid=147) INFO 04-29 06:18:07 [cuda.py:240] Cannot use FlashAttention-2 backend for Volta and Turing GPUs.
(VllmWorkerProcess pid=147) INFO 04-29 06:18:07 [cuda.py:289] Using XFormers backend.
(VllmWorkerProcess pid=146) INFO 04-29 06:18:07 [cuda.py:240] Cannot use FlashAttention-2 backend for Volta and Turing GPUs.
(VllmWorkerProcess pid=145) INFO 04-29 06:18:07 [cuda.py:240] Cannot use FlashAttention-2 backend for Volta and Turing GPUs.
(VllmWorkerProcess pid=146) INFO 04-29 06:18:07 [cuda.py:289] Using XFormers backend.
(VllmWorkerProcess pid=145) INFO 04-29 06:18:07 [cuda.py:289] Using XFormers backend.
INFO 04-29 06:18:08 [utils.py:1055] Found nccl from library libnccl.so.2
(VllmWorkerProcess pid=147) INFO 04-29 06:18:08 [utils.py:1055] Found nccl from library libnccl.so.2
INFO 04-29 06:18:08 [pynccl.py:69] vLLM is using nccl==2.21.5
(VllmWorkerProcess pid=147) INFO 04-29 06:18:08 [pynccl.py:69] vLLM is using nccl==2.21.5
(VllmWorkerProcess pid=146) INFO 04-29 06:18:08 [utils.py:1055] Found nccl from library libnccl.so.2
(VllmWorkerProcess pid=146) INFO 04-29 06:18:08 [pynccl.py:69] vLLM is using nccl==2.21.5
(VllmWorkerProcess pid=145) INFO 04-29 06:18:08 [utils.py:1055] Found nccl from library libnccl.so.2
(VllmWorkerProcess pid=145) INFO 04-29 06:18:08 [pynccl.py:69] vLLM is using nccl==2.21.5
INFO 04-29 06:18:09 [custom_all_reduce_utils.py:206] generating GPU P2P access cache in /root/.cache/vllm/gpu_p2p_access_cache_for_0,1,2,3.json
INFO 04-29 06:18:27 [custom_all_reduce_utils.py:244] reading GPU P2P access cache from /root/.cache/vllm/gpu_p2p_access_cache_for_0,1,2,3.json
(VllmWorkerProcess pid=146) INFO 04-29 06:18:27 [custom_all_reduce_utils.py:244] reading GPU P2P access cache from /root/.cache/vllm/gpu_p2p_access_cache_for_0,1,2,3.json
(VllmWorkerProcess pid=147) INFO 04-29 06:18:27 [custom_all_reduce_utils.py:244] reading GPU P2P access cache from /root/.cache/vllm/gpu_p2p_access_cache_for_0,1,2,3.json
(VllmWorkerProcess pid=145) INFO 04-29 06:18:27 [custom_all_reduce_utils.py:244] reading GPU P2P access cache from /root/.cache/vllm/gpu_p2p_access_cache_for_0,1,2,3.json
INFO 04-29 06:18:27 [shm_broadcast.py:266] vLLM message queue communication handle: Handle(local_reader_ranks=[1, 2, 3], buffer_handle=(3, 4194304, 6, 'psm_9d1fab20'), local_subscribe_addr='ipc:///tmp/6e19770e-3936-4b80-bc6f-8ed440791eef', remote_subscribe_addr=None, remote_addr_ipv6=False)
INFO 04-29 06:18:27 [parallel_state.py:1004] rank 0 in world size 4 is assigned as DP rank 0, PP rank 0, TP rank 0
(VllmWorkerProcess pid=146) INFO 04-29 06:18:27 [parallel_state.py:1004] rank 2 in world size 4 is assigned as DP rank 0, PP rank 0, TP rank 2
(VllmWorkerProcess pid=147) INFO 04-29 06:18:27 [parallel_state.py:1004] rank 3 in world size 4 is assigned as DP rank 0, PP rank 0, TP rank 3
(VllmWorkerProcess pid=145) INFO 04-29 06:18:27 [parallel_state.py:1004] rank 1 in world size 4 is assigned as DP rank 0, PP rank 0, TP rank 1
INFO 04-29 06:18:27 [model_runner.py:1108] Starting to load model Qwen/Qwen3-30B-A3B...
(VllmWorkerProcess pid=146) INFO 04-29 06:18:27 [model_runner.py:1108] Starting to load model Qwen/Qwen3-30B-A3B...
(VllmWorkerProcess pid=145) INFO 04-29 06:18:27 [model_runner.py:1108] Starting to load model Qwen/Qwen3-30B-A3B...
(VllmWorkerProcess pid=147) INFO 04-29 06:18:27 [model_runner.py:1108] Starting to load model Qwen/Qwen3-30B-A3B...
INFO 04-29 06:18:28 [weight_utils.py:265] Using model weights format ['*.safetensors']
(VllmWorkerProcess pid=146) INFO 04-29 06:18:28 [weight_utils.py:265] Using model weights format ['*.safetensors']
(VllmWorkerProcess pid=145) INFO 04-29 06:18:28 [weight_utils.py:265] Using model weights format ['*.safetensors']
(VllmWorkerProcess pid=147) INFO 04-29 06:18:28 [weight_utils.py:265] Using model weights format ['*.safetensors']
Loading safetensors checkpoint shards:   0% Completed | 0/16 [00:00<?, ?it/s]
Loading safetensors checkpoint shards:   6% Completed | 1/16 [00:01<00:17,  1.16s/it]
Loading safetensors checkpoint shards:  12% Completed | 2/16 [00:02<00:16,  1.17s/it]
Loading safetensors checkpoint shards:  19% Completed | 3/16 [00:03<00:15,  1.17s/it]
Loading safetensors checkpoint shards:  25% Completed | 4/16 [00:04<00:14,  1.17s/it]
Loading safetensors checkpoint shards:  31% Completed | 5/16 [00:05<00:12,  1.17s/it]
Loading safetensors checkpoint shards:  38% Completed | 6/16 [00:07<00:11,  1.17s/it]
Loading safetensors checkpoint shards:  44% Completed | 7/16 [00:08<00:10,  1.17s/it]
Loading safetensors checkpoint shards:  50% Completed | 8/16 [00:09<00:09,  1.17s/it]
Loading safetensors checkpoint shards:  56% Completed | 9/16 [00:10<00:08,  1.17s/it]
Loading safetensors checkpoint shards:  62% Completed | 10/16 [00:11<00:07,  1.17s/it]
Loading safetensors checkpoint shards:  69% Completed | 11/16 [00:12<00:04,  1.09it/s]
Loading safetensors checkpoint shards:  75% Completed | 12/16 [00:13<00:03,  1.01it/s]
Loading safetensors checkpoint shards:  81% Completed | 13/16 [00:14<00:03,  1.04s/it]
Loading safetensors checkpoint shards:  88% Completed | 14/16 [00:15<00:02,  1.08s/it]
Loading safetensors checkpoint shards:  94% Completed | 15/16 [00:16<00:01,  1.11s/it]
Loading safetensors checkpoint shards: 100% Completed | 16/16 [00:17<00:00,  1.13s/it]
Loading safetensors checkpoint shards: 100% Completed | 16/16 [00:17<00:00,  1.12s/it]

INFO 04-29 06:18:46 [loader.py:458] Loading weights took 17.93 seconds
INFO 04-29 06:18:46 [model_runner.py:1140] Model loading took 14.2474 GiB and 19.192954 seconds
(VllmWorkerProcess pid=146) INFO 04-29 06:18:47 [loader.py:458] Loading weights took 18.12 seconds
(VllmWorkerProcess pid=146) INFO 04-29 06:18:47 [model_runner.py:1140] Model loading took 14.2474 GiB and 19.835734 seconds
(VllmWorkerProcess pid=145) INFO 04-29 06:18:47 [loader.py:458] Loading weights took 18.24 seconds
(VllmWorkerProcess pid=147) INFO 04-29 06:18:48 [loader.py:458] Loading weights took 17.95 seconds
(VllmWorkerProcess pid=145) INFO 04-29 06:18:48 [model_runner.py:1140] Model loading took 14.2474 GiB and 20.417415 seconds
(VllmWorkerProcess pid=147) INFO 04-29 06:18:48 [model_runner.py:1140] Model loading took 14.2474 GiB and 20.602702 seconds
WARNING 04-29 06:18:52 [fused_moe.py:668] Using default MoE config. Performance might be sub-optimal! Config file not found at /usr/local/lib/python3.12/dist-packages/vllm/model_executor/layers/fused_moe/configs/E=128,N=192,device_name=Tesla_V100-DGXS-32GB.json
(VllmWorkerProcess pid=147) WARNING 04-29 06:18:52 [fused_moe.py:668] Using default MoE config. Performance might be sub-optimal! Config file not found at /usr/local/lib/python3.12/dist-packages/vllm/model_executor/layers/fused_moe/configs/E=128,N=192,device_name=Tesla_V100-DGXS-32GB.json
(VllmWorkerProcess pid=145) WARNING 04-29 06:18:52 [fused_moe.py:668] Using default MoE config. Performance might be sub-optimal! Config file not found at /usr/local/lib/python3.12/dist-packages/vllm/model_executor/layers/fused_moe/configs/E=128,N=192,device_name=Tesla_V100-DGXS-32GB.json
(VllmWorkerProcess pid=146) WARNING 04-29 06:18:52 [fused_moe.py:668] Using default MoE config. Performance might be sub-optimal! Config file not found at /usr/local/lib/python3.12/dist-packages/vllm/model_executor/layers/fused_moe/configs/E=128,N=192,device_name=Tesla_V100-DGXS-32GB.json
LLVM ERROR: Failed to compute parent layout for slice layout.
LLVM ERROR: Failed to compute parent layout for slice layout.
LLVM ERROR: Failed to compute parent layout for slice layout.
LLVM ERROR: Failed to compute parent layout for slice layout.
Traceback (most recent call last):
  File "<frozen runpy>", line 198, in _run_module_as_main
  File "<frozen runpy>", line 88, in _run_code
  File "/usr/local/lib/python3.12/dist-packages/vllm/entrypoints/openai/api_server.py", line 1130, in <module>
    uvloop.run(run_server(args))
  File "/usr/local/lib/python3.12/dist-packages/uvloop/__init__.py", line 109, in run
    return __asyncio.run(
           ^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/asyncio/runners.py", line 195, in run
    return runner.run(main)
           ^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/asyncio/runners.py", line 118, in run
    return self._loop.run_until_complete(task)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "uvloop/loop.pyx", line 1518, in uvloop.loop.Loop.run_until_complete
  File "/usr/local/lib/python3.12/dist-packages/uvloop/__init__.py", line 61, in wrapper
    return await main
           ^^^^^^^^^^
  File "/usr/local/lib/python3.12/dist-packages/vllm/entrypoints/openai/api_server.py", line 1078, in run_server
    async with build_async_engine_client(args) as engine_client:
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/contextlib.py", line 210, in __aenter__
    return await anext(self.gen)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/dist-packages/vllm/entrypoints/openai/api_server.py", line 146, in build_async_engine_client
    async with build_async_engine_client_from_engine_args(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/contextlib.py", line 210, in __aenter__
    return await anext(self.gen)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/dist-packages/vllm/entrypoints/openai/api_server.py", line 269, in build_async_engine_client_from_engine_args
    raise RuntimeError(
RuntimeError: Engine process failed to start. See stack trace for the root cause.
/usr/lib/python3.12/multiprocessing/resource_tracker.py:279: UserWarning: resource_tracker: There appear to be 16 leaked semaphore objects to clean up at shutdown
  warnings.warn('resource_tracker: There appear to be %d '
/usr/lib/python3.12/multiprocessing/resource_tracker.py:279: UserWarning: resource_tracker: There appear to be 1 leaked shared_memory objects to clean up at shutdown
  warnings.warn('resource_tracker: There appear to be %d '

Before submitting a new issue…

  • Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

About this issue

Most upvoted comments

I used the same arguments for Qwen/Qwen3-14B. It works well. So I think it’s the MoE problem.

Yes, all MOE models have issues, including Qwen3-30B-A3B and Qwen3-235B-A22B,because previously only DeepSeek-R1-671B was an MOE model, it was too large to test, now that there are smaller size 30B-A3B and 235B-A22B models, my V100 has the conditions to run them, but they have been failing consistently. Currently, I know that MOE+GPTQ is natively unsupported, MOE should be supported, but it seems it hasn’t been properly adapted with V100 yet.

同样的问题

vllm serve Qwen3-30B-A3B --port 8000 -tp 4 --dtype half --max-model-len 40960 --no-enable-chunked-prefill
WARNING 04-30 15:13:59 [fused_moe.py:668] Using default MoE config. Performance might be sub-optimal! Config file not found at /disk6/miniconda3/envs/vllm/lib/python3.11/site-packages/vllm/model_executor/layers/fused_moe/configs/E=128,N=192,device_name=Tesla_V100-SXM2-32GB.json

Same issue for Qwen/Qwen3-8B with Tesla V100 deployed on docker

command --model Qwen/Qwen3-8B --enable-reasoning --reasoning-parser deepseek_r1 --dtype half --uvicorn-log-level debug --no-enable-chunked-prefill --enable-prefix-caching --gpu-memory-utilization 0.85 --enforce-eager --max_model_len 10000

error log INFO 05-08 02:15:22 [engine.py:310] Added request chatcmpl-fd9f6952cc0e4cee8b49014b4f8de7f0. LLVM ERROR: Failed to compute parent layout for slice layout. ERROR 05-08 02:15:29 [client.py:305] RuntimeError('Engine process (pid 50) died.') ERROR 05-08 02:15:29 [client.py:305] NoneType: None INFO: Shutting down INFO: Waiting for connections to close. (CTRL+C to force quit) ERROR 05-08 02:15:39 [serving_chat.py:885] Error in chat completion stream generator. ERROR 05-08 02:15:39 [serving_chat.py:885] Traceback (most recent call last): ERROR 05-08 02:15:39 [serving_chat.py:885] File "/usr/local/lib/python3.12/dist-packages/vllm/entrypoints/openai/serving_chat.py", line 487, in chat_completion_stream_generator ERROR 05-08 02:15:39 [serving_chat.py:885] async for res in result_generator: ERROR 05-08 02:15:39 [serving_chat.py:885] File "/usr/local/lib/python3.12/dist-packages/vllm/engine/multiprocessing/client.py", line 667, in _process_request ERROR 05-08 02:15:39 [serving_chat.py:885] raise request_output ERROR 05-08 02:15:39 [serving_chat.py:885] vllm.engine.multiprocessing.MQEngineDeadError: Engine loop is not running. Inspect the stacktrace to find the original error: RuntimeError('Engine process (pid 50) died.').

same issue with Tesla V100

same issue with Tesla V100s,