vllm: [Bug]: ValueError: The output_size of gate's and up's weight = 192 is not divisible by weight quantization block_n = 128.
Your current environment
The output of python collect_env.py
INFO 05-02 08:12:42 [__init__.py:239] Automatically detected platform cuda.
Collecting environment information...
PyTorch version: 2.7.0+cu128
Is debug build: False
CUDA used to build PyTorch: 12.8
ROCM used to build PyTorch: N/A
OS: Ubuntu 22.04.5 LTS (x86_64)
GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
Clang version: 14.0.0-1ubuntu1.1
CMake version: version 3.22.1
Libc version: glibc-2.35
Python version: 3.12.9 | packaged by Anaconda, Inc. | (main, Feb 6 2025, 18:56:27) [GCC 11.2.0] (64-bit runtime)
Python platform: Linux-6.8.0-58-generic-x86_64-with-glibc2.35
Is CUDA available: True
CUDA runtime version: 12.8.93
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration:
GPU 0: NVIDIA GeForce RTX 5090 D
GPU 1: NVIDIA GeForce RTX 5090 D
GPU 2: NVIDIA GeForce RTX 5090 D
GPU 3: NVIDIA GeForce RTX 5090 D
GPU 4: NVIDIA GeForce RTX 5090 D
GPU 5: NVIDIA GeForce RTX 5090 D
GPU 6: NVIDIA GeForce RTX 5090 D
GPU 7: NVIDIA GeForce RTX 5090 D
Nvidia driver version: 570.133.07
cuDNN version: Could not collect
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True
CPU:
架构: x86_64
CPU 运行模式: 32-bit, 64-bit
Address sizes: 52 bits physical, 57 bits virtual
字节序: Little Endian
CPU: 64
在线 CPU 列表: 0-63
厂商 ID: AuthenticAMD
型号名称: AMD EPYC 9124 16-Core Processor
CPU 系列: 25
型号: 17
每个核的线程数: 2
每个座的核数: 16
座: 2
步进: 1
Frequency boost: enabled
CPU 最大 MHz: 3711.9141
CPU 最小 MHz: 1500.0000
BogoMIPS: 5999.61
标记: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good amd_lbr_v2 nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba perfmon_v2 ibrs ibpb stibp ibrs_enhanced vmmcall fsgsbase bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local user_shstk avx512_bf16 clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin cppc amd_ibpb_ret arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif x2avic v_spec_ctrl vnmi avx512vbmi umip pku ospke avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq la57 rdpid overflow_recov succor smca fsrm flush_l1d debug_swap
虚拟化: AMD-V
L1d 缓存: 1 MiB (32 instances)
L1i 缓存: 1 MiB (32 instances)
L2 缓存: 32 MiB (32 instances)
L3 缓存: 128 MiB (8 instances)
NUMA 节点: 2
NUMA 节点0 CPU: 0-15,32-47
NUMA 节点1 CPU: 16-31,48-63
Vulnerability Gather data sampling: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Mitigation; Safe RET
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; Enhanced / Automatic IBRS; IBPB conditional; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
Versions of relevant libraries:
[pip3] numpy==2.2.5
[pip3] nvidia-cublas-cu12==12.8.3.14
[pip3] nvidia-cuda-cupti-cu12==12.8.57
[pip3] nvidia-cuda-nvrtc-cu12==12.8.61
[pip3] nvidia-cuda-runtime-cu12==12.8.57
[pip3] nvidia-cudnn-cu12==9.7.1.26
[pip3] nvidia-cufft-cu12==11.3.3.41
[pip3] nvidia-cufile-cu12==1.13.0.11
[pip3] nvidia-curand-cu12==10.3.9.55
[pip3] nvidia-cusolver-cu12==11.7.2.55
[pip3] nvidia-cusparse-cu12==12.5.7.53
[pip3] nvidia-cusparselt-cu12==0.6.3
[pip3] nvidia-nccl-cu12==2.26.2
[pip3] nvidia-nvjitlink-cu12==12.8.61
[pip3] nvidia-nvtx-cu12==12.8.55
[pip3] pyzmq==26.4.0
[pip3] torch==2.7.0+cu128
[pip3] torchaudio==2.7.0+cu128
[pip3] torchvision==0.22.0+cu128
[pip3] transformers==4.51.3
[pip3] triton==3.3.0
[conda] numpy 2.2.5 pypi_0 pypi
[conda] nvidia-cublas-cu12 12.8.3.14 pypi_0 pypi
[conda] nvidia-cuda-cupti-cu12 12.8.57 pypi_0 pypi
[conda] nvidia-cuda-nvrtc-cu12 12.8.61 pypi_0 pypi
[conda] nvidia-cuda-runtime-cu12 12.8.57 pypi_0 pypi
[conda] nvidia-cudnn-cu12 9.7.1.26 pypi_0 pypi
[conda] nvidia-cufft-cu12 11.3.3.41 pypi_0 pypi
[conda] nvidia-cufile-cu12 1.13.0.11 pypi_0 pypi
[conda] nvidia-curand-cu12 10.3.9.55 pypi_0 pypi
[conda] nvidia-cusolver-cu12 11.7.2.55 pypi_0 pypi
[conda] nvidia-cusparse-cu12 12.5.7.53 pypi_0 pypi
[conda] nvidia-cusparselt-cu12 0.6.3 pypi_0 pypi
[conda] nvidia-nccl-cu12 2.26.2 pypi_0 pypi
[conda] nvidia-nvjitlink-cu12 12.8.61 pypi_0 pypi
[conda] nvidia-nvtx-cu12 12.8.55 pypi_0 pypi
[conda] pyzmq 26.4.0 pypi_0 pypi
[conda] torch 2.7.0+cu128 pypi_0 pypi
[conda] torchaudio 2.7.0+cu128 pypi_0 pypi
[conda] torchvision 0.22.0+cu128 pypi_0 pypi
[conda] transformers 4.51.3 pypi_0 pypi
[conda] triton 3.3.0 pypi_0 pypi
ROCM Version: Could not collect
Neuron SDK Version: N/A
vLLM Version: 0.8.5.dev407+g7423cf0a9 (git sha: 7423cf0a9)
vLLM Build Flags:
CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled
GPU Topology:
GPU0 GPU1 GPU2 GPU3 GPU4 GPU5 GPU6 GPU7 CPU Affinity NUMA Affinity GPU NUMA ID
GPU0 X NODE NODE NODE SYS SYS SYS SYS 0-15,32-47 0 N/A
GPU1 NODE X NODE NODE SYS SYS SYS SYS 0-15,32-47 0 N/A
GPU2 NODE NODE X NODE SYS SYS SYS SYS 0-15,32-47 0 N/A
GPU3 NODE NODE NODE X SYS SYS SYS SYS 0-15,32-47 0 N/A
GPU4 SYS SYS SYS SYS X NODE NODE NODE 16-31,48-63 1 N/A
GPU5 SYS SYS SYS SYS NODE X NODE NODE 16-31,48-63 1 N/A
GPU6 SYS SYS SYS SYS NODE NODE X NODE 16-31,48-63 1 N/A
GPU7 SYS SYS SYS SYS NODE NODE NODE X 16-31,48-63 1 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
LD_LIBRARY_PATH=/usr/local/cuda/lib64:
NCCL_CUMEM_ENABLE=0
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
CUDA_MODULE_LOADING=LAZY
🐛 Describe the bug
(VllmWorker rank=1 pid=1410768) WARNING 05-02 07:54:19 [custom_all_reduce.py:136] Custom allreduce is disabled because it's not supported on more than two PCIe-only GPUs. To silence this warning, specify disable_custom_all_reduce=True explicitly.
(VllmWorker rank=6 pid=1410773) WARNING 05-02 07:54:19 [custom_all_reduce.py:136] Custom allreduce is disabled because it's not supported on more than two PCIe-only GPUs. To silence this warning, specify disable_custom_all_reduce=True explicitly.
(VllmWorker rank=4 pid=1410771) WARNING 05-02 07:54:19 [custom_all_reduce.py:136] Custom allreduce is disabled because it's not supported on more than two PCIe-only GPUs. To silence this warning, specify disable_custom_all_reduce=True explicitly.
(VllmWorker rank=7 pid=1410774) WARNING 05-02 07:54:19 [custom_all_reduce.py:136] Custom allreduce is disabled because it's not supported on more than two PCIe-only GPUs. To silence this warning, specify disable_custom_all_reduce=True explicitly.
(VllmWorker rank=5 pid=1410772) WARNING 05-02 07:54:19 [custom_all_reduce.py:136] Custom allreduce is disabled because it's not supported on more than two PCIe-only GPUs. To silence this warning, specify disable_custom_all_reduce=True explicitly.
(VllmWorker rank=3 pid=1410770) WARNING 05-02 07:54:19 [custom_all_reduce.py:136] Custom allreduce is disabled because it's not supported on more than two PCIe-only GPUs. To silence this warning, specify disable_custom_all_reduce=True explicitly.
(VllmWorker rank=0 pid=1410767) WARNING 05-02 07:54:19 [custom_all_reduce.py:136] Custom allreduce is disabled because it's not supported on more than two PCIe-only GPUs. To silence this warning, specify disable_custom_all_reduce=True explicitly.
(VllmWorker rank=2 pid=1410769) WARNING 05-02 07:54:19 [custom_all_reduce.py:136] Custom allreduce is disabled because it's not supported on more than two PCIe-only GPUs. To silence this warning, specify disable_custom_all_reduce=True explicitly.
(VllmWorker rank=0 pid=1410767) INFO 05-02 07:54:19 [shm_broadcast.py:266] vLLM message queue communication handle: Handle(local_reader_ranks=[1, 2, 3, 4, 5, 6, 7], buffer_handle=(7, 4194304, 6, 'psm_938d0ae5'), local_subscribe_addr='ipc:///tmp/3733e0cb-f077-47fe-baea-66a5622fce09', remote_subscribe_addr=None, remote_addr_ipv6=False)
(VllmWorker rank=6 pid=1410773) INFO 05-02 07:54:19 [parallel_state.py:1004] rank 6 in world size 8 is assigned as DP rank 0, PP rank 0, TP rank 6
(VllmWorker rank=7 pid=1410774) INFO 05-02 07:54:19 [parallel_state.py:1004] rank 7 in world size 8 is assigned as DP rank 0, PP rank 0, TP rank 7
(VllmWorker rank=5 pid=1410772) INFO 05-02 07:54:19 [parallel_state.py:1004] rank 5 in world size 8 is assigned as DP rank 0, PP rank 0, TP rank 5
(VllmWorker rank=4 pid=1410771) INFO 05-02 07:54:19 [parallel_state.py:1004] rank 4 in world size 8 is assigned as DP rank 0, PP rank 0, TP rank 4
(VllmWorker rank=1 pid=1410768) INFO 05-02 07:54:19 [parallel_state.py:1004] rank 1 in world size 8 is assigned as DP rank 0, PP rank 0, TP rank 1
(VllmWorker rank=3 pid=1410770) INFO 05-02 07:54:19 [parallel_state.py:1004] rank 3 in world size 8 is assigned as DP rank 0, PP rank 0, TP rank 3
(VllmWorker rank=2 pid=1410769) INFO 05-02 07:54:19 [parallel_state.py:1004] rank 2 in world size 8 is assigned as DP rank 0, PP rank 0, TP rank 2
(VllmWorker rank=0 pid=1410767) INFO 05-02 07:54:19 [parallel_state.py:1004] rank 0 in world size 8 is assigned as DP rank 0, PP rank 0, TP rank 0
(VllmWorker rank=6 pid=1410773) INFO 05-02 07:54:19 [cuda.py:221] Using Flash Attention backend on V1 engine.
(VllmWorker rank=1 pid=1410768) INFO 05-02 07:54:19 [cuda.py:221] Using Flash Attention backend on V1 engine.
(VllmWorker rank=4 pid=1410771) INFO 05-02 07:54:19 [cuda.py:221] Using Flash Attention backend on V1 engine.
(VllmWorker rank=3 pid=1410770) INFO 05-02 07:54:19 [cuda.py:221] Using Flash Attention backend on V1 engine.
(VllmWorker rank=7 pid=1410774) INFO 05-02 07:54:19 [cuda.py:221] Using Flash Attention backend on V1 engine.
(VllmWorker rank=5 pid=1410772) INFO 05-02 07:54:19 [cuda.py:221] Using Flash Attention backend on V1 engine.
(VllmWorker rank=6 pid=1410773) WARNING 05-02 07:54:19 [topk_topp_sampler.py:69] FlashInfer is not available. Falling back to the PyTorch-native implementation of top-p & top-k sampling. For the best performance, please install FlashInfer.
(VllmWorker rank=1 pid=1410768) WARNING 05-02 07:54:19 [topk_topp_sampler.py:69] FlashInfer is not available. Falling back to the PyTorch-native implementation of top-p & top-k sampling. For the best performance, please install FlashInfer.
(VllmWorker rank=4 pid=1410771) WARNING 05-02 07:54:19 [topk_topp_sampler.py:69] FlashInfer is not available. Falling back to the PyTorch-native implementation of top-p & top-k sampling. For the best performance, please install FlashInfer.
(VllmWorker rank=3 pid=1410770) WARNING 05-02 07:54:19 [topk_topp_sampler.py:69] FlashInfer is not available. Falling back to the PyTorch-native implementation of top-p & top-k sampling. For the best performance, please install FlashInfer.
(VllmWorker rank=7 pid=1410774) WARNING 05-02 07:54:19 [topk_topp_sampler.py:69] FlashInfer is not available. Falling back to the PyTorch-native implementation of top-p & top-k sampling. For the best performance, please install FlashInfer.
(VllmWorker rank=5 pid=1410772) WARNING 05-02 07:54:19 [topk_topp_sampler.py:69] FlashInfer is not available. Falling back to the PyTorch-native implementation of top-p & top-k sampling. For the best performance, please install FlashInfer.
(VllmWorker rank=0 pid=1410767) INFO 05-02 07:54:19 [cuda.py:221] Using Flash Attention backend on V1 engine.
(VllmWorker rank=2 pid=1410769) INFO 05-02 07:54:19 [cuda.py:221] Using Flash Attention backend on V1 engine.
(VllmWorker rank=0 pid=1410767) WARNING 05-02 07:54:19 [topk_topp_sampler.py:69] FlashInfer is not available. Falling back to the PyTorch-native implementation of top-p & top-k sampling. For the best performance, please install FlashInfer.
(VllmWorker rank=2 pid=1410769) WARNING 05-02 07:54:19 [topk_topp_sampler.py:69] FlashInfer is not available. Falling back to the PyTorch-native implementation of top-p & top-k sampling. For the best performance, please install FlashInfer.
(VllmWorker rank=7 pid=1410774) INFO 05-02 07:54:19 [gpu_model_runner.py:1335] Starting to load model /home/user/Qwen3-235B-A22B-FP8/...
(VllmWorker rank=3 pid=1410770) INFO 05-02 07:54:19 [gpu_model_runner.py:1335] Starting to load model /home/user/Qwen3-235B-A22B-FP8/...
(VllmWorker rank=6 pid=1410773) INFO 05-02 07:54:19 [gpu_model_runner.py:1335] Starting to load model /home/user/Qwen3-235B-A22B-FP8/...
(VllmWorker rank=2 pid=1410769) INFO 05-02 07:54:19 [gpu_model_runner.py:1335] Starting to load model /home/user/Qwen3-235B-A22B-FP8/...
(VllmWorker rank=6 pid=1410773) WARNING 05-02 07:54:19 [utils.py:168] The model class Qwen3MoeForCausalLM has not defined packed_modules_mapping, this may lead to incorrect mapping of quantized or ignored modules
(VllmWorker rank=3 pid=1410770) WARNING 05-02 07:54:19 [utils.py:168] The model class Qwen3MoeForCausalLM has not defined packed_modules_mapping, this may lead to incorrect mapping of quantized or ignored modules
(VllmWorker rank=7 pid=1410774) WARNING 05-02 07:54:19 [utils.py:168] The model class Qwen3MoeForCausalLM has not defined packed_modules_mapping, this may lead to incorrect mapping of quantized or ignored modules
(VllmWorker rank=5 pid=1410772) INFO 05-02 07:54:19 [gpu_model_runner.py:1335] Starting to load model /home/user/Qwen3-235B-A22B-FP8/...
(VllmWorker rank=4 pid=1410771) INFO 05-02 07:54:19 [gpu_model_runner.py:1335] Starting to load model /home/user/Qwen3-235B-A22B-FP8/...
(VllmWorker rank=1 pid=1410768) INFO 05-02 07:54:19 [gpu_model_runner.py:1335] Starting to load model /home/user/Qwen3-235B-A22B-FP8/...
(VllmWorker rank=0 pid=1410767) INFO 05-02 07:54:19 [gpu_model_runner.py:1335] Starting to load model /home/user/Qwen3-235B-A22B-FP8/...
(VllmWorker rank=0 pid=1410767) WARNING 05-02 07:54:19 [utils.py:168] The model class Qwen3MoeForCausalLM has not defined packed_modules_mapping, this may lead to incorrect mapping of quantized or ignored modules
(VllmWorker rank=2 pid=1410769) WARNING 05-02 07:54:19 [utils.py:168] The model class Qwen3MoeForCausalLM has not defined packed_modules_mapping, this may lead to incorrect mapping of quantized or ignored modules
(VllmWorker rank=5 pid=1410772) WARNING 05-02 07:54:19 [utils.py:168] The model class Qwen3MoeForCausalLM has not defined packed_modules_mapping, this may lead to incorrect mapping of quantized or ignored modules
(VllmWorker rank=4 pid=1410771) WARNING 05-02 07:54:19 [utils.py:168] The model class Qwen3MoeForCausalLM has not defined packed_modules_mapping, this may lead to incorrect mapping of quantized or ignored modules
(VllmWorker rank=1 pid=1410768) WARNING 05-02 07:54:19 [utils.py:168] The model class Qwen3MoeForCausalLM has not defined packed_modules_mapping, this may lead to incorrect mapping of quantized or ignored modules
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] WorkerProc failed to start.
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] Traceback (most recent call last):
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/v1/executor/multiproc_executor.py", line 409, in worker_main
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] worker = WorkerProc(*args, **kwargs)
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] ^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/v1/executor/multiproc_executor.py", line 306, in __init__
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] self.worker.load_model()
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/v1/worker/gpu_worker.py", line 162, in load_model
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] self.model_runner.load_model()
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/v1/worker/gpu_model_runner.py", line 1338, in load_model
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] self.model = get_model(vllm_config=self.vllm_config)
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/model_executor/model_loader/__init__.py", line 14, in get_model
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] return loader.load_model(vllm_config=vllm_config)
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/model_executor/model_loader/loader.py", line 452, in load_model
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] model = _initialize_model(vllm_config=vllm_config)
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/model_executor/model_loader/loader.py", line 133, in _initialize_model
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] return model_class(vllm_config=vllm_config, prefix=prefix)
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/model_executor/models/qwen3_moe.py", line 487, in __init__
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] self.model = Qwen3MoeModel(vllm_config=vllm_config,
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/compilation/decorators.py", line 151, in __init__
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] old_init(self, vllm_config=vllm_config, prefix=prefix, **kwargs)
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/model_executor/models/qwen3_moe.py", line 334, in __init__
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] self.start_layer, self.end_layer, self.layers = make_layers(
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] ^^^^^^^^^^^^
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/model_executor/models/utils.py", line 610, in make_layers
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] maybe_offload_to_cpu(layer_fn(prefix=f"{prefix}.{idx}"))
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/model_executor/models/qwen3_moe.py", line 336, in <lambda>
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] lambda prefix: Qwen3MoeDecoderLayer(config=config,
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/model_executor/models/qwen3_moe.py", line 278, in __init__
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] self.mlp = Qwen3MoeSparseMoeBlock(config=config,
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/model_executor/models/qwen3_moe.py", line 113, in __init__
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] self.experts = FusedMoE(num_experts=config.num_experts,
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/model_executor/layers/fused_moe/layer.py", line 517, in __init__
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] self.quant_method.create_weights(layer=self, **moe_quant_params)
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] File "/home/user/vllm/vllm/model_executor/layers/quantization/fp8.py", line 478, in create_weights
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] raise ValueError(
(VllmWorker rank=6 pid=1410773) ERROR 05-02 07:54:19 [multiproc_executor.py:435] ValueError: The output_size of gate's and up's weight = 192 is not divisible by weight quantization block_n = 128.
[rank0]:[W502 07:54:20.856288655 ProcessGroupNCCL.cpp:1476] Warning: WARNING: destroy_process_group() was not called before program exit, which can leak resources. For more info, please see https://pytorch.org/docs/stable/distributed.html#shutdown (function operator())
ERROR 05-02 07:54:22 [core.py:400] EngineCore failed to start.
ERROR 05-02 07:54:22 [core.py:400] Traceback (most recent call last):
ERROR 05-02 07:54:22 [core.py:400] File "/home/user/vllm/vllm/v1/engine/core.py", line 391, in run_engine_core
ERROR 05-02 07:54:22 [core.py:400] engine_core = EngineCoreProc(*args, **kwargs)
ERROR 05-02 07:54:22 [core.py:400] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 05-02 07:54:22 [core.py:400] File "/home/user/vllm/vllm/v1/engine/core.py", line 333, in __init__
ERROR 05-02 07:54:22 [core.py:400] super().__init__(vllm_config, executor_class, log_stats,
ERROR 05-02 07:54:22 [core.py:400] File "/home/user/vllm/vllm/v1/engine/core.py", line 65, in __init__
ERROR 05-02 07:54:22 [core.py:400] self.model_executor = executor_class(vllm_config)
ERROR 05-02 07:54:22 [core.py:400] ^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 05-02 07:54:22 [core.py:400] File "/home/user/vllm/vllm/executor/executor_base.py", line 52, in __init__
ERROR 05-02 07:54:22 [core.py:400] self._init_executor()
ERROR 05-02 07:54:22 [core.py:400] File "/home/user/vllm/vllm/v1/executor/multiproc_executor.py", line 91, in _init_executor
ERROR 05-02 07:54:22 [core.py:400] self.workers = WorkerProc.wait_for_ready(unready_workers)
ERROR 05-02 07:54:22 [core.py:400] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 05-02 07:54:22 [core.py:400] File "/home/user/vllm/vllm/v1/executor/multiproc_executor.py", line 370, in wait_for_ready
ERROR 05-02 07:54:22 [core.py:400] raise e from None
ERROR 05-02 07:54:22 [core.py:400] Exception: WorkerProc initialization failed due to an exception in a background process. See stack trace for root cause.
Process EngineCore_0:
Traceback (most recent call last):
File "/root/anaconda3/envs/vllm/lib/python3.12/multiprocessing/process.py", line 314, in _bootstrap
self.run()
File "/root/anaconda3/envs/vllm/lib/python3.12/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/user/vllm/vllm/v1/engine/core.py", line 404, in run_engine_core
raise e
File "/home/user/vllm/vllm/v1/engine/core.py", line 391, in run_engine_core
engine_core = EngineCoreProc(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/user/vllm/vllm/v1/engine/core.py", line 333, in __init__
super().__init__(vllm_config, executor_class, log_stats,
File "/home/user/vllm/vllm/v1/engine/core.py", line 65, in __init__
self.model_executor = executor_class(vllm_config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/user/vllm/vllm/executor/executor_base.py", line 52, in __init__
self._init_executor()
File "/home/user/vllm/vllm/v1/executor/multiproc_executor.py", line 91, in _init_executor
self.workers = WorkerProc.wait_for_ready(unready_workers)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/user/vllm/vllm/v1/executor/multiproc_executor.py", line 370, in wait_for_ready
raise e from None
Exception: WorkerProc initialization failed due to an exception in a background process. See stack trace for root cause.
Traceback (most recent call last):
File "/root/anaconda3/envs/vllm/lib/python3.12/weakref.py", line 666, in _exitfunc
f()
File "/root/anaconda3/envs/vllm/lib/python3.12/weakref.py", line 590, in __call__
return info.func(*info.args, **(info.kwargs or {}))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/user/vllm/vllm/v1/executor/multiproc_executor.py", line 228, in shutdown
for w in self.workers:
^^^^^^^^^^^^
AttributeError: 'MultiprocExecutor' object has no attribute 'workers'
Traceback (most recent call last):
File "<frozen runpy>", line 198, in _run_module_as_main
File "<frozen runpy>", line 88, in _run_code
File "/home/user/vllm/vllm/entrypoints/openai/api_server.py", line 1129, in <module>
uvloop.run(run_server(args))
File "/root/anaconda3/envs/vllm/lib/python3.12/site-packages/uvloop/__init__.py", line 109, in run
return __asyncio.run(
^^^^^^^^^^^^^^
File "/root/anaconda3/envs/vllm/lib/python3.12/asyncio/runners.py", line 195, in run
return runner.run(main)
^^^^^^^^^^^^^^^^
File "/root/anaconda3/envs/vllm/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 "/root/anaconda3/envs/vllm/lib/python3.12/site-packages/uvloop/__init__.py", line 61, in wrapper
return await main
^^^^^^^^^^
File "/home/user/vllm/vllm/entrypoints/openai/api_server.py", line 1077, in run_server
async with build_async_engine_client(args) as engine_client:
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/anaconda3/envs/vllm/lib/python3.12/contextlib.py", line 210, in __aenter__
return await anext(self.gen)
^^^^^^^^^^^^^^^^^^^^^
File "/home/user/vllm/vllm/entrypoints/openai/api_server.py", line 146, in build_async_engine_client
async with build_async_engine_client_from_engine_args(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/anaconda3/envs/vllm/lib/python3.12/contextlib.py", line 210, in __aenter__
return await anext(self.gen)
^^^^^^^^^^^^^^^^^^^^^
File "/home/user/vllm/vllm/entrypoints/openai/api_server.py", line 178, in build_async_engine_client_from_engine_args
async_llm = AsyncLLM.from_vllm_config(
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/user/vllm/vllm/v1/engine/async_llm.py", line 151, in from_vllm_config
return cls(
^^^^
File "/home/user/vllm/vllm/v1/engine/async_llm.py", line 118, in __init__
self.engine_core = core_client_class(
^^^^^^^^^^^^^^^^^^
File "/home/user/vllm/vllm/v1/engine/core_client.py", line 649, in __init__
super().__init__(
File "/home/user/vllm/vllm/v1/engine/core_client.py", line 400, in __init__
self._wait_for_engine_startup()
File "/home/user/vllm/vllm/v1/engine/core_client.py", line 432, in _wait_for_engine_startup
raise RuntimeError("Engine core initialization failed. "
RuntimeError: Engine core initialization failed. See root cause above.
/root/anaconda3/envs/vllm/lib/python3.12/multiprocessing/resource_tracker.py:255: 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 '
The command is python -m vllm.entrypoints.openai.api_server --model /home/user/Qwen3-235B-A22B-FP8/ --tensor-parallel-size 8 --gpu-memory-utilization 1 --served-model-name deepmoney VLLM_USE_V1=1 --max-model-len 32768 ,which indicates that the model’s weight sizes are incompatible with the quantization block size configuration. And it seems happen in Qwen3-fp8 model only
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
- Original URL
- State: open
- Created 2 years ago
- Comments: 17 (2 by maintainers)
Qwen3官方文档里提到了这个问题以及解决办法:https://qwen.readthedocs.io/zh-cn/latest/deployment/vllm.html
I got same error , but --enable-expert-parallel works for me . but output token is slow, the log says is only “Avg generation throughput: 7.9 tokens/s” ,