vllm - 💡(How to fix) Fix [Bug]: vllm 0.17.0 启动Qwen3.5-122B-A10B失败 [1 comments, 2 participants]

vllm2026-03-09 08:58:39

ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

GitHub issue URL

Helpful · Quick feedback

GitHub stats

vllm-project/vllm#36476•Fetched 2026-04-08 00:36:44

View on GitHub

Comments

Participants

Timeline

Reactions

Author

Lee-xeo

Participants

Lee-xeo

tonibofarull

Timeline (top)

commented ×1labeled ×1subscribed ×1

Error Message

1. 环境与参数确认

(APIServer pid=1222) INFO 03-09 12:26:19 [utils.py:302] vllm version 0.17.0 (APIServer pid=1222) INFO 03-09 12:26:19 [utils.py:302] model /mnt/user/models/Qwen3.5/Qwen3.5-122B-A10B (APIServer pid=1222) INFO 03-09 12:26:19 [utils.py:238] non-default args: {'model': '...', 'tensor_parallel_size': 2} (APIServer pid=1222) INFO 03-09 12:26:19 [model.py:531] Resolved architecture: Qwen3_5MoeForConditionalGeneration (APIServer pid=1222) INFO 03-09 12:26:19 [scheduler.py:231] Chunked prefill is enabled with max_num_batched_tokens=8192.

2. 分布式初始化 (并行状态)

(Worker pid=1441) INFO 03-09 12:26:42 [parallel_state.py:1393] world_size=2 rank=0 local_rank=0 distributed_init_method=tcp://127.0.0.1:56759 backend=nccl (Worker pid=1441) INFO 03-09 12:26:42 [pynccl.py:111] vLLM is using nccl==2.27.5 (Worker pid=1441) (Worker_TP0 pid=1441) INFO 03-09 12:26:51 [cuda.py:405] Using FLASH_ATTN attention backend.

3. 模型权重加载

Loading safetensors checkpoint shards: 0% Completed | 0/39 [00:00<?, ?it/s] ... Loading safetensors checkpoint shards: 100% Completed | 39/39 [10:01<00:00, 15.41s/it] (Worker pid=1441) (Worker_TP0 pid=1441) INFO 03-09 12:36:54 [gpu_model_runner.py:4338] Model loading took 114.35 GiB memory

4. 关键阶段：CUDA Graph 捕获与编译 (崩溃发生在此处)

(Worker pid=1441) (Worker_TP0 pid=1441) INFO 03-09 12:37:08 [backends.py:350] Cache the graph of compile range (1, 8192) for later use Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 100%|██████████| 51/51 [00:06<00:00, 7.49it/s] Capturing CUDA graphs (decode, FULL): 0%| | 0/51 [00:00<?, ?it/s]

5. 核心错误堆栈 (Worker Rank 1 触发 AssertionError)

(Worker pid=1442) (Worker_TP1 pid=1442) ERROR 03-09 12:37:45 [multiproc_executor.py:880] Traceback (most recent call last): File ".../vllm/v1/worker/gpu_worker.py", line 522, in compile_or_warm_up_model cuda_graph_memory_bytes = self.model_runner.capture_model() File ".../vllm/v1/worker/gpu_model_runner.py", line 5337, in capture_model self._capture_cudagraphs(...) File ".../vllm/v1/worker/gpu_model_runner.py", line 5430, in _capture_cudagraphs dummy_run(...) File ".../vllm/v1/worker/gpu_model_runner.py", line 4976, in _dummy_run outputs = self.model(...) File ".../vllm/model_executor/models/qwen3_5.py", line 738, in forward hidden_states = self.language_model.model(...) File ".../vllm/model_executor/models/qwen3_next.py", line 1132, in forward (AOT compiled) File "<eval_with_key>.121", line 5, in forward gdn_attention_core = torch.ops.vllm.gdn_attention_core(...) File ".../vllm/model_executor/models/qwen3_next.py", line 1451, in gdn_attention_core self._forward_core(...) File ".../vllm/model_executor/models/qwen3_next.py", line 715, in _forward_core mixed_qkv_non_spec = causal_conv1d_update(...) File ".../vllm/model_executor/layers/mamba/ops/causal_conv1d.py", line 1162, in causal_conv1d_update assert num_cache_lines >= batch AssertionError

6. 系统收尾报错

(EngineCore_DP0 pid=1335) ERROR 03-09 12:37:45 [core.py:1100] EngineCore failed to start. (APIServer pid=1222) RuntimeError: Engine core initialization failed. See root cause above.

Root Cause

6. 系统收尾报错

(EngineCore_DP0 pid=1335) ERROR 03-09 12:37:45 [core.py:1100] EngineCore failed to start. (APIServer pid=1222) RuntimeError: Engine core initialization failed. See root cause above.

Fix Action

Fix / Workaround

============================== CPU Info

Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Address sizes: 46 bits physical, 57 bits virtual Byte Order: Little Endian CPU(s): 192 On-line CPU(s) list: 0 Off-line CPU(s) list: 1-191 Vendor ID: GenuineIntel Model name: INTEL(R) XEON(R) PLATINUM 8558 CPU family: 6 Model: 207 Thread(s) per core: 2 Core(s) per socket: 48 Socket(s): 2 Stepping: 2 CPU max MHz: 4000.0000 CPU min MHz: 800.0000 BogoMIPS: 4200.00 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 art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor 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 cat_l2 cdp_l3 invpcid_single cdp_l2 ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local split_lock_detect avx_vnni avx512_bf16 wbnoinvd dtherm ida arat pln pts hwp hwp_act_window hwp_epp hwp_pkg_req avx512vbmi umip pku ospke waitpkg avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq la57 rdpid bus_lock_detect cldemote movdiri movdir64b enqcmd fsrm md_clear serialize tsxldtrk pconfig arch_lbr amx_bf16 avx512_fp16 amx_tile amx_int8 flush_l1d arch_capabilities Virtualization: VT-x L1d cache: 4.5 MiB (96 instances) L1i cache: 3 MiB (96 instances) L2 cache: 192 MiB (96 instances) L3 cache: 520 MiB (2 instances) NUMA node(s): 2 NUMA node0 CPU(s): 0-47,96-143 NUMA node1 CPU(s): 48-95,144-191 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 Retbleed: Not affected Vulnerability Spec rstack overflow: 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; Enhanced IBRS, IBPB conditional, RSB filling, PBRSB-eIBRS SW sequence Vulnerability Srbds: Not affected Vulnerability Tsx async abort: Not affected

Code Example

==============================
        System Info
==============================
OS                           : Ubuntu 22.04 LTS (x86_64)
GCC version                  : (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
Clang version                : Could not collect
CMake version                : version 4.2.1
Libc version                 : glibc-2.35

==============================
       PyTorch Info
==============================
PyTorch version              : 2.10.0+cu128
Is debug build               : False
CUDA used to build PyTorch   : 12.8
ROCM used to build PyTorch   : N/A

==============================
      Python Environment
==============================
Python version               : 3.12.0 | packaged by conda-forge | (main, Oct  3 2023, 08:43:22) [GCC 12.3.0] (64-bit runtime)
Python platform              : Linux-5.15.0-89-generic-x86_64-with-glibc2.35

==============================
       CUDA / GPU Info
==============================
Is CUDA available            : True
CUDA runtime version         : 12.8.93
CUDA_MODULE_LOADING set to   : 
GPU models and configuration : 
GPU 0: NVIDIA H200
GPU 1: NVIDIA H200

Nvidia driver version        : 570.133.20
cuDNN version                : Could not collect
HIP runtime version          : N/A
MIOpen runtime version       : N/A
Is XNNPACK available         : True

==============================
          CPU Info
==============================
Architecture:                       x86_64
CPU op-mode(s):                     32-bit, 64-bit
Address sizes:                      46 bits physical, 57 bits virtual
Byte Order:                         Little Endian
CPU(s):                             192
On-line CPU(s) list:                0
Off-line CPU(s) list:               1-191
Vendor ID:                          GenuineIntel
Model name:                         INTEL(R) XEON(R) PLATINUM 8558
CPU family:                         6
Model:                              207
Thread(s) per core:                 2
Core(s) per socket:                 48
Socket(s):                          2
Stepping:                           2
CPU max MHz:                        4000.0000
CPU min MHz:                        800.0000
BogoMIPS:                           4200.00
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 art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor 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 cat_l2 cdp_l3 invpcid_single cdp_l2 ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local split_lock_detect avx_vnni avx512_bf16 wbnoinvd dtherm ida arat pln pts hwp hwp_act_window hwp_epp hwp_pkg_req avx512vbmi umip pku ospke waitpkg avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq la57 rdpid bus_lock_detect cldemote movdiri movdir64b enqcmd fsrm md_clear serialize tsxldtrk pconfig arch_lbr amx_bf16 avx512_fp16 amx_tile amx_int8 flush_l1d arch_capabilities
Virtualization:                     VT-x
L1d cache:                          4.5 MiB (96 instances)
L1i cache:                          3 MiB (96 instances)
L2 cache:                           192 MiB (96 instances)
L3 cache:                           520 MiB (2 instances)
NUMA node(s):                       2
NUMA node0 CPU(s):                  0-47,96-143
NUMA node1 CPU(s):                  48-95,144-191
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 Retbleed:             Not affected
Vulnerability Spec rstack overflow: 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; Enhanced IBRS, IBPB conditional, RSB filling, PBRSB-eIBRS SW sequence
Vulnerability Srbds:                Not affected
Vulnerability Tsx async abort:      Not affected

==============================
Versions of relevant libraries
==============================
[pip3] flashinfer-python==0.6.4
[pip3] numpy==2.2.0
[pip3] nvidia-cublas-cu12==12.8.4.1
[pip3] nvidia-cuda-cupti-cu12==12.8.90
[pip3] nvidia-cuda-nvrtc-cu12==12.8.93
[pip3] nvidia-cuda-runtime-cu12==12.8.90
[pip3] nvidia-cudnn-cu12==9.10.2.21
[pip3] nvidia-cudnn-frontend==1.16.0
[pip3] nvidia-cufft-cu12==11.3.3.83
[pip3] nvidia-cufile-cu12==1.13.1.3
[pip3] nvidia-curand-cu12==10.3.9.90
[pip3] nvidia-cusolver-cu12==11.7.3.90
[pip3] nvidia-cusparse-cu12==12.5.8.93
[pip3] nvidia-cusparselt-cu12==0.7.1
[pip3] nvidia-cutlass-dsl==4.4.1
[pip3] nvidia-cutlass-dsl-libs-base==4.4.1
[pip3] nvidia-ml-py==13.590.44
[pip3] nvidia-nccl-cu12==2.27.5
[pip3] nvidia-nvjitlink-cu12==12.8.93
[pip3] nvidia-nvshmem-cu12==3.4.5
[pip3] nvidia-nvtx-cu12==12.8.90
[pip3] onnxruntime==1.24.1
[pip3] pyzmq==27.1.0
[pip3] sentence-transformers==5.2.0
[pip3] torch==2.10.0+cu128
[pip3] torch_c_dlpack_ext==0.1.5
[pip3] torchaudio==2.10.0+cu128
[pip3] torchcodec==0.10.0+cu128
[pip3] torchsde==0.2.6
[pip3] torchvision==0.25.0+cu128
[pip3] transformers==4.57.6
[pip3] triton==3.6.0
[pip3] x-transformers==2.16.2
[conda] flashinfer-python                           0.6.4                     pypi_0                 pypi
[conda] libopenvino-pytorch-frontend                2025.2.0                  hecca717_1             http://mirrors.i.h.pjlab.org.cn/repository/conda-tsinghua-cloud/conda-forge
[conda] numpy                                       2.2.0                     pypi_0                 pypi
[conda] nvidia-cublas-cu12                          12.8.4.1                  pypi_0                 pypi
[conda] nvidia-cuda-cupti-cu12                      12.8.90                   pypi_0                 pypi
[conda] nvidia-cuda-nvrtc-cu12                      12.8.93                   pypi_0                 pypi
[conda] nvidia-cuda-runtime-cu12                    12.8.90                   pypi_0                 pypi
[conda] nvidia-cudnn-cu12                           9.10.2.21                 pypi_0                 pypi
[conda] nvidia-cudnn-frontend                       1.16.0                    pypi_0                 pypi
[conda] nvidia-cufft-cu12                           11.3.3.83                 pypi_0                 pypi
[conda] nvidia-cufile-cu12                          1.13.1.3                  pypi_0                 pypi
[conda] nvidia-curand-cu12                          10.3.9.90                 pypi_0                 pypi
[conda] nvidia-cusolver-cu12                        11.7.3.90                 pypi_0                 pypi
[conda] nvidia-cusparse-cu12                        12.5.8.93                 pypi_0                 pypi
[conda] nvidia-cusparselt-cu12                      0.7.1                     pypi_0                 pypi
[conda] nvidia-cutlass-dsl                          4.4.1                     pypi_0                 pypi
[conda] nvidia-cutlass-dsl-libs-base                4.4.1                     pypi_0                 pypi
[conda] nvidia-ml-py                                13.590.44                 pypi_0                 pypi
[conda] nvidia-nccl-cu12                            2.27.5                    pypi_0                 pypi
[conda] nvidia-nvjitlink-cu12                       12.8.93                   pypi_0                 pypi
[conda] nvidia-nvshmem-cu12                         3.4.5                     pypi_0                 pypi
[conda] nvidia-nvtx-cu12                            12.8.90                   pypi_0                 pypi
[conda] pyzmq                                       27.1.0                    pypi_0                 pypi
[conda] sentence-transformers                       5.2.0                     pypi_0                 pypi
[conda] torch                                       2.10.0+cu128              pypi_0                 pypi
[conda] torch-c-dlpack-ext                          0.1.5                     pypi_0                 pypi
[conda] torchaudio                                  2.10.0+cu128              pypi_0                 pypi
[conda] torchcodec                                  0.10.0+cu128              pypi_0                 pypi
[conda] torchsde                                    0.2.6                     pypi_0                 pypi
[conda] torchvision                                 0.25.0+cu128              pypi_0                 pypi
[conda] transformers                                4.57.6                    pypi_0                 pypi
[conda] triton                                      3.6.0                     pypi_0                 pypi
[conda] x-transformers                              2.16.2                    pypi_0                 pypi

==============================
         vLLM Info
==============================
ROCM Version                 : Could not collect
vLLM Version                 : 0.17.0
vLLM Build Flags:
  CUDA Archs: Not Set; ROCm: Disabled
GPU Topology:
        GPU0    GPU1    NIC0    NIC1    NIC2    NIC3    NIC4    NIC5    NIC6    NIC7    NIC8    NIC9    NIC10   NIC11   NIC12   NIC13   NIC14   NIC15NIC16   NIC17   CPU Affinity    NUMA Affinity   GPU NUMA ID
GPU0     X      NV18    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODEPIX      NODE    0       0               N/A
GPU1    NV18     X      NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODENODE     PIX                             N/A
NIC0    PIX     NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODEPIX      NODE
NIC1    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODENODE     NODE
NIC2    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODENODE     NODE
NIC3    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODENODE     NODE
NIC4    NODE    PIX     NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODENODE     PIX
NIC5    NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODENODE     NODE
NIC6    NODE    NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODENODE     NODE
NIC7    NODE    NODE    NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX NODE     NODE
NIC8    PIX     NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODEPIX      NODE
NIC9    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODENODE     NODE
NIC10   NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODENODE     NODE
NIC11   NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODENODE     NODE
NIC12   NODE    PIX     NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODENODE     PIX
NIC13   NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODENODE     NODE
NIC14   NODE    NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODENODE     NODE
NIC15   NODE    NODE    NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X  NODE     NODE
NIC16   PIX     NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE X       NODE
NIC17   NODE    PIX     NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODENODE      X 

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

NIC Legend:

  NIC0: mlx5_0
  NIC1: mlx5_1
  NIC2: mlx5_2
  NIC3: mlx5_3
  NIC4: mlx5_4
  NIC5: mlx5_5
  NIC6: mlx5_6
  NIC7: mlx5_7
  NIC8: mlx5_20
  NIC9: mlx5_21
  NIC10: mlx5_22
  NIC11: mlx5_23
  NIC12: mlx5_24
  NIC13: mlx5_25
  NIC14: mlx5_26
  NIC15: mlx5_27
  NIC16: mlx5_bond_0
  NIC17: mlx5_data_0

==============================
     Environment Variables
==============================
LD_LIBRARY_PATH=/mnt/shared-storage-user/colab-share/cuda-12.8/lib64:/mnt/shared-storage-user/colab-share/cuda-12.8/lib64:/usr/local/nvidia/lib:/usr/local/nvidia/lib64
CUDA_HOME=/mnt/shared-storage-user/colab-share/cuda-12.8
CUDA_HOME=/mnt/shared-storage-user/colab-share/cuda-12.8
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
TORCHINDUCTOR_CACHE_DIR=/tmp

---

vllm serve /mnt/models/Qwen3.5/Qwen3.5-122B-A10B --tensor-parallel-size 2`

---

# 1. 环境与参数确认
(APIServer pid=1222) INFO 03-09 12:26:19 [utils.py:302]  vllm version 0.17.0
(APIServer pid=1222) INFO 03-09 12:26:19 [utils.py:302]  model   /mnt/user/models/Qwen3.5/Qwen3.5-122B-A10B
(APIServer pid=1222) INFO 03-09 12:26:19 [utils.py:238] non-default args: {'model': '...', 'tensor_parallel_size': 2}
(APIServer pid=1222) INFO 03-09 12:26:19 [model.py:531] Resolved architecture: Qwen3_5MoeForConditionalGeneration
(APIServer pid=1222) INFO 03-09 12:26:19 [scheduler.py:231] Chunked prefill is enabled with max_num_batched_tokens=8192.

# 2. 分布式初始化 (并行状态)
(Worker pid=1441) INFO 03-09 12:26:42 [parallel_state.py:1393] world_size=2 rank=0 local_rank=0 distributed_init_method=tcp://127.0.0.1:56759 backend=nccl
(Worker pid=1441) INFO 03-09 12:26:42 [pynccl.py:111] vLLM is using nccl==2.27.5
(Worker pid=1441) (Worker_TP0 pid=1441) INFO 03-09 12:26:51 [cuda.py:405] Using FLASH_ATTN attention backend.

# 3. 模型权重加载
Loading safetensors checkpoint shards:   0% Completed | 0/39 [00:00<?, ?it/s]
...
Loading safetensors checkpoint shards: 100% Completed | 39/39 [10:01<00:00, 15.41s/it]
(Worker pid=1441) (Worker_TP0 pid=1441) INFO 03-09 12:36:54 [gpu_model_runner.py:4338] Model loading took 114.35 GiB memory

# 4. 关键阶段：CUDA Graph 捕获与编译 (崩溃发生在此处)
(Worker pid=1441) (Worker_TP0 pid=1441) INFO 03-09 12:37:08 [backends.py:350] Cache the graph of compile range (1, 8192) for later use
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 100%|██████████| 51/51 [00:06<00:00,  7.49it/s]
Capturing CUDA graphs (decode, FULL):   0%|          | 0/51 [00:00<?, ?it/s]

# 5. 核心错误堆栈 (Worker Rank 1 触发 AssertionError)
(Worker pid=1442) (Worker_TP1 pid=1442) ERROR 03-09 12:37:45 [multiproc_executor.py:880] Traceback (most recent call last):
  File ".../vllm/v1/worker/gpu_worker.py", line 522, in compile_or_warm_up_model
    cuda_graph_memory_bytes = self.model_runner.capture_model()
  File ".../vllm/v1/worker/gpu_model_runner.py", line 5337, in capture_model
    self._capture_cudagraphs(...)
  File ".../vllm/v1/worker/gpu_model_runner.py", line 5430, in _capture_cudagraphs
    dummy_run(...)
  File ".../vllm/v1/worker/gpu_model_runner.py", line 4976, in _dummy_run
    outputs = self.model(...)
  File ".../vllm/model_executor/models/qwen3_5.py", line 738, in forward
    hidden_states = self.language_model.model(...)
  File ".../vllm/model_executor/models/qwen3_next.py", line 1132, in forward (AOT compiled)
  File "<eval_with_key>.121", line 5, in forward
    gdn_attention_core = torch.ops.vllm.gdn_attention_core(...)
  File ".../vllm/model_executor/models/qwen3_next.py", line 1451, in gdn_attention_core
    self._forward_core(...)
  File ".../vllm/model_executor/models/qwen3_next.py", line 715, in _forward_core
    mixed_qkv_non_spec = causal_conv1d_update(...)
  File ".../vllm/model_executor/layers/mamba/ops/causal_conv1d.py", line 1162, in causal_conv1d_update
    assert num_cache_lines >= batch
AssertionError

# 6. 系统收尾报错
(EngineCore_DP0 pid=1335) ERROR 03-09 12:37:45 [core.py:1100] EngineCore failed to start.
(APIServer pid=1222) RuntimeError: Engine core initialization failed. See root cause above.

RAW_BUFFERClick to expand / collapse

Your current environment

<details> <summary>The output of <code>python collect_env.py</code></summary>

==============================
        System Info
==============================
OS                           : Ubuntu 22.04 LTS (x86_64)
GCC version                  : (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
Clang version                : Could not collect
CMake version                : version 4.2.1
Libc version                 : glibc-2.35

==============================
       PyTorch Info
==============================
PyTorch version              : 2.10.0+cu128
Is debug build               : False
CUDA used to build PyTorch   : 12.8
ROCM used to build PyTorch   : N/A

==============================
      Python Environment
==============================
Python version               : 3.12.0 | packaged by conda-forge | (main, Oct  3 2023, 08:43:22) [GCC 12.3.0] (64-bit runtime)
Python platform              : Linux-5.15.0-89-generic-x86_64-with-glibc2.35

==============================
       CUDA / GPU Info
==============================
Is CUDA available            : True
CUDA runtime version         : 12.8.93
CUDA_MODULE_LOADING set to   : 
GPU models and configuration : 
GPU 0: NVIDIA H200
GPU 1: NVIDIA H200

Nvidia driver version        : 570.133.20
cuDNN version                : Could not collect
HIP runtime version          : N/A
MIOpen runtime version       : N/A
Is XNNPACK available         : True

==============================
          CPU Info
==============================
Architecture:                       x86_64
CPU op-mode(s):                     32-bit, 64-bit
Address sizes:                      46 bits physical, 57 bits virtual
Byte Order:                         Little Endian
CPU(s):                             192
On-line CPU(s) list:                0
Off-line CPU(s) list:               1-191
Vendor ID:                          GenuineIntel
Model name:                         INTEL(R) XEON(R) PLATINUM 8558
CPU family:                         6
Model:                              207
Thread(s) per core:                 2
Core(s) per socket:                 48
Socket(s):                          2
Stepping:                           2
CPU max MHz:                        4000.0000
CPU min MHz:                        800.0000
BogoMIPS:                           4200.00
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 art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor 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 cat_l2 cdp_l3 invpcid_single cdp_l2 ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local split_lock_detect avx_vnni avx512_bf16 wbnoinvd dtherm ida arat pln pts hwp hwp_act_window hwp_epp hwp_pkg_req avx512vbmi umip pku ospke waitpkg avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq la57 rdpid bus_lock_detect cldemote movdiri movdir64b enqcmd fsrm md_clear serialize tsxldtrk pconfig arch_lbr amx_bf16 avx512_fp16 amx_tile amx_int8 flush_l1d arch_capabilities
Virtualization:                     VT-x
L1d cache:                          4.5 MiB (96 instances)
L1i cache:                          3 MiB (96 instances)
L2 cache:                           192 MiB (96 instances)
L3 cache:                           520 MiB (2 instances)
NUMA node(s):                       2
NUMA node0 CPU(s):                  0-47,96-143
NUMA node1 CPU(s):                  48-95,144-191
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 Retbleed:             Not affected
Vulnerability Spec rstack overflow: 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; Enhanced IBRS, IBPB conditional, RSB filling, PBRSB-eIBRS SW sequence
Vulnerability Srbds:                Not affected
Vulnerability Tsx async abort:      Not affected

==============================
Versions of relevant libraries
==============================
[pip3] flashinfer-python==0.6.4
[pip3] numpy==2.2.0
[pip3] nvidia-cublas-cu12==12.8.4.1
[pip3] nvidia-cuda-cupti-cu12==12.8.90
[pip3] nvidia-cuda-nvrtc-cu12==12.8.93
[pip3] nvidia-cuda-runtime-cu12==12.8.90
[pip3] nvidia-cudnn-cu12==9.10.2.21
[pip3] nvidia-cudnn-frontend==1.16.0
[pip3] nvidia-cufft-cu12==11.3.3.83
[pip3] nvidia-cufile-cu12==1.13.1.3
[pip3] nvidia-curand-cu12==10.3.9.90
[pip3] nvidia-cusolver-cu12==11.7.3.90
[pip3] nvidia-cusparse-cu12==12.5.8.93
[pip3] nvidia-cusparselt-cu12==0.7.1
[pip3] nvidia-cutlass-dsl==4.4.1
[pip3] nvidia-cutlass-dsl-libs-base==4.4.1
[pip3] nvidia-ml-py==13.590.44
[pip3] nvidia-nccl-cu12==2.27.5
[pip3] nvidia-nvjitlink-cu12==12.8.93
[pip3] nvidia-nvshmem-cu12==3.4.5
[pip3] nvidia-nvtx-cu12==12.8.90
[pip3] onnxruntime==1.24.1
[pip3] pyzmq==27.1.0
[pip3] sentence-transformers==5.2.0
[pip3] torch==2.10.0+cu128
[pip3] torch_c_dlpack_ext==0.1.5
[pip3] torchaudio==2.10.0+cu128
[pip3] torchcodec==0.10.0+cu128
[pip3] torchsde==0.2.6
[pip3] torchvision==0.25.0+cu128
[pip3] transformers==4.57.6
[pip3] triton==3.6.0
[pip3] x-transformers==2.16.2
[conda] flashinfer-python                           0.6.4                     pypi_0                 pypi
[conda] libopenvino-pytorch-frontend                2025.2.0                  hecca717_1             http://mirrors.i.h.pjlab.org.cn/repository/conda-tsinghua-cloud/conda-forge
[conda] numpy                                       2.2.0                     pypi_0                 pypi
[conda] nvidia-cublas-cu12                          12.8.4.1                  pypi_0                 pypi
[conda] nvidia-cuda-cupti-cu12                      12.8.90                   pypi_0                 pypi
[conda] nvidia-cuda-nvrtc-cu12                      12.8.93                   pypi_0                 pypi
[conda] nvidia-cuda-runtime-cu12                    12.8.90                   pypi_0                 pypi
[conda] nvidia-cudnn-cu12                           9.10.2.21                 pypi_0                 pypi
[conda] nvidia-cudnn-frontend                       1.16.0                    pypi_0                 pypi
[conda] nvidia-cufft-cu12                           11.3.3.83                 pypi_0                 pypi
[conda] nvidia-cufile-cu12                          1.13.1.3                  pypi_0                 pypi
[conda] nvidia-curand-cu12                          10.3.9.90                 pypi_0                 pypi
[conda] nvidia-cusolver-cu12                        11.7.3.90                 pypi_0                 pypi
[conda] nvidia-cusparse-cu12                        12.5.8.93                 pypi_0                 pypi
[conda] nvidia-cusparselt-cu12                      0.7.1                     pypi_0                 pypi
[conda] nvidia-cutlass-dsl                          4.4.1                     pypi_0                 pypi
[conda] nvidia-cutlass-dsl-libs-base                4.4.1                     pypi_0                 pypi
[conda] nvidia-ml-py                                13.590.44                 pypi_0                 pypi
[conda] nvidia-nccl-cu12                            2.27.5                    pypi_0                 pypi
[conda] nvidia-nvjitlink-cu12                       12.8.93                   pypi_0                 pypi
[conda] nvidia-nvshmem-cu12                         3.4.5                     pypi_0                 pypi
[conda] nvidia-nvtx-cu12                            12.8.90                   pypi_0                 pypi
[conda] pyzmq                                       27.1.0                    pypi_0                 pypi
[conda] sentence-transformers                       5.2.0                     pypi_0                 pypi
[conda] torch                                       2.10.0+cu128              pypi_0                 pypi
[conda] torch-c-dlpack-ext                          0.1.5                     pypi_0                 pypi
[conda] torchaudio                                  2.10.0+cu128              pypi_0                 pypi
[conda] torchcodec                                  0.10.0+cu128              pypi_0                 pypi
[conda] torchsde                                    0.2.6                     pypi_0                 pypi
[conda] torchvision                                 0.25.0+cu128              pypi_0                 pypi
[conda] transformers                                4.57.6                    pypi_0                 pypi
[conda] triton                                      3.6.0                     pypi_0                 pypi
[conda] x-transformers                              2.16.2                    pypi_0                 pypi

==============================
         vLLM Info
==============================
ROCM Version                 : Could not collect
vLLM Version                 : 0.17.0
vLLM Build Flags:
  CUDA Archs: Not Set; ROCm: Disabled
GPU Topology:
        GPU0    GPU1    NIC0    NIC1    NIC2    NIC3    NIC4    NIC5    NIC6    NIC7    NIC8    NIC9    NIC10   NIC11   NIC12   NIC13   NIC14   NIC15NIC16   NIC17   CPU Affinity    NUMA Affinity   GPU NUMA ID
GPU0     X      NV18    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODEPIX      NODE    0       0               N/A
GPU1    NV18     X      NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODENODE     PIX                             N/A
NIC0    PIX     NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODEPIX      NODE
NIC1    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODENODE     NODE
NIC2    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODENODE     NODE
NIC3    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODENODE     NODE
NIC4    NODE    PIX     NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODENODE     PIX
NIC5    NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODENODE     NODE
NIC6    NODE    NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODENODE     NODE
NIC7    NODE    NODE    NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX NODE     NODE
NIC8    PIX     NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODE    NODEPIX      NODE
NIC9    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODE    NODENODE     NODE
NIC10   NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODE    NODENODE     NODE
NIC11   NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODE    NODENODE     NODE
NIC12   NODE    PIX     NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODE    NODENODE     PIX
NIC13   NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODE    NODENODE     NODE
NIC14   NODE    NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X      NODENODE     NODE
NIC15   NODE    NODE    NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE     X  NODE     NODE
NIC16   PIX     NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE X       NODE
NIC17   NODE    PIX     NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODE    NODE    NODE    NODE    NODE    PIX     NODE    NODE    NODENODE      X 

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

NIC Legend:

  NIC0: mlx5_0
  NIC1: mlx5_1
  NIC2: mlx5_2
  NIC3: mlx5_3
  NIC4: mlx5_4
  NIC5: mlx5_5
  NIC6: mlx5_6
  NIC7: mlx5_7
  NIC8: mlx5_20
  NIC9: mlx5_21
  NIC10: mlx5_22
  NIC11: mlx5_23
  NIC12: mlx5_24
  NIC13: mlx5_25
  NIC14: mlx5_26
  NIC15: mlx5_27
  NIC16: mlx5_bond_0
  NIC17: mlx5_data_0

==============================
     Environment Variables
==============================
LD_LIBRARY_PATH=/mnt/shared-storage-user/colab-share/cuda-12.8/lib64:/mnt/shared-storage-user/colab-share/cuda-12.8/lib64:/usr/local/nvidia/lib:/usr/local/nvidia/lib64
CUDA_HOME=/mnt/shared-storage-user/colab-share/cuda-12.8
CUDA_HOME=/mnt/shared-storage-user/colab-share/cuda-12.8
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
TORCHINDUCTOR_CACHE_DIR=/tmp

</details>

🐛 Describe the bug

运行：

vllm serve /mnt/models/Qwen3.5/Qwen3.5-122B-A10B --tensor-parallel-size 2`

日志输出：

# 1. 环境与参数确认
(APIServer pid=1222) INFO 03-09 12:26:19 [utils.py:302]  vllm version 0.17.0
(APIServer pid=1222) INFO 03-09 12:26:19 [utils.py:302]  model   /mnt/user/models/Qwen3.5/Qwen3.5-122B-A10B
(APIServer pid=1222) INFO 03-09 12:26:19 [utils.py:238] non-default args: {'model': '...', 'tensor_parallel_size': 2}
(APIServer pid=1222) INFO 03-09 12:26:19 [model.py:531] Resolved architecture: Qwen3_5MoeForConditionalGeneration
(APIServer pid=1222) INFO 03-09 12:26:19 [scheduler.py:231] Chunked prefill is enabled with max_num_batched_tokens=8192.

# 2. 分布式初始化 (并行状态)
(Worker pid=1441) INFO 03-09 12:26:42 [parallel_state.py:1393] world_size=2 rank=0 local_rank=0 distributed_init_method=tcp://127.0.0.1:56759 backend=nccl
(Worker pid=1441) INFO 03-09 12:26:42 [pynccl.py:111] vLLM is using nccl==2.27.5
(Worker pid=1441) (Worker_TP0 pid=1441) INFO 03-09 12:26:51 [cuda.py:405] Using FLASH_ATTN attention backend.

# 3. 模型权重加载
Loading safetensors checkpoint shards:   0% Completed | 0/39 [00:00<?, ?it/s]
...
Loading safetensors checkpoint shards: 100% Completed | 39/39 [10:01<00:00, 15.41s/it]
(Worker pid=1441) (Worker_TP0 pid=1441) INFO 03-09 12:36:54 [gpu_model_runner.py:4338] Model loading took 114.35 GiB memory

# 4. 关键阶段：CUDA Graph 捕获与编译 (崩溃发生在此处)
(Worker pid=1441) (Worker_TP0 pid=1441) INFO 03-09 12:37:08 [backends.py:350] Cache the graph of compile range (1, 8192) for later use
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 100%|██████████| 51/51 [00:06<00:00,  7.49it/s]
Capturing CUDA graphs (decode, FULL):   0%|          | 0/51 [00:00<?, ?it/s]

# 5. 核心错误堆栈 (Worker Rank 1 触发 AssertionError)
(Worker pid=1442) (Worker_TP1 pid=1442) ERROR 03-09 12:37:45 [multiproc_executor.py:880] Traceback (most recent call last):
  File ".../vllm/v1/worker/gpu_worker.py", line 522, in compile_or_warm_up_model
    cuda_graph_memory_bytes = self.model_runner.capture_model()
  File ".../vllm/v1/worker/gpu_model_runner.py", line 5337, in capture_model
    self._capture_cudagraphs(...)
  File ".../vllm/v1/worker/gpu_model_runner.py", line 5430, in _capture_cudagraphs
    dummy_run(...)
  File ".../vllm/v1/worker/gpu_model_runner.py", line 4976, in _dummy_run
    outputs = self.model(...)
  File ".../vllm/model_executor/models/qwen3_5.py", line 738, in forward
    hidden_states = self.language_model.model(...)
  File ".../vllm/model_executor/models/qwen3_next.py", line 1132, in forward (AOT compiled)
  File "<eval_with_key>.121", line 5, in forward
    gdn_attention_core = torch.ops.vllm.gdn_attention_core(...)
  File ".../vllm/model_executor/models/qwen3_next.py", line 1451, in gdn_attention_core
    self._forward_core(...)
  File ".../vllm/model_executor/models/qwen3_next.py", line 715, in _forward_core
    mixed_qkv_non_spec = causal_conv1d_update(...)
  File ".../vllm/model_executor/layers/mamba/ops/causal_conv1d.py", line 1162, in causal_conv1d_update
    assert num_cache_lines >= batch
AssertionError

# 6. 系统收尾报错
(EngineCore_DP0 pid=1335) ERROR 03-09 12:37:45 [core.py:1100] EngineCore failed to start.
(APIServer pid=1222) RuntimeError: Engine core initialization failed. See root cause above.

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.

extent analysis

Fix Plan

To fix the AssertionError caused by num_cache_lines >= batch in the causal_conv1d_update function, we need to ensure that the number of cache lines is sufficient for the batch size. Here are the steps:

Check the batch size: Verify that the batch size is not exceeding the expected value. You can do this by adding a print statement or a logger before the assert statement to print the batch size.
Increase the cache size: If the batch size is correct, you may need to increase the cache size to accommodate the batch. You can do this by modifying the num_cache_lines variable or by adjusting the model configuration to use more cache lines.
Modify the causal convolution layer: If increasing the cache size is not feasible, you may need to modify the causal convolution layer to handle larger batch sizes. This could involve changing the layer's architecture or implementing a more efficient caching mechanism.

Here's an example code snippet that demonstrates how to modify the causal_conv1d_update function to handle larger batch sizes:

def causal_conv1d_update(self, input, cache):
    # ... (rest of the function remains the same)
    num_cache_lines = self.num_cache_lines
    batch = input.shape    
    # Check if the batch size exceeds the cache size
    if batch > num_cache_lines:
        # Either increase the cache size or modify the layer to handle larger batch sizes
        num_cache_lines = batch
    
    # ... (rest of the function remains the same)
    mixed_qkv_non_spec = torch.ops.vllm.causal_conv1d_update(input, cache, num_cache_lines)
    return mixed_qkv_non_spec

In this example, we've added a check to see if the batch size exceeds the cache size. If it does, we increase the cache size to accommodate the batch. This is just one possible solution, and you may need to modify the code further to suit your specific use case.

Verification

To verify that the fix worked, you can run the model with the modified causal_conv1d_update function and check for any errors. You can also add print statements or loggers to monitor the batch size and cache size during runtime.

Extra Tips

Make sure to test the model with different batch sizes to ensure that the fix works for all scenarios.
Consider implementing a more efficient caching mechanism to reduce memory usage and improve performance.
If you're using a pre-trained model, you may need to retrain the model with the modified causal_conv1d_update function to ensure that the weights are updated correctly.

Vote matrix · Quick signals

Works

Did the solution work? Tap to confirm.

Easy Fix

Was it a quick fix?

Time Saver

Did it save you time?

Blocking

Was it severely blocking?

Common Issue

Are others likely hitting this too?

Flaky / Intermittent

Is it intermittent?

Verified / Reproducible

Can you reproduce it reliably?

#api #ssr #installation #model loading #dependency error #configuration error #environment variable

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Data

Security

Network

Code

UI/UX

Text

System

Multimedia

Protocol

API

Engineering

vllm - 💡(How to fix) Fix [Bug]: vllm 0.17.0 启动Qwen3.5-122B-A10B失败 [1 comments, 2 participants]

Recommended Tools

GitHub issue graph ai analysis

Error Message

1. 环境与参数确认

2. 分布式初始化 (并行状态)

3. 模型权重加载

4. 关键阶段：CUDA Graph 捕获与编译 (崩溃发生在此处)

5. 核心错误堆栈 (Worker Rank 1 触发 AssertionError)

6. 系统收尾报错

Root Cause

6. 系统收尾报错

Fix Action

Fix / Workaround

============================== CPU Info

Code Example

Your current environment

🐛 Describe the bug

Before submitting a new issue...

extent analysis

Fix Plan

Verification

Extra Tips

Still need to ship something?

TRENDING

vllm - 💡(How to fix) Fix [Bug]: vllm 0.17.0 启动Qwen3.5-122B-A10B失败 [1 comments, 2 participants]

Recommended Tools

GitHub issue graph ai analysis

Error Message

1. 环境与参数确认

2. 分布式初始化 (并行状态)

3. 模型权重加载

4. 关键阶段：CUDA Graph 捕获与编译 (崩溃发生在此处)

5. 核心错误堆栈 (Worker Rank 1 触发 AssertionError)

6. 系统收尾报错

Root Cause

6. 系统收尾报错

Fix Action

Fix / Workaround

============================== CPU Info

Code Example

Your current environment

🐛 Describe the bug

Before submitting a new issue...

extent analysis

Fix Plan

Verification

Extra Tips

Still need to ship something?

RELATED_DISCOVERY

TRENDING