Skip to content

RuntimeError:RunOpsApiV2:build/CMakeFiles/torch_npu_dir/compile_depend.ts:243 NPU function error:c10_npu::acl::AclrtSynchronizeStreamWithTimeOut(stream),error code is 107027 #77

@frankie-ys

Description

@frankie-ys

I have meet this problem when use two piece 910B, start options is
export ASCEND_LAUNCH_BLOCKING=1 export VLLM_ALLOW_LONG_MAX_MODEL_LEN=1 vllm serve /data/qwen3-32B --port 8080 --max-model-len 80000 -tp 2 --enable-prefix-caching --enable-chunked-prefill --max-num-batched-tokens 4096 --rope-scaling '{"rope_type":"yarn","factor":4.0,"original_max_position_embeddings":80000}'
after safetensors checkpoint load completed, when waitting for pendding NCCL work to finish before starting graph capture. It occured EE9999 Inner Error and descripe like the title. How can I solve this problem?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions