Skip to content

[Bug]: v0.11.0rc0部署Qwen3-Next-80B-A3B-Instruct在问到数字相关问题是精度较差 #4057

@qinghuanguayu

Description

@qinghuanguayu

Your current environment

The output of `python collect_env.py`
Your output of above commands here

🐛 Describe the bug

我根据 https://vllm-ascend.readthedocs.io/en/latest/tutorials/multi_npu_qwen3_next.html 的教程用v0.11.0rc0部署了Qwen3-Next-80B-A3B-Instruct模型,正常对话精度没有问题,但是当问的问题中有数字的时候,则会出现胡言乱语的情况

Image

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't working

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions