Skip to content

max_tokens 500报错 #87

@MingCangx

Description

@MingCangx

0.4.9及以后版本
站点报错:status_code=500, max_tokens=1000000cannot be greater than max_model_len=max_total_tokens=262144. Please request fewer output tokens. (parameter=max_tokens, value=1000000.
尝试在env文件里调整,INKOS_LLM_MAX_TOKENS=100000改成50000,20000均无效果,上文报错是1000000,高于默认值10倍。
回退到0.4.5可以正常使用。

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions