这是indexloc提供的服务,不要输入任何密码
Skip to content

[Installation]: Gemma2 Installing Flash Infer [rank0]: TypeError: 'NoneType' object is not callable #6237

@robertgshaw2-redhat

Description

@robertgshaw2-redhat

The source of your error is that you have installed the wrong version of FlashInfer.

FlashInfer builds wheels for specific torch and cuda versions. vLLM v0.5.1 uses torch==2.3 and cuda==12.1. So you will likely want to download the following wheel:

pip install flashinfer -i https://flashinfer.ai/whl/cu121/torch2.3

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions