Skip to content

Flash Attention 2 #64

@jacobfulano

Description

@jacobfulano

Hi @Zhihan1996,

You might also be interested in this PR which updates MosaicBERT to FlashAttention 2 and removes all the issues with the custom Triton FlashAttention + ALiBi implementation.

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