-
-
Notifications
You must be signed in to change notification settings - Fork 11.8k
[WIP][Attention][FA3] Update FA3 to include new swizzle optimization #23465
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: main
Are you sure you want to change the base?
[WIP][Attention][FA3] Update FA3 to include new swizzle optimization #23465
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Code Review
This pull request updates the flash-attention dependency to a new commit hash, likely to incorporate the 'FA3 swizzle optimization' mentioned in the PR title. While pinning to a commit is good for reproducibility, using a commit hash that is not part of a tag or a long-lived branch can pose a risk for future builds and maintenance. I've added a comment suggesting the use of a git tag for better long-term stability.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Pinning dependencies to a specific commit hash is good for reproducibility. However, for long-term maintenance and release builds, it's better to use a git tag. Commit hashes that are not part of a branch or tag can be garbage collected by git, or become hard to track. Since this is a work-in-progress, it's acceptable for now, but it would be best to create a tag in the vllm-project/flash-attention repository for this commit before this PR is merged.
Signed-off-by: Lucas Wilkinson <[email protected]>
ed1629a to
7a4e376
Compare
Signed-off-by: Lucas Wilkinson <[email protected]>
Signed-off-by: Lucas Wilkinson <[email protected]>
Signed-off-by: Lucas Wilkinson <[email protected]>
Signed-off-by: Lucas Wilkinson <[email protected]>
|
This pull request has been automatically marked as stale because it has not had any activity within 90 days. It will be automatically closed if no further activity occurs within 30 days. Leave a comment if you feel this pull request should remain open. Thank you! |
WIP; vLLM side of vllm-project/flash-attention#82 merge that first
This PR is just marked ready to validate CI and building before merging vllm-project/flash-attention#82