add attention backend recommendation for Minimax 2.5#512
Merged
Conversation
Signed-off-by: Faradawn Yang <73060648+faradawn@users.noreply.github.com>
Contributor
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Code Review
This pull request updates the deployment guide for MiniMax-M2.5 to recommend specific attention backends (FLASHINFER or FLASH_ATTN) based on sequence length for H200 FP8. The reviewer suggested phrasing improvements to make the instructions clearer and more direct, along with minor punctuation corrections.
Comment on lines
+190
to
+193
| For H200 FP8, choose the attention backend by sequence length for best performance: | ||
| shorter sequences (e.g. 1024) should keep the command above with `--attention-backend FLASHINFER` | ||
| and `--enable-flashinfer-autotune`, while longer input sequences (e.g. 8192) | ||
| can prefer FlashAttention by replacing those flags with `--attention-backend FLASH_ATTN`. |
Contributor
There was a problem hiding this comment.
The phrase 'can prefer FlashAttention' is slightly awkward. It is clearer and more direct to state that longer sequences 'should use' FlashAttention for optimal performance. Also, adding commas after 'e.g.' is standard style.
For H200 FP8, choose the attention backend by sequence length for best performance:
shorter sequences (e.g., 1024) should keep the command above with `--attention-backend FLASHINFER`
and `--enable-flashinfer-autotune`, while longer input sequences (e.g., 8192)
should use FlashAttention by replacing those flags with `--attention-backend FLASH_ATTN`.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
For longer sequence lengths, use flash attention backend for best performance for minimax on H200 FP8.
Reference: SemiAnalysisAI/InferenceX#1668