minimaxm2.5-fp8-h200-vllm: switch 8k/1k attention backend to FLASH_ATTN#1667
Closed
RohitNagraj wants to merge 2 commits into
Closed
minimaxm2.5-fp8-h200-vllm: switch 8k/1k attention backend to FLASH_ATTN#1667RohitNagraj wants to merge 2 commits into
RohitNagraj wants to merge 2 commits into
Conversation
…SH_ATTN Switch the attention backend for the 8k/1k cell of minimaxm2.5-fp8-h200-vllm from FLASHINFER to FLASH_ATTN. ISL-conditional: the 1k/1k cell is unchanged (keeps FLASHINFER + --enable-flashinfer-autotune, byte-identical to prior behavior); only ISL=8192 triggers the swap. Appends a perf-changelog entry.
![claude[bot]](https://avatars.githubusercontent.com/in/1236702?s=60&v=4){kind=small}
Collaborator
Author
|
Superseded by #1668, which is opened from a branch in this repository instead of a fork. Closing in favor of that PR. |
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Switch the attention backend for the 8k/1k cell of minimaxm2.5-fp8-h200-vllm from FLASHINFER to FLASH_ATTN. ISL-conditional: the 1k/1k cell is unchanged (keeps FLASHINFER + --enable-flashinfer-autotune, byte-identical to prior behavior); only ISL=8192 triggers the swap.
Appends a perf-changelog entry.
Note
Low Risk
Benchmark-only vLLM flag selection for one ISL cell; no auth, data, or production serving paths.
Overview
For MiniMax-M2.5 FP8 H200 vLLM (
minimaxm2.5_fp8_h200.sh),vllm servenow picks the attention stack fromISL: whenISL=8192(8k/1k), it usesFLASH_ATTNand does not pass--enable-flashinfer-autotune; for all other input lengths it keepsFLASHINFERplus FlashInfer autotune, matching prior 1k/1k behavior.A
perf-changelog.yamlentry documents this underminimaxm2.5-fp8-h200-vllm.Reviewed by Cursor Bugbot for commit 424fe77. Bugbot is set up for automated code reviews on this repo. Configure here.