fix(qwen-tts): install flash-attn on cuda13 images (#9293)#10293
Open
localai-bot wants to merge 1 commit into
Open
fix(qwen-tts): install flash-attn on cuda13 images (#9293)#10293localai-bot wants to merge 1 commit into
localai-bot wants to merge 1 commit into
Conversation
The cuda12 image installs flash-attn via requirements-cublas12-after.txt, but there was no cublas13 equivalent, so cuda13-qwen-tts never installed flash_attn and fell back to SDPA with warnings. Add the matching requirements-cublas13-after.txt. Assisted-by: claude:claude-opus-4-8 [Claude Code] Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Re: #9293 (the qwen-tts backend part)
Problem
On CUDA-13 the Qwen TTS backend logs flash-attn warnings and falls back to SDPA. The cuda12 image installs flash-attn via
requirements-cublas12-after.txt, but there was norequirements-cublas13-after.txt, socuda13-qwen-ttsnever installedflash_attn.Fix
Add
backend/python/qwen-tts/requirements-cublas13-after.txtcontainingflash-attn, mirroring the cublas12 variant.Scope
This addresses the qwen-backend performance part of the issue. The separate vllm-omni-fails-entirely part (related to #8536) needs its own reproduction and is not covered here. Additive requirements change, not built locally.
Assisted-by: claude:claude-opus-4-8 [Claude Code]