Skip to content

Use one mma warp group to optimize decoding performance#34

Merged
Qiaolin-Yu merged 4 commits intosgl-kernelfrom
qiaolin_dev
Feb 20, 2026
Merged

Use one mma warp group to optimize decoding performance#34
Qiaolin-Yu merged 4 commits intosgl-kernelfrom
qiaolin_dev

Conversation

@Qiaolin-Yu
Copy link
Copy Markdown
Collaborator

No description provided.

@Qiaolin-Yu
Copy link
Copy Markdown
Collaborator Author

Works well on my local server. sglang side pr: sgl-project/sglang#18985
Merge it just for better e2e test, free free to revert it if bug found.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant