Skip to content

Dispatch MXFP4 weight conversion for sm70 & sm75#3937

Merged
lvhan028 merged 2 commits intoInternLM:mainfrom
lzhangzz:mxfp4b
Sep 5, 2025
Merged

Dispatch MXFP4 weight conversion for sm70 & sm75#3937
lvhan028 merged 2 commits intoInternLM:mainfrom
lzhangzz:mxfp4b

Conversation

@lzhangzz
Copy link
Copy Markdown
Collaborator

@lzhangzz lzhangzz commented Sep 4, 2025

Resolves #3934

@lvhan028
Copy link
Copy Markdown
Collaborator

lvhan028 commented Sep 5, 2025

Test pased on V100 platform. Amazing job!!

@lvhan028 lvhan028 added the enhancement New feature or request label Sep 5, 2025
@lvhan028 lvhan028 merged commit b77f157 into InternLM:main Sep 5, 2025
8 of 9 checks passed
littlegy pushed a commit to littlegy/lmdeploy that referenced this pull request Sep 11, 2025
* simplify weight conversion dispatch

* fix sm70 window attention
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

enhancement New feature or request

Projects

None yet

Development

Successfully merging this pull request may close these issues.

gpt-oss-20b crashed with turbomind (sm=70 not implemented)

2 participants