Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Speed up update_weights_from_tensor #2695

Merged
merged 23 commits into from
Jan 2, 2025

Conversation

fzyzcjy
Copy link
Collaborator

@fzyzcjy fzyzcjy commented Jan 1, 2025

Motivation

Speed up by avoiding tensor serialization and deserialization. Since it was only used for tests before and never released, the old one was OK, and the new one with API breaking changes seems OK.

Modifications

Checklist

  • Format your code according to the Contributor Guide.
  • Add unit tests as outlined in the Contributor Guide.
  • Update documentation as needed, including docstrings or example tutorials.

@fzyzcjy
Copy link
Collaborator Author

fzyzcjy commented Jan 1, 2025

CI looks green now

@fzyzcjy
Copy link
Collaborator Author

fzyzcjy commented Jan 2, 2025

Hmm, quick question firstly: is it possible this is flakyness? Commit 5867d7a fails accuracy-test-2, while 8f8c9c1 (same logic) fails accuracy-test-1. And the failures does not seem to be super related to the current PR at first glance.

@merrymercy merrymercy merged commit 9183c23 into sgl-project:main Jan 2, 2025
14 of 15 checks passed
@merrymercy
Copy link
Contributor

@fzyzcjy Yes, some these are flaky.

@fzyzcjy
Copy link
Collaborator Author

fzyzcjy commented Jan 2, 2025

I see

YAMY1234 pushed a commit to YAMY1234/sglang that referenced this pull request Jan 2, 2025
XiaotongJiang pushed a commit to XiaotongJiang/sglang that referenced this pull request Jan 3, 2025
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.

2 participants