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

[save_pretrained ] Skip collecting duplicated weight #36409

Open
wants to merge 2 commits into
base: main
Choose a base branch
from

Conversation

wejoncy
Copy link
Contributor

@wejoncy wejoncy commented Feb 26, 2025

What does this PR do?

Many weight in sub moduoles are collected multiple times. It's useless and time-wasting. Sometimes it even induces some issue when CPU/disk offload is enabled.

Fixes # (issue)

Before submitting

  • This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
  • Did you read the contributor guideline,
    Pull Request section?
  • Was this discussed/approved via a Github issue or the forum? Please add a link
    to it if that's the case.
  • Did you make sure to update the documentation with your changes? Here are the
    documentation guidelines, and
    here are tips on formatting docstrings.
  • Did you write any new necessary tests?

@Rocketknight1 @ArthurZucker

@wejoncy wejoncy changed the title Skip collecting duplicated weight [save_pretrained ] Skip collecting duplicated weight Feb 26, 2025
@gante
Copy link
Member

gante commented Feb 26, 2025

Hi @wejoncy 👋

I'm not a reviewer for this PR, but I can foresee a request: please share with us a script with an example where your change results in an upgrade 🤗

Copy link
Member

@SunMarc SunMarc left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sounds reasonable ! It would be nice to have comparison with a model before and after this PR as @gante pointed out

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.

4 participants