Skip to content

Conversation

@Ki-Seki
Copy link

@Ki-Seki Ki-Seki commented Oct 20, 2025

What does this PR do?

Refactor the parameter freezing logic to simplify the handling of shared and unshared parameters. The original implementation was confusing; this version is clearer. Additionally, the current shared approach merely performs a deep copy—consider adding a TODO to switch to a shallow copy for shared layers in the future to save GPU memory.

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? Please add a link
    to it if that's the case.
  • Did you make sure to update the documentation with your changes?
  • Did you write any new necessary tests?

Who can review?

@lvwerra @lewtun @albertvillanova @qgallouedec

Ki-Seki and others added 2 commits October 20, 2025 11:55
Refactor parameter freezing logic to simplify shared and unshared handling.
@HuggingFaceDocBuilderDev

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

@Ki-Seki
Copy link
Author

Ki-Seki commented Oct 21, 2025

@qgallouede, thank you for running checks. I found that two CI checks failed, which may be some issues with the main branch. This PR is just a simple refactor.🤗

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.

3 participants