Skip to content

Conversation

@Potabk
Copy link
Contributor

@Potabk Potabk commented Dec 2, 2025

Purpose

This patch purpose pin transformers version to latest 4.57.3, for the release 4.57.2, there was a hidden bug when loading models with local_files_only=True, see huggingface/transformers#42389

Test Plan

Test Result


Essential Elements of an Effective PR Description Checklist
  • The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
  • The test plan, such as providing test command.
  • The test results, such as pasting the results comparison before and after, or e2e results
  • (Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.
  • (Optional) Release notes update. If your change is user facing, please update the release notes draft in the Google Doc.

Signed-off-by: wangli <wangli858794774@gmail.com>
@mergify mergify bot added the ci/build label Dec 2, 2025
Copy link
Contributor

@gemini-code-assist gemini-code-assist bot left a comment

Choose a reason for hiding this comment

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

Code Review

This pull request updates the transformers dependency to version >= 4.57.3. As explained in the description, this is a necessary change to fix a bug in version 4.57.2 related to local_files_only=True. The change is correct and I approve it.

@DarkLight1337
Copy link
Member

Our Docker images are built using the latest available version so that should be the case for most users already.

For those who install vLLM manually, I think that it should be up to the users to upgrade their vLLM version to resolve the issue. We should not force a minimum version unless it affects the coupling between vLLM and Transformers.

@Potabk
Copy link
Contributor Author

Potabk commented Dec 2, 2025

looks reasonable

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

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants