Skip to content

Fix type annotation to allow ref model to be null#1091

Merged
cbalioglu merged 4 commits intomainfrom
ansong/fix_dpo_precompute_type
Mar 25, 2025
Merged

Fix type annotation to allow ref model to be null#1091
cbalioglu merged 4 commits intomainfrom
ansong/fix_dpo_precompute_type

Conversation

@niansong1996
Copy link
Copy Markdown
Contributor

@niansong1996 niansong1996 commented Mar 22, 2025

What does this PR do? Please describe:
Current type annotation for reference_model in DpoFinetuneConfig would give a structure error when setting reference_model=null and use precomputed reference scores instead.

Does your PR introduce any breaking changes? If yes, please list them:
Small change, N/A

Check list:

  • Was the content of this PR discussed and approved via a GitHub issue? (no need for typos or documentation improvements)
  • Did you read the contributor guideline?
  • Did you make sure that your PR does only one thing instead of bundling different changes together?
  • Did you make sure to update the documentation with your changes? (if necessary)
  • Did you write any new necessary tests?
  • Did you verify new and existing tests pass locally with your changes?
  • Did you update the CHANGELOG? (no need for typos, documentation, or minor internal changes)

@niansong1996 niansong1996 requested a review from cbalioglu as a code owner March 22, 2025 05:38
@facebook-github-bot facebook-github-bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Mar 22, 2025
@niansong1996 niansong1996 requested a review from uralik March 22, 2025 05:38
Copy link
Copy Markdown
Contributor

@cbalioglu cbalioglu left a comment

Choose a reason for hiding this comment

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

LGTM! Thanks for the fix @niansong1996!

@cbalioglu cbalioglu merged commit 5d0b40a into main Mar 25, 2025
15 checks passed
@cbalioglu cbalioglu deleted the ansong/fix_dpo_precompute_type branch March 25, 2025 13:16
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants