⚖️ Fix scale_rewards issue in GRPO #3992
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
LeonEricsson
reviewed
Sep 2, 2025
trl/trainer/grpo_config.py
Outdated
Comment on lines
618
to
623
| self.scale_rewards = {"true": "group", "false": "none"}.get(self.scale_rewards.lower(), self.scale_rewards.lower()) | ||
| if self.scale_rewards not in ["batch", "none", "group"]: | ||
| raise ValueError( | ||
| f"Invalid value for scale_rewards: {self.scale_rewards}. Must be one of 'batch', 'group', or 'none'." | ||
| ) | ||
|
|
Collaborator
There was a problem hiding this comment.
@qgallouedec I noticed that you moved this into the trainer in the original PR. Thoughts? I prefer eager validation like this.
Member
There was a problem hiding this comment.
Yes, I usually try to minimize what we have in post-init. But here I guess it's ok
|
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. |
qgallouedec
reviewed
Sep 2, 2025
qgallouedec
reviewed
Sep 2, 2025
qgallouedec
reviewed
Sep 2, 2025
Member
qgallouedec
left a comment
qgallouedec
left a comment
There was a problem hiding this comment.
lgtm, thanks a lot, a patch release will soon be released
qgallouedec
changed the title
qgallouedec
added a commit
that referenced
this pull request
Sep 3, 2025
Co-authored-by: Leon <[email protected]> Co-authored-by: Quentin Gallouédec <[email protected]>
SamY724
pushed a commit
to SamY724/trl
that referenced
this pull request
Sep 6, 2025
Co-authored-by: Leon <[email protected]> Co-authored-by: Quentin Gallouédec <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What does this PR do?
Fixes scale_rewards malfunctioned in GRPOTrainer #3991
Before submitting
Pull Request section?
to it if that's the case.
Who can review?
Amazing team members such as @qgallouedec