Move _get_kl_completion_ids into _get_kl_dataset#6181
Open
qgallouedec wants to merge 3 commits into
Open
Conversation
|
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. |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
_get_kl_completion_idswas a module-level helper used in exactly one place. Move it into its only caller,KTOTrainer._get_kl_dataset, as a nested function, alongside the existingrename_kl_fn.The direct unit test that imported the helper is removed; its behavior stays covered through the trainer's KL-dataset preparation.
Note
Low Risk
Pure structural refactor with identical KL dataset mapping logic; only test surface area shrinks for the former public helper.
Overview
Refactors where KL mismatching is defined: the module-level
_get_kl_completion_idshelper is removed and the same batched completion cycling logic lives as a nestedget_kl_completion_idsinsideKTOTrainer._get_kl_dataset, next torename_kl_fn. The vision collator comment now points at that nested name.Tests drop the standalone unit test that imported
_get_kl_completion_idsand asserted rotation on a syntheticDataset; imports are trimmed accordingly. KL pairing behavior is unchanged and remains covered via trainer/collator paths (e.g.test_kl_cycling, end-to-end KTO training).Reviewed by Cursor Bugbot for commit 35edb17. Bugbot is set up for automated code reviews on this repo. Configure here.