refactor: Move CPOTrainer to experimental module #4470
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
Move CPOTrainer and CPOConfig to trl.experimental.cpo following the pattern established in #4312 for BCO. Changes: - Create trl/experimental/cpo/ module with CPOTrainer and CPOConfig - Replace original files with deprecation stubs - Update all imports in tests, examples, and documentation - Move CPO from Trainers to Experimental section in docs - Maintain backward compatibility with deprecation warnings until TRL 0.29 Closes #4460
- Remove unused imports (os, warnings) from experimental/cpo/cpo_trainer.py - Reorder config import after utils imports - Split combined import into separate lines in trainer/cpo_trainer.py - Add missing torch import for type hints in deprecation stub All ruff checks now pass.
|
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. |
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Outdated suggestions cannot be applied.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.
Summary
Moves
CPOTrainerandCPOConfigtotrl.experimental.cpofollowing the pattern established in #4312 for BCO migration.This is part of the V1 refactoring effort tracked in #4223.
Changes
trl/experimental/cpo/module with full CPOTrainer and CPOConfig implementationtest_cpo_trainer.py,test_trainers_args.py)examples/scripts/cpo.py)_toctree.ymlcpo_trainer.mdpaper_index.mdBackward Compatibility
The old import path still works but issues a
FutureWarning:New import path:
Testing
All existing tests updated to use the new import path. The migration follows the exact same pattern as BCO (#4312), which has been successfully integrated.
Closes #4460