Some fixes for frozen weights#270
Merged
Merged
Conversation
jlamypoirier
approved these changes
May 15, 2025
Collaborator
jlamypoirier
left a comment
There was a problem hiding this comment.
Seems good, thanks for the fixes!
Contributor
Author
|
Added another fix for cases where loaded and current config don't have the same set of frozen weights @jlamypoirier . |
Contributor
Author
|
Thanks for the review @jlamypoirier ! Let me know if the new changes are OK |
jlamypoirier
approved these changes
May 16, 2025
| ], | ||
| compare=f"test_{TEST_MODEL}_checkpoint_and_eval", | ||
| prepare_fn=_prepare_resume_fn, | ||
| do_compare=False, |
Collaborator
There was a problem hiding this comment.
Why not just removing compare?
Contributor
Author
There was a problem hiding this comment.
the prepare_resume_fn uses compare to copy the checkpoints
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
✨ Description
Closes #265
Closes #256
1 - log gradients only for parameters that require_grad
2 - Fix shard validation when loading from distributed format
3 - In
copy_shard_overlaps, skip when current shard is empty (exp_avgs,exp_avgs_sqfor frozen weights), and fill with zero when loaded shard is empty.🔍 Type of change
Select all that apply:
🗒️ Additional Notes
Include any additional context, information, or considerations here, such as known issues, follow-up tasks, or backward compatibility concerns.