unsloth checkpointing fix for latest transformers==4.52.x #2674
+11
−1
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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Outdated suggestions cannot be applied.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.
transformers==4.52.x introduced a GradientCheckpointingLayer and refactored the checkpointing logic. This adjusts FastLanguageModel to bypass our custom checkpoint logic when decoder layers are an instance of GradientCheckpointingLayer.
I have 3 notebooks to compare.
transformers==4.51.3
https://colab.research.google.com/drive/19tEe55Z-b3oz61S6R5diBoHAxHZVnOlp?usp=sharing
transformers==4.52.4
https://colab.research.google.com/drive/1IQZGdYYoF73NqG3WtO_ar7rfunllWGSr?usp=sharing
transformers==4.52.4 + checkpointing fix
https://colab.research.google.com/drive/1nfA9Sc20lBAbTzmo2S2VgAW-P6ZkOkMq?usp=sharing
as you can see in the second notebook the loss remains the same but timing is longer. In the final notebook loss and speed matches original transformers.