Warn when fine-tuning chromosomes overlap with pretraining data#184
Open
LiudengZhang wants to merge 1 commit intoGenentech:mainfrom
Open
Warn when fine-tuning chromosomes overlap with pretraining data#184LiudengZhang wants to merge 1 commit intoGenentech:mainfrom
LiudengZhang wants to merge 1 commit intoGenentech:mainfrom
Conversation
Add a warnings.warn() in tune_on_dataset() that checks whether the fine-tuning dataset's chromosomes overlap with the pretrained model's training chromosomes, which may indicate data leakage. The check uses chromosome metadata already stored in the checkpoint via data_params['train']['intervals']. It is skipped when either the pretrained model or the new dataset lacks interval information. Closes Genentech#59
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Closes #59.
When calling
tune_on_dataset(), this checks whether the fine-tuning dataset's chromosomes overlap with the pretrained model's training chromosomes and emits awarnings.warn()if so, since this may indicate data leakage.The check uses chromosome metadata already stored in the checkpoint via
data_params['train']['intervals']. It is skipped when either the pretrained model or the new dataset lacks interval information (e.g. string-based sequences).Tests added:
test_lightning_model_finetune_chrom_overlap_warning: simulates a pretrained model trained on chr1, fine-tunes with a chr1 dataset, asserts warning firestest_lightning_model_finetune_no_chrom_warning: fine-tunes with string-based dataset, asserts no warning