[transformers] Prompt masking #2192
Open
Conversation
This file contains 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
bfineran
previously approved these changes
Mar 20, 2024
Satrat
suggested changes
Mar 26, 2024
There was a problem hiding this comment.
Calculation of the mask looks good but pretty sure its being applied backwards. Lets add a test case to test the whole pipeline so we catch this error(ie text -> tokenization -> text and confirm the labels are masked out only on the prompts.
Also, can you update the internal_docs repo with instructions on how to use this new feature on a custom dataset: https://github.com/neuralmagic/internal-docs/blob/main/teams/ml-engineering/sparseml/text_generation/custom_datasets.md
Satrat
suggested changes
Apr 2, 2024
Satrat
previously approved these changes
Apr 5, 2024
bfineran
previously approved these changes
Apr 8, 2024
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Outdated suggestions cannot be applied.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.
Applying character masks to a prompts in the format
[foo]some text here\n[bar]response here, to mask characters owned by[bar]