remove padding mask for input embeddings #1799
Merged
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
ebsmothers
approved these changes
Jun 21, 2022
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Outdated suggestions cannot be applied.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.
This PR removes the masking at Padded tokens for the input embeddings.
In fairseq, the masking is applied to input embedding here but not in HF implementation. This causes MisMatch in output embedding for the padded tokens.
Ideally, it should not matter the output for padded tokens. Per the investigations from @ebsmothers, it is somehow causing results to differ for MDETR model. In order for Torch MM to upstream dependency on torchtext for RoBERTa encoder, this change is necessary.