OPT/BioGPT: Improved attention mask shape exception #23270
Merged
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
amyeroberts
approved these changes
May 12, 2023
|
The documentation is not available anymore as the PR was closed or merged. |
sheonhan
pushed a commit
to sheonhan/transformers
that referenced
this pull request
Jun 1, 2023
gojiteji
pushed a commit
to gojiteji/transformers
that referenced
this pull request
Jun 5, 2023
novice03
pushed a commit
to novice03/transformers
that referenced
this pull request
Jun 23, 2023
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Outdated suggestions cannot be applied.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.
What does this PR do?
Related exception: #23197
Currently, in OPT/BioGPT, if we don't pass an attention mask or if we pass an attention mask with a wrong shape, an exception will be printed in the attention layer:
Attention mask should be of size {(bsz, 1, tgt_len, src_len)}, but is {attention_mask.size()}. This exception has the following problems:bsz,tgt_len, orsrc_len, the exception will be misleading.In #23197 we found that when the length of
past_key_valuesis equal to the length of theattention_mask,tgt_lenandsrc_lenwill be wrong (in at least these 2 models), triggering the exception with an incorrect message. This PR solves both issues: it prevents the incorrect computation oftgt_lenandsrc_lenby checking the shape ofattention_maskin the main model class, printing a user-friendly message.If this PR gets approved, I will add a similar check to the other models.