Fix Training Error with AdapterDrop and Prefix Tuning #673
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
…refix tuning with AdapterDrop
calpt
reviewed
Apr 11, 2024
| # For Prefix Tuning, when training with AdapterDrop, we must additionally check if the dimensions of the | ||
| # position_bias given from previous layers match the dimensions of the scores of the current layer to make | ||
| # sure that the position_bias is adequately recomputed if previous layers have been skipped. | ||
| if position_bias is None or position_bias.shape != scores.shape: |
There was a problem hiding this comment.
can we specifically check the seqlen dimensions here to make this condition more precise?
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Outdated suggestions cannot be applied.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.
Fixes #669
Changes in this PR:
RuntimeErrordue to dimension mismatch occuring when passing the positional encoding from layers dropped by AdapterDrop to layers modified by prefix tuning.