Update prepare 4d causal attention call #2678
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
The latest transformers removes the device argument to _prepare_4d_causal_attention_mask_with_cache_position, and since we provide device arg via _fast_prepare_inputs_for_generation, inference fails for models that have this function defined. Qwen2 is one such examples.
I have a backward compatible fix that inpsect the function signature for device, and calls the function appropriately.
test before fix
https://colab.research.google.com/drive/1EMVkur2wSTgvvbZawGuvtoxLPPPlj_QQ?usp=sharing
after fix
https://colab.research.google.com/drive/1m15_peSfF7TxcMUC7Y2nsp_kGk1QyDIH?usp=sharing
you'll see before that inference fails, but after it works as it should.