transformers
0142aab7 - don't zero out the attention_mask when using sliding window with flash attention (#31670)

Commit

1 year ago

don't zero out the attention_mask when using sliding window with flash attention (#31670) * don't zero out the attention_mask when using sliding window with flash attention * chore: lint

References

#31670 - don't zero out the attention_mask when using sliding window with flash attention

Author

winglian

Parents

1c68f2ca

transformers 0142aab7 - don't zero out the attention_mask when using sliding window with flash attention (#31670)

transformers
0142aab7 - don't zero out the attention_mask when using sliding window with flash attention (#31670)