optimum
2678e74d - Allow `attention_mask=None` for BetterTransformer in the inference batched case for gpt2 & gpt-neo (#1180)

Commit

2 years ago

Allow `attention_mask=None` for BetterTransformer in the inference batched case for gpt2 & gpt-neo (#1180) fix if mask is none

References

Author

fxmarty

Parents