Add type hints for BERTGeneration

[robsmith155]

What does this PR do?

I added type hints for the BERTGenerationEncoder and BERTGenerationDecoder classes as requested in #16059 and demonstrated in #16074.

I wasn't completely sure on what to use for the past_key_values argument. I set it to Optional[Tuple[Tuple[torch.FloatTensor]]], but let me know if this is wrong. Also, not sure if I should also add type hints for the BertGenerationConfig class?

@Rocketknight1

[HuggingFaceDocBuilderDev]

The documentation is not available anymore as the PR was closed or merged.

[Rocketknight1]

Looks good to me, thank you!

[Rocketknight1]

I wasn't completely sure on what to use for the past_key_values argument. I set it to Optional[Tuple[Tuple[torch.FloatTensor]]], but let me know if this is wrong. Also, not sure if I should also add type hints for the BertGenerationConfig class?

What you put here is fine, and type hints for BertGenerationConfig are nice but optional - if you want to do them you can, but the main thing we're interested in is the core model classes. Let me know either way - if you don't want to do it, this is ready to merge now!

[robsmith155]

Okay great, you can go ahead and merge it then. I'll run your notebook to see what else needs to be done and work on some of those instead. Cheers

[Rocketknight1]

Got it. Thanks for the PR!