[Docs] Improve PyTorch, Flax generate API

[patrickvonplaten]

What does this PR do?

This PR is the first step to make generate a 1st class citizen in the docs. It improves the generate API for PyTorch and Flax generate, improves the examples for PyTorch and adds PyTorch to the example doc tests.

Once the TF generate refactor is complete - it's API can also be improved with better examples.

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline,
  Pull Request section?
• Was this discussed/approved via a Github issue or the forum? Please add a link
  to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

Who can review?

Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

[patrickvonplaten]

No doc-builder triggered here? 😢

[LysandreJik]

Docs live here: https://moon-ci-docs.huggingface.co/docs/transformers/pr_15988/en/index

[patrickvonplaten]

Docs live here: https://moon-ci-docs.huggingface.co/docs/transformers/pr_15988/en/index

The docs are not updated on the link if the PR is changed (or it takes too long). Will build the docs locally now, but I think it makes it quite difficult for the community to add/change docs.

[LysandreJik]

The job updates the docs. Are they not up to date here? https://moon-ci-docs.huggingface.co/docs/transformers/pr_15988/en/main_classes/text_generation

[patrickvonplaten]

The job updates the docs. Are they not up to date here? https://moon-ci-docs.huggingface.co/docs/transformers/pr_15988/en/main_classes/text_generation

Nope

[sgugger]

Thanks for expanding the doc!

[stevhliu]

I think replacing the first paragraph with the suggestion below makes it easier for the user to map each generate to its GenerationMixin class :)

Each framework has a generate method for auto-regressive text generation implemented in their respective GenerationMixin class:

• PyTorch [~generation_utils.GenerationMixin.generate] is implemented in [~generation_utils.GenerationMixin].
• TensorFlow [~generation_tf_utils.TFGenerationMixin.generate] is implemented in [~generation_tf_utils.TFGenerationMixin].
• Flax/JAX [~generation_flax_utils.FlaxGenerationMixin.generate] is implemented in [~generation_flax_utils.FlaxGenerationMixin].

[stevhliu]

Small typo for each of the GenerationMixin classes: GenerationMixin, TFGenerationMixin, FlaxGenerationMixin.

[patrickvonplaten]

Correcting it here: #16133 (comment) . Thanks!