Accept token in trainer.push_to_hub()

[mapmeld]

What does this PR do?

The trainer's HF token is set to self.args.hub_token when it is initialized.
I use Transformers and trainer.train() in CoLab notebooks. If I start running with no token, a read-only token, or some other issue with Secrets/env vars, I can't correct the mistake later when calling trainer.push_to_hub()

With this change, I could try again with trainer.push_to_hub(token="****")

Alternatives: if overwriting self.args.hub_token is undesirable here, it could be passed to create_repo and update_folder in this file.

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?

[amyeroberts]

Thanks for adding this! Seems reasonable to me, especially as token is an arg for other push_to_hub methods.

Just a comment on state modification.

cc @muellerzr for reference

[amyeroberts]

Thanks for iterating and adding this capability!

[amyeroberts]

Merging as the failing tests are unrelated and have been resolved on main