What does this PR do?

While the input sequences are padded by tokenizer.pad which makes the length of the sequences to be a multiple of pad_to_multiple_of, the label sequences are padded to the maximum length among them without considering pad_to_multiple_of. This PR changes DataCollatorForSeq2Seq class to make sure that the labels have their lengths to be a multiple of the given value.

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?

@sgugger

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.