Added Deberta model type support

[FilipposVentirozos]

What does this PR do?

Fixes in the token_classification example the support for Deberta. Since it was not included in an 'if' clause. Specifically, I added there since it's mentioned that Deberta needs the 'add_prefix_space' to be True and that 'if' clause covers this parameter. I tested it and it works well.

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.

[LysandreJik]

Hey @FilipposVentirozos, have you tried running the script with DeBERTa? Does it run/train successfully?

[FilipposVentirozos]

Hello @LysandreJik yes it runs as intended. That is Deberta version 1 (microsoft/deberta-base). The newer versions (i.e. v2, v3) still not supported cause of the need of the slow tokenisation instead, may look into it more later.

[FilipposVentirozos]

I tried again with microsoft/deberta-v3-base and it actually works. FYI, I tested it locally on my Mac and on my custom dataset and on Conll2003 that was for train, eval and predict. If you would like me to do any other experiment, please let me know.

[LysandreJik]

Ok let's merge it then! Thanks for your efforts