Add mean_resizing for every models' resizing_token_embeddings()

[YenFuLin]

What does this PR do?

The function, resizing_token_embeddings(), is defined in model_utils.py, but in some VLMs overrided it and forget adding this new argument, mean_resizing, so that user cannot change mean_resizing value. However, I think this argument shall be flexible to VLMs so that I refine it.

Fixes #35357

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?

@Rocketknight1

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.

[zucchini-nlp]

Great, thanks for passing the mean_resizing

For VLMs I am trying to remove the overriden methods and will merge #34603 soon, probably today. So you might need to rebase after that :)

[YenFuLin]

Ok, thanks for your friendly reminder.

[YenFuLin]

Continue working on #35717