Add Pytorch Tensor Parallel support for Mistral

[VladOS95-cyber]

What does this PR do?

This PR uses the torch.distributed.tensor.parallel subpackage to implement Tensor Parallel for Mistral.

Fixes # (issue)

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. Link:Add Tensor Parallel support for ALL models #34789
• [] 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.

[Rocketknight1]

cc @ArthurZucker since I think he was handling the tensor parallelism issue - let me know if I should have pinged someone else instead!

[ArthurZucker]

LGTM thanks for adding! 🤗

[DebarshiChanda]

@VladOS95-cyber @ArthurZucker Hi guys, I had a small doubt. Why aren't the changes in the query_states, key_states and value_states applied to other attention classes (like MistralFlashAttention2, MistralSdpaAttention) as well?

[VladOS95-cyber]

@VladOS95-cyber @ArthurZucker Hi guys, I had a small doubt. Why aren't the changes in the query_states, key_states and value_states applied to other attention classes (like MistralFlashAttention2, MistralSdpaAttention) as well?

Hey @DebarshiChanda! Thanks for noticed that, i think you are right and we should apply the same infer logic for other attention classes. I am going to fix it in another PR #35007