Improve gguf tensor processing

[VladOS95-cyber]

What does this PR do?

Adding more and more architectures with custom weights conversion is making tensor processing system bigger, less flexible and readable.
Improve GGUF tensor processing system in modeling_gguf_pytorch_utils.py by adding TensorProcessor class for general cases without custom weights changes and all necessary separated tensor processor classes based on architecture with custom process logic.

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: Community contribution: Adding GGUF support for more architectures #33260
• 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?

Regarding the task @SunMarc @LysandreJik @ArthurZucker .

[VladOS95-cyber]

Hey @SunMarc! I made some refactoring changes regarding tensor GGUF processing. Please, take a look. What do you think?

[VladOS95-cyber]

Hey @SunMarc! Just a kind ping, does it make sense to have these changes?

[SunMarc]

Thanks for making tensor processing cleaner. I think it makes sense to do that as it will be easier to read the code. Left a few minor comments. Can you have a second look @Isotr0py ?

[VladOS95-cyber]

Hey @SunMarc , @Isotr0py! I made some refactoring based on your comments, please, take a look again.

[Isotr0py]

Overall LGTM! Thanks for improving this!

[VladOS95-cyber]

Hey @SunMarc! Do you have any further comments or it looks good to you?

[VladOS95-cyber]

Hey @SunMarc! Just a kind reminder in order to not loose this PR.

[SunMarc]

Thanks for the PR @VladOS95-cyber as always ! We were away for a week, sorry for the delay. LGTM, just a nit

[SunMarc]

maybe add a comment above to say that you modified name to None in order to skip the processing done below as we already did it above with processor.process. Or we can refactor a bit to also include what we have below to be included in processor.process

[LysandreJik]

I very much like this split! Thanks @VladOS95-cyber!

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[LysandreJik]

Feel free to merge when happy @SunMarc