Avoid incorrect generations for KV caches containing more than sliding_window tokens

[TimFelixBeyer]

What does this PR do?

Gemma3 generates incoherent output when manually calling forward with an instance of HybridCache which contains more than sliding_window tokens of content.

This is because the call to past_key_values.get_seq_len() always returns the sequence length as measured by the cache of the very first layer. Because this is a local attention layer, its sequence_length never extends beyond config.sliding_window. This leads to an incorrect computation of cache_position and incoherent generations down the line.

To fix it you can simply provide the correct cache_position manually.
This behavior is impossible to fix without changing get_seq_len of HybridCache, so I propose to simply raise an informative error message for now.

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?

@ArthurZucker

[github-actions]

Hi 👋, thank you for opening this pull request! The pull request is converted to draft by default. The CI will be paused while the PR is in draft mode. When it is ready for review, please click the Ready for review button (at the bottom of the PR page). This will assign reviewers and trigger CI.

[Rocketknight1]

cc @gante as well!

[ArthurZucker]

sounds good, we are gonna adress this with the cache refactor! #38077

[manueldeprada]

This is being fixed more generally in the cache refactor, yes. Thanks for the pointer! I will add a test for it. @TimFelixBeyer do you have a simple test snippet at hand?

I think it is better to wait for the general fix rather than adding a safeguard just for Gemma.

[manueldeprada]

As announced, this was fixed with cache refactors. PR can be closed!

[ArthurZucker]

okay, closing