Refactor autoregressive model tests

[pstjohn]

Moves shared auto-regressive model tests to modeling_common, making it easier to add new auto-regressive models

Summary by CodeRabbit

• Tests

  • Added generation tests with KV-cache support across model architectures (ESM2, LLaMA3, Mixtral), including batched and beam search variants.
  • Expanded test infrastructure to support flexible configuration overrides in quantized model tests.

• Bug Fixes

  • Improved numerical stability in Mixtral by adjusting tensor dtype handling for expert routing operations.

• Refactor

  • Consolidated generation test implementations across models into shared base test class to reduce duplication.

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• ✅ Review completed - (🔄 Check again to review again)

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[jomitchellnv]

lgtm