[Model] Adding Granite model.

[shawntan]

Adds Granite model class to vLLM.

Model will be on Hugging Face once huggingface/transformers#31502 is merged.

[github-actions]

👋 Hi! Thank you for contributing to the vLLM project.
Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which consists a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of default ones by unblocking the steps in your fast-check build on Buildkite UI.

Once the PR is approved and ready to go, please make sure to run full CI as it is required to merge (or just use auto-merge).

To run full CI, you can do one of these:

• Comment /ready on the PR
• Add ready label to the PR
• Enable auto-merge.

🚀

[njhill]

Thanks @shawntan! Could you rebase on the latest main branch, and add a test in https://github.com/vllm-project/vllm/tree/main/tests/models similar to the other model architectures?

[njhill]

@shawntan I have moved this to draft since it currently depends on a future version of the transformers library.

How about including the GraniteConfig class in its own file here in the meantime? Then we can clean that up later once vLLM moves to the necessary transformers version.

[njhill]

@shawntan to avoid duplication / maintenance overhead would it make more sense to just add the optional multipliers to llama.py?

[shawntan]

@njhill HF PR merged.

[njhill]

Thanks @shawntan, looks great.

However, we may need to keep the config class here for the time being, until a new transformers version is released containing it and vLLM moves to that version. Could you reinstate that?

And take it out of draft if it's now ready to be merged?