Fixing Cody Snider's PR for LlamaCPP local support (no server)

[IMJONEZZ]

Ran ruff check . --fix-only

Added import llama_cpp to fix build test

Added install commands for CPU, CUDA, and Metal with link to pypi for other configurations

Added comment explaining model: str

[arnavsinghvi11]

Thanks for adding to the LlamaCPP support @IMJONEZZ ! The changes still fail the tests as llama_cpp is not imported . To fix this, you can follow the try/except Exception structure here
and/or add llama_cpp to the optional dependencies in the pyproject.toml.

It would also be great if you could add documentation for the LlamaCpp LM to provide context as done for the LMs in our documentation here?. maybe can pull from the example in @codysnider 's PR #862 too.