Add support for llama.cpp llm evaluator

[lbux]

Is your feature request related to a problem? Please describe.
As of now, Haystack's evaluators which extend LLMEvaluator only support OpenAI. I would like for support through llama.cpp to be added for local/offline/"free" evaluation.

Describe the solution you'd like
After implementing LlamaCppChatGenerator (deepset-ai/haystack-core-integrations#723), it is possible to constrain the output to json. As such, we can split up the instructions of the evaluator into ChatMessages (system for instructions/examples, user for input tuples) and then have the result be in the json format expected.

I have a WIP implementation, and I would like feedback on how to handle the integration of llama.cpp into the existing evaluators.

As of now, I am manually calling the Chat Generator in llm_evaluator to ensure it would be possible. I am hard coding the model and generation kwargs.

My two different ideas to resolve this are as follows:

• Allow for passing in of a generator through evaluators and handle logic differently depending on whether api = "openai" or "llama.cpp".
• Allow for passing in model_kwargs and generation_kwargs, discard model kwargs when api = "openai", create LlamaCppChatGenerator with kwargs in llm_evaluator using both kwargs when api = "llama.cpp"

Another idea but I don't think it would be ideal is to just have separate evaluator components for llama.cpp.

If I can have some feedback, I can submit a PR with revised changes in a couple of days.

Describe alternatives you've considered
N/A
Additional context
N/A

[shadeMe]

Thanks for your help with this! Instead of passing generator instances, we can just expand the current approach in the following manner:

• Expose an api_params: Optional[Dict[str, Any]] = None init parameter in LlmEvaluator.
• Add llama_cpp as a new API and use a LlamaCppChatGenerator as a backend when it's chosen.
• Allow users to pass the required API-specific parameters via api_params and, internally, conditionally handle the contents of that dictionary based on the chosen API.