OpenAI LLM Generators for Haystack 2.0

[ZanSara]

See the proposal: #5540

---

OpenAI chat models like ChatGPT and GPT4 can be queried through an identical API, therefore we will have a single component for both GPT4 and ChatGPT.

We can either call it OpenAIGenerator and make it support also GPT3 models, or call it GPT4Generator and make an alias for it called ChatGPTGenerator.

Draft API for GPT4Generator:

@component
class GPT4Generator:

    def __init__(self, streaming_callback: Optional[Callable] = None, ... gpt3.5/4 parameters ...):
        ...

    @component.output_types(replies=List[List[str]])
    def run(self, prompts: List[str], ... chatgpt specific params...):
        ...
        return {'replies': [...], 'metadata': [...]}

Note how the component takes a list of prompts and LLM parameters only, but no variables nor templates, and returns only strings. This is because input rendering and output parsing are delegated to PromptBuilder.

In order to support token streaming, we make this component accept a callback in __init__, and that callback will be called every time a new chunk of the streamed response is received.