Remove openai module dependency from streaming.py

[auxon]

Currently, the openai module is required, even if I'm using another LLM. Why is it required, and can it be removed?

def openai_aiosession(func):
    """Decorator to set openai.aiosession for StreamingResponse."""

    @wraps(func)
    async def wrapper(*args, **kwargs):
        try:
            import openai  # type: ignore
        except ImportError:
            raise ImportError(
                "openai is not installed. Install it with `pip install 'lanarky[openai]'`."
            )

        openai.aiosession.set(aiohttp.ClientSession())
        logger.debug(f"opeanai.aiosession set: {openai.aiosession.get()}")

[ajndkr]

@auxon given the popularity of openai, the library started with supporting first. this is the first issue towards non-openai solutions. do you have a particular LLM provider in mind?

I can then make openai optional similar to langchain.

[auxon]

@ajndkr I'm using Llama2 7b-Chat right now, but may switch models over time.

[ajndkr]

@auxon Right now, the implementation is tightly coupled with langchain's openai integration. So until there's supported for more LLM providers, I dont want to prematurely work towards making openai optional.

if you are interested to contribute towards this, maybe you can propose a plan and we can work towards supported more LLM providers.

[auxon]

@ajndkr I'll take a look but things seem to work fine as is, so it may be low priority.

[ajndkr]

@auxon i've removed the openai_aiosession decorator after openai_v1 migration. see #149

the changes will come into effect with v0.8 release. i have other features/bugfixes in mind. Will take a week or so until i make the new release.

P.S. i checked and openai was always an optional dependency. see: https://github.com/ajndkr/lanarky/blob/90bd4bf2fe3b66b0c708f1ebb6c73006cc372627/pyproject.toml#L19