run_async's response_stream is a generator and not an async generator

[jonchun]

Note: Running python3.12 on Ubuntu (WSL)

Problem:
Attempting to use the run_async method gives me the following error:

 line 208, in run_async
    async for partial_response in response_stream:
TypeError: 'async for' requires an object with __aiter__ method, got generator

After digging around for a bit, here's what I found in the instructor documentation:
https://python.useinstructor.com/concepts/partial/?h=partial#understanding-partial-responses

extraction_stream = client.chat.completions.create_partial(
    model="gpt-4",
    response_model=MeetingInfo,
    messages=[
        {
            "role": "user",
            "content": f"Get the information about the meeting and the users {text_block}",
        },
    ],
    stream=True,
)
console = Console()

for extraction in extraction_stream:
    obj = extraction.model_dump()
    console.clear()
    console.print(obj)

It's clear from the docs that

When specifying a create_partial and setting stream=True, the response from instructor becomes a Generator[T]. As the generator yields results, you can iterate over these incremental updates. The last value yielded by the generator represents the completed extraction!

this is supposed to return a Generator rather than an AsyncGenerator. The async for needs to be used OUTSIDE of the agent with the above fixes:

e.g.

async for partial in agent.run_async(user_input):
    console.clear()
    console.print(partial )

[KennyVaneetvelde]

I actually opened a new issue: #76

Reason being that the changes from the PR broke other implementations so I reverted it and made a new ticket to better support it.

That being said, your error likely was not a true error but happened due to using the wrong client

You actually should have used openai.AsyncOpenAI as in the examples