[bug]Unawaited Coroutine Warning in openinference: AsyncAPIResponse.parse

[SiyuanQi]

A RuntimeWarning is raised during the execution of the program, indicating that a coroutine was never awaited. Specifically, the error occurs in the file openinference/instrumentation/openai/_request.py at line 191. The warning message is:
RuntimeWarning: coroutine 'AsyncAPIResponse.parse' was never awaited.

.../lib/python3.12/site-packages/openinference/instrumentation/openai/_request.py:191: RuntimeWarning: coroutine 
'AsyncAPIResponse.parse' was never awaited
  response.parse()
RuntimeWarning: Enable tracemalloc to get the object allocation traceback
.../lib/python3.12/site-packages/openinference/instrumentation/openai/_request.py:250: RuntimeWarning: coroutine 
'AsyncAPIResponse.parse' was never awaited
  _finish_tracing(
RuntimeWarning: Enable tracemalloc to get the object allocation traceback

Environment:

• OS: macOS 14.6.1
• Version: Python 3.12 (Anaconda environment)
• openinference-instrumentation 0.1.18
• openinference-instrumentation-groq 0.1.4
• openinference-instrumentation-openai 0.1.17
• openinference-semantic-conventions 0.1.12

[axiomofjoy]

Thanks @SiyuanQi. Do you have a code snippet to reproduce the issue? Or can you let us know which interface you are using when you hit the issue?

[SiyuanQi]

Sure! Below is a code snippet that reproduces the issue:

import os
import dotenv
import asyncio

from openai import AsyncClient

from phoenix.otel import register
from openinference.instrumentation.openai import OpenAIInstrumentor


dotenv.load_dotenv("../.env.local")
tracer_provider = register(
    project_name="openinference-test",
    endpoint=os.getenv("PHOENIX_COLLECTOR_ENDPOINT"),
)
OpenAIInstrumentor().instrument(tracer_provider=tracer_provider)


async def run_tts():
    client = AsyncClient(api_key=os.getenv("OPENAI_API_KEY"))
    async with client.audio.speech.with_streaming_response.create(
        input="Hello",
        model="tts-1",
        voice="alloy",
    ) as audio_stream:
        async for data in audio_stream.iter_bytes():
            pass


if __name__ == "__main__":
    asyncio.run(run_tts())

[axiomofjoy]

Thanks @SiyuanQi, we'll take a look.