Using latest audio transcribe from openai.

[BirenMer]

Thank you for contributing to LangChain!

• openai bug fix in audio.py: "community: found a bug while testing langchain in another repo found the bug in the main langchian repo as well hence correcting it"

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

1 Skipped Deployment

Name	Status	Preview	Comments	Updated (UTC)
langchain	⬜️ Ignored (Inspect)	Visit Preview		Dec 22, 2024 6:01pm

[BirenMer]

Changing the below line of code :
transcript = openai.Audio.transcribe("whisper-1", file_obj) # type: ignore[attr-defined]

As above functionality is no longer supported by openai.
To make the code more consistent and stable.

The new line is as follows:
transcript = openai.audio.transcriptions.create(
model="whisper-1",
file=file_obj, )

[ccurme]

You should only hit this line if you are using an old (pre-1.0) version of the openai sdk. For these versions the line looks correct:

https://github.com/openai/openai-python/blob/f7ccce126325ea35b6e5224ab954652c97a74896/openai/api_resources/audio.py#L46

Let me know if I've misunderstood something.