OpenAIWhisperParser : You tried to access openai.Audio, but this is no longer supported in openai>=1.0.0

[lucasseidor33]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

try:                
            print(linkYT)
            local = False    
            # save_dir = "~/Downloads/YouTube"
            # urls = ["https://www.youtube.com/watch?v=oF0Or9XxQJQ&ab_channel=TodoNoticias"]    
            video_name = linkYT.split("=")[-1]
            save_dir = './uploads/'
            video_path = os.path.join(save_dir, video_name)
            if not os.path.exists(video_path):
                os.makedirs(video_path)
            urls = [linkYT]
            SUPABASE_URL = config('SUPABASE_BASE_URL')
            SUPABASE_SERVICE_KEY = config('SUPABASE_SERVICE_ROLE_KEY')
            supabase: Client = create_client(SUPABASE_URL, SUPABASE_SERVICE_KEY)
            if local:
                print('local')
            else:
                # loader = GenericLoader(YoutubeAudioLoader(urls, save_dir), OpenAIWhisperParser())
                # docs = loader.load()    
                loader = GenericLoader(YoutubeAudioLoader(urls, video_path), OpenAIWhisperParser())
                docs = loader.load()
            rawDocs = loader.load()
            print('rawDocs loaded')
            text_splitter = RecursiveCharacterTextSplitter(
                # Set a really small chunk size, just to show.
                chunk_size = 1000,
                chunk_overlap  = 200,
            )

Description

when I try to transcribe my video to text with Python via OpenAIWhisperParser method I get the following error:

You tried to access openai.Audio, but this is no longer supported in openai>=1.0.0 - see the README at https://github.com/openai/openai-python for the API.

You can run openai migrate to automatically upgrade your codebase to use the 1.0.0 interface.

Alternatively, you can pin your installation to the old version, e.g. pip install openai==0.28

My version of openai is 1.2.1. Any help is welcome!

System Info

python version is 3.11.3, langchain==0.0.333
in windows

[dosubot[bot]]

🤖

Hey @lucasseidor33! Great to see you back here seeking adventures in the code jungle 🌴

Based on the error message you're seeing, it appears that the OpenAI API has changed in version 1.0.0 and no longer supports the openai.Audio class. This is causing the error when the OpenAIWhisperParser class in your code tries to use openai.Audio to transcribe audio.

To resolve this issue, you should use the openai.Audio.transcriptions.create method instead of openai.Audio.transcribe when using OpenAI version 1.0.0 or later. Here's how you can modify your code:

# Transcribe
print(f"Transcribing part {split_number+1}!")
attempts = 0
while attempts < 3:
    try:
        transcript = client.audio.transcriptions.create(
            model="whisper-1", file=file_obj
        )
        break
    except Exception as e:
        attempts += 1
        print(f"Attempt {attempts} failed. Exception: {str(e)}")
        time.sleep(5)
else:
    print("Failed to transcribe after 3 attempts.")
    continue

This code tries to transcribe the audio using the openai.Audio.transcriptions.create method. If it fails, it retries up to 3 times before giving up.

This change should resolve the error you're encountering. If you have any other questions or run into any other issues, please let me know!

For more information about the OpenAIWhisperParser class and how it works, you can refer to the LangChain codebase.

Sources

• Issue: openai.ChatCompletion is no longer supported in openai>=1.0.0
• libs/langchain/langchain/document_loaders/parsers/audio.py

About Dosu

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.

[eponce92]

did you figured it out ?

[Martijnhofstra]

Not at all, have you fixed it by downgrading? V2 is out, so I am a bit confused as well. I hope I remember by the time I fixed it.

[sglebs]

This works:

transcription = openai.audio.transcriptions.create(model=audio_model_name, file=file_name, **extra_args)

Unfortunately I have not been able to find the equivalent of "atranscribe" (the async/await equivalent from the old API). This makes the call abobe synchronous/blocking (super bad).

[sglebs]

Regarding async, I can see that _module_client.py has this: audio: resources.Audio = AudioProxy().__as_proxied__()

but there does not seem to be any asyncaudio equivalent, unfortunately. I can see a resources.AsyncAudio and I can see its transcriptions (an AsyncTranscriptionsobject/class) supports an async def create. But I can't get to it.