Integrate new method into OpenAIModel for transcription

[rroblf01]

Since OpenAI supports transcribing audios, it would be interesting to integrate this method into OpenAIModel

Example of Audio Transcription with OpenAIModel

This example shows how to use the OpenAIModel class from the pydantic_ai.models.openai module to transcribe an audio file using the OpenAI API.

Example Code

import asyncio

from pydantic_ai.models.openai import OpenAIModel

# Initialize the OpenAI model with the desired model name
openai_model = OpenAIModel(model_name='whisper-1')


# Asynchronous function to get the transcription of the audio file
async def get_transcription():
    # Open the audio file in binary mode
    with open('/path/to/audio.mp3', 'rb') as audio_file:
        # Call the transcription method of the model
        transcription_text = await openai_model.transcription(audio_file)

    return transcription_text


if __name__ == '__main__':
    # Run the asynchronous function
    asyncio.run(get_transcription())

An example of a response could be

Lorem ipsum dolor sit amet consectetur adipiscing elit congue, ridiculus dapibus porttitor suspendisse torquent netus elementum ornare, eros quisque lacinia taciti imperdiet consequat magna. Lectus ad vel mollis pretium eget commodo donec velit, elementum phasellus libero himenaeos dictumst nascetur nibh penatibus egestas, proin varius condimentum curabitur venenatis ultrices hendrerit. Iaculis etiam fringilla viverra porttitor cras tortor molestie dictum, tristique lacus tincidunt torquent ultrices natoque ad lectus nisl, facilisis phasellus quam nibh consequat quisque pulvinar.