Add support for splitting and transcribing large audio files

[mentatbot[bot]]

This update addresses the issue where the app could not handle audio files larger than 25 MB due to Whisper's max input file size limitation. The solution involves splitting audio files greater than 25 MB into smaller chunks, each of which can be transcribed by the API. The results are then combined into a single transcript.

Changes include:

• Added a split_audio_file function to handle splitting of large audio files.
• Modified the transcribe_audio function to process and transcribe each chunk separately and then combine the results.
• Removed the file size check in the main processing flow since the splitting logic now handles it.

This ensures that larger audio files can be processed without hitting the size limit, while also considering API cost and rate limits.

Closes #8

[Bklieger]

@mentatbot Please review this PR. Running the code results in this error when uploading an .m4a file which previously worked: Error code: 400 - {'error': {'message': 'file must be one of the following types: [flac mp3 mp4 mpeg mpga m4a ogg opus wav webm]', 'type': 'invalid_request_error'}}