It is suggested that Markitdown [audio-transcription] should be able to utilize the local Whisper model to transcribe MP3 and MP4 files

[zhangweildlh]

It is suggested that Markitdown [audio-transcription] should be able to utilize the local Whisper model to transcribe MP3 and MP4 files
I am unable to connect and use the OpenAI Whisper API for MP3 and MP4 transcription. I can only install ffmpeg, torch, and openai-whisper through the following commands and perform transcription locally. However, Markdown does not support invoking the local Whisper model for transcription. I hope this feature will be supported in future versions.

• sudo apt install ffmpeg
• pip install 'markitdown[pptx,docx,xlsx,xls,pdf,audio-transcription]'
• pip install torch --index-url https://download.pytorch.org/whl/cpu
• pip install openai-whisper
• pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu