search

tomchang25 / whisper-auto-transcribe

Auto transcribe tool based on whisper
MIT License
205 stars 14 forks source link

What'are the differences between transcribe models? #42

Closed mxgc closed 1 year ago

mxgc commented 1 year ago

In cli.py, three model types are listed whisper, whisper_timestams, stable_whisper. Could you elaborate the difference between them? @tomchang25

Thanks for making this tool available. I used it to transcribe my entire audiobook library and it worked great!

tomchang25 commented 1 year ago

Whisper: The original version of OpenAI's language model.

Whisper with timestamps: A GitHub project that adds VAD (Voice Activity Detection) and other improvements. In my opinion, this is the most stable version.

Stable Whisper: Another open-source project like Whisper with timestamps, but it has its own format. It is powerful, but not as stable in my opinion. It may default back to Whisper timestamps in the next version

mxgc commented 1 year ago

Great! Thanks for the clarification.