/Whisperer

Batch speech to text using OpenAI's whisper.

Primary LanguageC#

Whisperer lets you generate subtitles for any kind of video/audio files.

It uses OpenAI's whisper, it runs as many instances simultaneously as your GPU's memory allows.

Update: now uses https://github.com/Const-me/Whisper GPU enabled whispercpp, thanks to Kosta for this major speed improvement!

You will need the models from here: https://huggingface.co/ggerganov/whisper.cpp/tree/main
Do NOT use v3 it is not supported yet!

Requires ffmpeg on your path, not included.

scrshot