a fork of https://github.com/ANonEntity/WhisperWithVAD (DeepL integration for translation)
adapted to Faster-Whisper with built-in SileroVAD: https://github.com/guillaumekln/faster-whisper
for vocals extraction, replace spleeter with demucs, inspired from: https://github.com/EtienneAb3d/WhisperHallu
several enhancements for Japanese from somewhere on internet
to open the notebook in Google Colab.
- Run the Setup Whisper cell.
- Upload your input audio to either the runtime itself, Google Drive, or a file hosting service with direct download links.
- Set the audio_path and language variables, and then run the Run Whisper cell. (Note: Audio path is set automatically if you use the Upload cell)
- Once it's done, the notebook will automatically download the generated SRT file.
bonus: youtube version (currently not working)