Repository to create the COG image for the Swiss-German SOTA Whisper model.
Whisper4SG-SRG-V2-Full-MC-DE-SG-Corpus-V2 on Hugging Face
- Run
download_model.sh
- Predict with
cog predict -i audio_file=@<PATH TO FILE>
- Run
download_model.sh
- Run
cog push <YOUR REPOSITORY ON REPLICATE>
(best to read the guide on Replicate)