How to use microsoft/unispeech-sat-large-sd for diarization

#1
by shripadbhat - opened

I would like to use the model, to diarize audio file and get timestamps of each speaker, how can I achieve that ?

I've tried it even in small audio files and always get CUDA oom. I think maybe it's scientifically good but practically not usable.

Sign up or log in to comment