pyewts
torch
torchaudio
transformers
gradio>=4.0.0
numpy
pyannote.audio