### Requirements
librosa
torch
torchaudio
soundfile
numpy
speechbrain
wget
torchmetrics