search

auspicious3000 / autovc

AutoVC: Zero-Shot Voice Style Transfer with Only Autoencoder Loss
https://arxiv.org/abs/1905.05879
MIT License
987 stars 205 forks source link

The process of converting spectrograms into audio is very slow #125

Closed idaoi closed 1 month ago

idaoi commented 1 month ago

The process of converting the spectrogram output by the model to audio using wavegen is very slow. In the example of vocoder.ipynb, each audio segment requires 20 minutes of conversion time. Is this normal? My graphics card is RTX2060 with CUDA.