A possible bug in meldataset.py

yl4579 / StarGANv2-VC

StarGANv2-VC: A Diverse, Unsupervised, Non-parallel Framework for Natural-Sounding Voice Conversion

MIT License

479 stars 107 forks source link

A possible bug in meldataset.py #41

Closed Charlottecuc closed 2 years ago

Charlottecuc commented 2 years ago

Hi. It seems that you use torchaudio.transforms.MelSpectrogram to extract melspec without passing "sample_rate" to this function, so the function will still use the default param, which is 16khz. Does it matter..?

Charlottecuc commented 2 years ago

https://github.com/yl4579/StarGANv2-VC/issues/10#issuecomment-920453833