Adliyan

Retraining the model with 16k sampling rate data

6

I modified the sampling rate parameter of preprocess.yaml to 16k, and retrained fastspeech2 and hifigan with data with a sampling rate of 16k, but the resultant synthesized speech was very...