Inconvergence in pitch and energy loss

[周晴]

hi，
When using my own data for training, the pitch and energy loss did not converge, and the Mel loss decreased to 1. The sampling rate of my data is 44100. I have modified the sampling rate parameter in the preprocess.yaml file to 44100, and the other parameters have not been modified. After training for 100k, the pitch and energy loss did not converge. The final synthesized speech is also entirely noisy.