xiawenxing

Results 1 issues of xiawenxing

Different result in train and eval and synthesize mode

I trained my model and run eval.py regularly. Then I open tensorboard to listen to the audio sample in both mode. The audio in train mode is very clear but...