tensorboard result: the generated audio of generate.py is 0 seconds
Ahapy opened this issue · comments
Environment: win10 + Python3.7, cuda10.2:
- Copy part of the audio of VCTK / wav48 / to form a wav folder for training
2.New logdir - Modify "SILENCE_THRESHOLD" in the original program to 0.1
tensorboard result: the generated speech of generate.py is 0 seconds