FastSpeech2
FastSpeech2 copied to clipboard
torch.nn.modules.module.ModuleAttributeError: 'FastSpeech2' object has no attribute 'module' how to reduce
File "train.py", line 198, in
when you train a model with multi GPU, state_dict will be saved as model.module.state_dict while model.state_dict when single GPU.
Thank you, I have found this problem and solved it