UserWarning: Using a target size (torch.Size([4, 1, 128, 80])) that is different to the input size (torch.Size([4, 128, 80])). This will likely lead to incorrect results due to broadcasting. Please ensure they have the same size.

auspicious3000 / autovc

AutoVC: Zero-Shot Voice Style Transfer with Only Autoencoder Loss

https://arxiv.org/abs/1905.05879

MIT License

983 stars 207 forks source link

UserWarning: Using a target size (torch.Size([4, 1, 128, 80])) that is different to the input size (torch.Size([4, 128, 80])). This will likely lead to incorrect results due to broadcasting. Please ensure they have the same size. #66

Closed BenjaminChua closed 3 years ago

BenjaminChua commented 3 years ago

To train my own model with the default dataset given in this repo, I have firstly generated spec and then the training metadata. However, while training using the main.py script I get this UserWarning. May I know why?

BenjaminChua commented 3 years ago

I am using PyTorch 1.6, so I think that's the main reason for this issue. Resolved this by commenting/removing lines 199-200 in model_vc.py:

mel_outputs = mel_outputs.unsqueeze(1)
mel_outputs_postnet = mel_outputs_postnet.unsqueeze(1)