WavThruVec_pytorch
WavThruVec_pytorch copied to clipboard
p1an-lin-jung
An implementation of Charactr, Inc's "WavThruVec: Latent speech representation as intermediate features for neural speech synthesis"
![Uploading 微信图片_20240421144204.png…]() Epoch[192/2001],step [202000/211200]:W2V Feat Loss: nan, W2V Feat PostNet Loss: nan,attn binarization loss:nan,Current Learning Rate is 0.100000.Time Used: 142124.986s,Estimated Time Remaining: 6594.002s.
Thank you for putting up this code. I am interested in the txt2vec model (that you said works well in the other issue). Is the training stable? How long does...
Metadata
Owner
Metadata
An implementation of Charactr, Inc's "WavThruVec: Latent speech representation as intermediate features for neural speech synthesis"