In this work we used two LSTM layers, followed by a reshape layer, then a dense-layer and finally the output layer. PyTorch implementation of VQ-VAE + ...
確定! 回上一頁