Open CharlesNord opened 4 years ago
I would like also to say thank you for the nice-written code.
One question regarding the training procedure: which gan type is preferable? LSGAN or WGAN? In the training loop there is no gradient penalty or weight clipping, so probably training WGAN is not a stable process in this configuration?
Hello, I really appreciate your elegant implementation of the gated convolution and the coarse to fine structure. I am very curious about how long does it take to get the comparable result as the original paper?