Issues
- 1
NaN losses
#8 opened by DandiC - 2
Reproduce the plain VAE result
#4 opened by yangw1234 - 0
train_size is not effective
#7 opened by kinsumliu - 0
Paper and code difference
#6 opened by rien333 - 1
I want to use the encoder part only as a CNN, what layers should I use and after which one I put the softmax layer?
#5 opened by fernande2000 - 0
Reconstruct from existing image
#2 opened by galsasson - 1
UpSampling2dLayer+Conv2d or DeConv2d?
#1 opened by LynnHo