yzwxx/vae-celebA

Issues

• NaN losses

  #8 opened 5 years ago by DandiC
  1

• Reproduce the plain VAE result

  #4 opened 7 years ago by yangw1234
  2

• train_size is not effective

  #7 opened 6 years ago by kinsumliu
  0

• Paper and code difference

  #6 opened 6 years ago by rien333
  0

• 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 6 years ago by fernande2000
  1

• Reconstruct from existing image

  #2 opened 7 years ago by galsasson
  0

• UpSampling2dLayer+Conv2d or DeConv2d?

  #1 opened 7 years ago by LynnHo
  1