i use the dcgan.py to train dataset with mnist, after 25 epochs the netD`s 
performance pretty well , train acc is around 0.9, but the netG`s loss is a 
little high, and it looks like don`t converge 

below is netD loss and netG loss:
![d_and_g_loss](https://user-images.githubusercontent.com/28485566/45806511-d582e400-bcf3-11e8-9972-462bece22ccb.png)






[ Full content available at: 
https://github.com/apache/incubator-mxnet/issues/12612 ]
This message was relayed via gitbox.apache.org for [email protected]

Reply via email to