about err_d and err_g_adv

[theSha1do1w]

thx for u source code.
and when i train my own dataset,I found that err_d and err_g_adv is unchanged.
both of them are from Dnet.
How does your loss function change during training