about err_d and err_g_adv

samet-akcay / skip-ganomaly

Source code for Skip-GANomaly paper

MIT License

192 stars 64 forks source link

about err_d and err_g_adv #7

Open theSha1do1w opened 5 years ago

theSha1do1w commented 5 years ago

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