Closed cangwang closed 9 months ago
You can try to tune the hyperparameters
--wadvd 10.0\ # Aversarial loss weight for D
--wadvg 10.0\ # Aversarial loss weight for G
--wcon 1.5\ # Content loss weight
--wgra 3.0\ # Gram loss weight
--wcol 30.0\ # Color loss weight
And consider stop training if the loss start increase
You have use how many epoch to change? Use tune hyperparameters is another function to change ,and we can‘’t feel the source image and fake images direct。
I have run 100 epoch for shikai dataset ,and adv loss is more and more larger。and the result is not real like shikai style and do it wrong to get result?