loss does not decrease at all (it rather increases...)

[kylemin]

I used exact same image preprocessing, including cropping and resizing.
However, as training goes on, g_loss increases gradually, and the training and test image look very strange.
(no matter what channel/parameter/optimizer is used)

I wonder how did you guys get the results.