The return tensor of CompositeGenerator don't match the tensor in train.py

[donghaoye]

network.py line 199:
return img_final, flow, weight, img_raw, img_feat, flow_feat, img_fg_feat

train.py line 102: fake_B, fake_B_raw, flow, weight, real_A, real_Bp, fake_B_last = modelG(input_A, input_B, inst_A, fake_B_last)

[tcwang0509]

Why do they need to match? modelG calls the generator network, but is not the network itself.

[donghaoye]

the 2nd return variable of modelG is flow, but you write it as fake_B_raw.

network.py line 199: return img_final, flow, weight, img_raw, img_feat, flow_feat, img_fg_feat

train.py line 102: fake_B, fake_B_raw, flow, weight, real_A, real_Bp, fake_B_last = modelG(input_A, input_B, inst_A, fake_B_last)

[donghaoye]

Sorry. I get it.