Question: monet2photo training loss about pytorch-cyclegan-and-pix2pix HOT 5 CLOSED

I"m not sure how representative this is, but here's my final loss for the discriminators and generators.
There's visible oscillation from about 20 epochs to 100 epochs for Generator B as well as D_B.

Once the learning rate started to get lowered at epoch 100, G_B loss slowly increased and G_A seemed to converge in the .35 to .40 range. Both Discriminators stopped oscillating and gradually got lower.

Perhaps this will be useful to someone doing the same. I used instance norm. Perhaps I should have used batch norm since I was running batch_size = 8

from pytorch-cyclegan-and-pix2pix.

The losses are not so interpretable as G and D are optimizing a minimax game. The plots you posted here looks quite typical to me (except the spike). I will mainly focus on the quality of images.

from pytorch-cyclegan-and-pix2pix.

@filmo : Have you find the reason why the spike in the loss and how to solve it?

from pytorch-cyclegan-and-pix2pix.

No, I didn't end up exploring it further.

from pytorch-cyclegan-and-pix2pix.

Thanks. Let follows the issue how it will be solved
#807

from pytorch-cyclegan-and-pix2pix.