Checkpoint images examples? #103
Description
I was wondering if anybody could provide some references for how should checkpoint images look after a certain number of iterations. I am using colab for training and it is going quite slow. I am on ~11500 iteration. I did a few tests and the model hasn't learned what I am training it to learn so far. I do not want to be impatient, but it would be very helpful if I could compare my training images to training images from other projects so I could get an idea if the training is going well and it needs more time or if I need to improve my dataset/configuration parameters.
Also, I couldn't find an explanation of what do those checkpoint grid images actually represent?
attaching current training image for a2b (11500 iteration from a total of 1000 images database, x250 in each folder [train A,B + test A,B]). Domain A are images of men, Domain B are images of men with muscles.
