nnaisense / conditional-style-transfer Goto Github PK

Could you provide a simple sample datasets or the structure of directory ?
Look forward to your reply.

I have run your code using the comand python test.py --checkpoints_dir=./samples/models --name GanAuxPretrained --model gan_aux --netG=resnet_residual --netD=disc_noisy --gpu_ids=0 --num_style_samples=1 --loadSize=512 --fineSize=512 --knn=5 --peer_reg=bidir --epoch=200 --content_folder=./samples/data/content_imgs --style_folder=./samples/data/style_imgs --output_folder=output.
Then I got the results with strange color.I just have no idea why my results is so bizarre.

When I was training the model, I had this problem:
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [256, 256, 3, 3]] is at version 2; expected version 1 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True).
How to solve this problem?
Look forward to your reply.

Hello, I want to use your code to train my own dataset, but I am not sure about the meaning of trainA, trainB, valA, valB, etc in the dataroot, and where should my dataset be placed?