Jun-Yan Zhu

I am not sure if it is related to our codebase. If you are not familiar with Lua, you may want to use our PyTorch [version](https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix) with similar or better...

For the complex transformation, I would recommend a small lambda_cycle (e.g. 1 or 2). Could you show some your results as well as loss plots? How big are your G...

We haven't tried it. It is worth trying. I can imagine there might be overfitting issue as the discriminator only needs to memorize one image.

It's hard to train a GAN with only one target image. One thing you can try is to use a smaller patch size. Let's assume that your image is 512x512....

How much memory does your GPU have?

It takes 3.2-3.9 GB to run the test code on my GTX 1080. You can resize input images to 256x256 and see if you can run the models.

We obtained the results with our PyTorch [version](https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix). We will add a training script and more details in the PyTorch repo soon.

@taesung89

Not sure if it is related to this repo. Could you run other GPU jobs with Torch?

have you downloaded the test data?