Lucy Chai

Thanks for pointing it out! It seems that it is not downloading through the `gdown` tool, but you can open the link directly in the browser to download it. I've...

Hi, it looks like I'm using conda 4.7.11, python=3.6.11, tensorflow-gpu=1.14.0, but I'll also attach my full environment spec here from the patch-forensics-data environment (it's messier than the one in the...

I think specifically for that model you might need to change the cbase parameter using `--cbase=16384` (and maybe also `--cmax=512`) to match the number of channel dimensions.

I think it should be possible, but didn't have time to fully investigate before deadlines. It should be similar to patch training for the normal setup but finetuning from the...

We only tried unconditional generation, but it should be compatible with the proper modifications. I suppose you'll need to modify the function `compute_scale_inputs`; in particular the `mapped_scale = G.scale_mapping(scale, None)`....

Hi, unfortunately I haven't tested it on a Windows system. I'm not certain if it would be caused by input path; seems that the UnpairedDataset object takes in a path...