Dmitry Ulyanov

Oh, sorry, I've been very busy lately, the code should be prepared for public, but I will create a branch for it.

See "diversity" branch.

Torch should use OpenMP for parallel computations, the fact that only one core was used tells something went wrong with installation. For me it uses all the cores. Try manually...

This is strange. Convolution implementation uses matrix mul, so neural nets should be parallel as well..

Hm, it can be because of using threads for dataloader. I have no idea how to deal with it.

This is very strange, since it is there in `nn` package. Try to rebuild `nn` and `cunn`.

I'll take a look tomorrow.

Yes, due to pooling layers the width and height should be a multiple of 32.

Hi, try different `image_size` and `style_size`. It will make the textures larger/smaller. Stylizing large images results as you noticed in small details. But what you can do is to downscale...

Looks like script cannot find the images. If you use symlinks make sure they are in absolute paths. Make sure `dataset/train/dummy` actually has something. Try `rm gen/*` to clean cache.