Implementation of A Style-Based Generator Architecture for Generative Adversarial Networks (https://arxiv.org/abs/1812.04948) in PyTorch
Usage:
python train.py -d {celeba} PATH
I have mixed styles at 4^2 - 8^2 scale. I can't get samples as dramatic as samles in the original paper. I think my model too dependent on 4^2 scale features - it seems like that much of details determined in that scale, so little variations can be acquired after it.