Awesome

News

28 June 2019: We re-implement these GANs by Pytorch 1.1! The old version is here: v0 or in the "v0" directory.
Tensorflow 2 Version

<p align="center"> GANs - Pytorch </p>

Pytorch implementations of DCGAN, LSGAN, WGAN-GP(LP) and DRAGAN.

Exemplar results

Fashion-MNIST

DCGAN	LSGAN	WGAN-GP	DRAGAN
<img src="./pics/fashion-mnist_gan.jpg">	<img src="./pics/fashion-mnist_lsgan.jpg">	<img src="./pics/fashion-mnist_wgan-gp.jpg">	<img src="./pics/fashion-mnist_dragan.jpg">

CelebA

DCGAN	LSGAN
<img src="./pics/celeba_gan.jpg">	<img src="./pics/celeba_lsgan.jpg">
WGAN-GP	DRAGAN
<img src="./pics/celeba_wgan-gp.jpg">	<img src="./pics/celeba_dragan.jpg">
WGAN-LP	DRAGAN-LP
<img src="./pics/celeba_wgan-lp.jpg">	<img src="./pics/celeba_dragan-lp.jpg">

Anime

WGAN-GP	DRAGAN
<img src="./pics/anime_wgan-gp.jpg">	<img src="./pics/anime_dragan.jpg">

Usage

Prerequisites
- PyTorch 1.1
- tensorboardX
- scikit-image, oyaml, tqdm
- Python 3.6
Datasets
- Fashion-MNIST will be automatically downloaded
- CelebA should be prepared by yourself in ./data/img_align_celeba/*.jpg
  - dataset link: Dropbox
  - the above link might be inaccessible, the alternatives are (find "img_align_celeba.zip")
    - Baidu Netdisk or
    - Google Drive
- the Anime dataset should be prepared by yourself in ./data/faces/*.jpg
  - dataset link: https://pan.baidu.com/s/1eSifHcA, password: g5qa
  - reference: https://zhuanlan.zhihu.com/p/24767059

Examples of training

Fashion-MNIST DCGAN

CUDA_VISIBLE_DEVICES=0 python train.py --dataset=fashion_mnist --epoch=25 --adversarial_loss_mode=gan

CelebA DRAGAN

CUDA_VISIBLE_DEVICES=0 python train.py --dataset=celeba --epoch=25 --adversarial_loss_mode=gan --gradient_penalty_mode=1-gp --gradient_penalty_sample_mode=dragan

Anime WGAN-GP

CUDA_VISIBLE_DEVICES=0 python train.py --dataset=anime --epoch=200 --adversarial_loss_mode=wgan --gradient_penalty_mode=1-gp --gradient_penalty_sample_mode=line --n_d=5

see more training exampls in commands.sh

tensorboard for loss visualization

tensorboard --logdir ./output/fashion_mnist_gan/summaries --port 6006