Pytorch implementation of WGAN-GP with a projection operator.

• Modification of link
• igul222/improved_wgan_training
• caogang/wgan-gp
• LayerNorm

• Python >= 3.6
• Pytorch >= v1.0.0
• Numpy
• SciPy
• tensorboardX (installation here). It is very convenient to see costs and results during training with TensorboardX for Pytorch
• TensorFlow for tensorboardX

• gan_train.py: This model is mainly based on GoodGenerator and GoodDiscriminator from Improved Training of Wasserstein GANs. Generator and discriminator are modified for 128x128 (width, height) dataset.

• Usage example, run: python gan_train.py --dataset 'circle' --output 'output_dir'. Then, the output file will be written to ./output_dir folder.

• This is the toy example dataset. You can use "train_toyCircle_3Ch_128.h5" dataset for the training. 
• https://drive.google.com/drive/folders/1eQCZtni4UvilOI4-nQHBhRQyMvTpOizN?usp=sharing

During the implementation of this model, we built a test module to compare the result between original model (Tensorflow) and our model (Pytorch) for every layer we implemented. It is available at compare-tensorflow-pytorch

Results such as costs, generated images (every 200 iters) for tensorboard will be written to ./runs folder.

To display the results to tensorboard, run: tensorboard --logdir runs