pytorch-GANs

My items: [Tensorflow version]

The original code address

https://github.com/hwalsuklee/tensorflow-generative-model-collections

Project progress

It's adapted to the cifar10. Details can be reached via email.

Results for mnist

The following results can be reproduced with command:

python main.py --dataset mnist --gan_type <TYPE> --epoch 40 --batch_size 64

Fixed generation

All results are generated from the fixed noise vector.

Name	Epoch 1	Epoch 20	Epoch 40	GIF
GAN
CGAN
VAE
CVAE
WGAN
LSGAN
EBGAN
ACGAN
infoGAN
SAGAN

GANs for label

Name	Epoch 1	Epoch 20	Epoch 40	GIF
CGAN
CVAE
ACGAN
infoGAN

Loss plot

Name	Loss
GAN
CGAN
VAE
CVAE
WGAN
LSGAN
EBGAN
ACGAN
infoGAN
SAGAN

Folder structure

The following shows basic folder structure.

├── main.py # gateway
├── data
│   ├── mnist # mnist data (not included in this repo)
│       ├── t10k-images-idx3-ubyte.gz
│       ├── t10k-labels-idx1-ubyte.gz
│       ├── train-images-idx3-ubyte.gz
│       └── train-labels-idx1-ubyte.gz
│
├── GAN.py # vainilla GAN
├── utils.py # utils
├── models # model files to be saved here
└── results # generation results to be saved here

Development Environment

• Ubuntu 16.04 LTS
• NVIDIA GTX 1080
• cuda 9.0
• Python 3.5.2
• pytorch 0.4.0
• torchvision 0.2.1

Acknowledgements

This implementation has been based on tensorflow-generative-model-collections and tested with Pytorch on Ubuntu 16.04 using GPU.