The Numerics of GANs

This repository contains the code to reproduce the core results from the paper The Numerics of GANs.

To cite this work, please use

@INPROCEEDINGS{Mescheder2017NIPS,
  author = {Lars Mescheder and Sebastian Nowozin and Andreas Geiger},
  title = {The Numerics of GANs},
  booktitle = {Advances in neural information processing systems},
  year = {2017}
}

You can find further details on our project page.

Dependencies

This project uses Python 3.5.2. Before running the code, you have to install

• Tensorflow 1.0
• Numpy
• Scipy
• Matplotlib
• tqdm

These dependencies can be installed using pip by running

pip install tensorflow-gpu numpy scipy matplotlib tqdm

Instructions

First download cifar-10 and celebA and put them into the ./data directory (in seperate folders). The experiments folder contains scripts for starting the different experiments.