An introduction to Generative Adversarial Networks

This is the code that we used to generate our GAN 1D Gaussian approximation. For more information see our blog post: http://blog.aylien.com/introduction-generative-adversarial-networks-code-tensorflow.

Installing dependencies

Written for Python 2.7.x.

For the Python dependencies, first install the requirements file:

$ pip install -r requirements.txt

You should then install TensorFlow 0.12, see: https://www.tensorflow.org/get_started/os_setup#pip_installation.

If you want to also generate the animations, you need to also make sure that ffmpeg is installed and on your path.

Training

For a full list of parameters, run:

$ python gan.py --help

To run without minibatch discrimination (and plot the resulting distributions):

$ python gan.py

To run with minibatch discrimination (and plot the resulting distributions):

$ python gan.py --minibatch True