README.md

DCGAN for MNIST digit generation

Tensorflow implementation of a Generative Adverserial Network (GAN) to generate MNIST digits, using the new way to override Model.train_step in keras.

Training vs Generated Images

Install dependencies

Create a Python 3 virtual environment and activate it:

virtualenv -p python3 venv
source ./venv/bin/activate

Next, install the required dependencies:

pip install -r requirements.txt

Train model

Start the model training by running:

python train.py

To track metrics, start Tensorboard

tensorboard --logdir path/to/log/dir

and then go to localhost:6006.