In this part of Homework 3, your will implement several a GAN, and run experiments using the MNIST and CIFAR-10 datasets.
You can download the datasets here (extract them to the data/
directory in the root of this repository). I've provided code for loading and processing this data into minibatches in mnist.py
and cifar10.py
.
Scaffolding for your GAN models is provided in models.py
and utils.py
.
Framework code for training your models is found in mnist.ipynb
(Problem 5) and cifar10.ipynb
(Problem 6). I recommend using Google Colab to execute these notebooks (with a GPU accelerator attached).