malzantot / Pytorch-conditional-GANs

Implementation of Conditional Generative Adversarial Networks in PyTorch

Conditional Deep Convolutional Generative Adversarial Network

Conditional Generation of MNIST images using conditional DC-GAN in PyTorch.

Based on the following papers:

• Conditional Generative Adversarial Nets
• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks

Implementation inspired by the PyTorch examples implementation of DCGAN.

Sample Results

Example of sampling results shown below. Each row is conditioned on a different digit label:

Usage

python conditional_dcgan.py --cuda --save_dir=models --samples_dir=samples --epochs=25

Questions and comments:

Feel free to reach to me at malzantot [at] ucla [dot] edu for any questions or comments.