This is an implementation of the papers Generative Adversarial Networks and Conditional Generative Adversarial Nets in PyTorch, done for personal enrichment. To get started, follow the setup instructions and then the run instructions.
- Install PyTorch from https://pytorch.org/ - Note: they removed conda support, but you can still install it using pip in a conda environment (that's what I did)
pip install tqdm pillow pandas numpy matplotlib torchsummary
For GAN:
- Adjust the hyperparameters near the top of train.py as needed
python train.py
For Conditional GAN:
- Adjust the hyperparameters near the top of train_conditional.py as needed
python train_conditional.py