This project demonstrates a GAN built with PyTorch, using a subset of 5000 CelebA images. It leverages Wasserstein GAN with Gradient Penalty (WGAN-GP) for facial image generation. The provided models ...