PyTorch-GANs

PyTorch implementations of various types of Generative Adversarial Networks (GAN). Each folder contains its own implementation.

Generative Adversarial Networks:

• ConditionalGAN (cGAN)
• CycleGAN
• DCGAN
  • DCGAN Original Architecture for 64 x 64 images
  • DCGAN_32 for generating 32 x 32 images
• LSGAN
  • LSGAN Original Architecture for 112 x 112 images
  • LSGAN_28 Original Architecture for 28 x 28 images
• StarGAN_Digits for generating 32 x 32 digits
• TinyGAN A small GAN network
• VanillaGAN
• WassersteinGAN-GP Wasserstein GAN with Gradient Penalty
• WassersteinGAN