https://github.com/dragen1860/gan

Pytorch Implementations for popular GAN algorithms.
https://github.com/dragen1860/gan

Last synced: 5 months ago
JSON representation

Pytorch Implementations for popular GAN algorithms.

• Host: GitHub
• URL: https://github.com/dragen1860/gan
• Owner: dragen1860
• Created: 2018-02-18T12:28:40.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2018-02-21T03:03:18.000Z (over 7 years ago)
• Last Synced: 2025-04-01T02:52:10.147Z (7 months ago)
• Language: Python
• Homepage:
• Size: 1.54 MB
• Stars: 5
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# GAN

## 50 lines of GAN

It implements a simple G and D network, where both G and D network consist of 3 fully connected layers. The training data is generated from gaussian distribution whose mean = 4 and variance = 1.25.

Here is the training curve which indicates how generated data converge to true guassian distribution. 

![50lines](res/50.gif)

## WGAN-GP

Wasserstein GAN with gradient penalty

![wgan-gp](res/wgan-gp.gif)