Need help with lsgan?
Click the “chat” button below for chat support from the developer who created it, or find similar developers for support.

About the developer

208 Stars 49 Forks 31 Commits 3 Opened issues


This is the project for LS-GAN (Loss-Sensitive GAN)

Services available


Need anything else?

Contributors list

# 312,466
31 commits

Loss-Sensitive GAN

Author: Guo-Jun Qi, Date: 1/9/2017, most recent update: 2/8/2017

Questions about the source codes can be directed to Dr. Guo-Jun Qi at [email protected] All copyrights reserved.

Please cite the following paper when referring to the following algorithms (LS-GAN and CLS-GAN)

Guo-Jn Qi. Loss-Sensitive Generative Adversarial Networks on Lipschitz Densities. arXiv:1701.06264 [pdf]

We are keeping updating this repository of source codes, and more results and algorithms will be released soon.

  • We now have a new project generalizing LS-GAN to a more general form, called Generalized LS-GAN (GLS-GAN). It unifies Wasserstein GAN as well as this LS-GAN in an integrated framework, along with a super class of GLS-GANs including various new GAN members to explore. Check the github project at, and the Appendix D of the newly updated prepint

  • We also show an incomplete map of GANs at, which plots the terrority of basic GANs in our view. We will update the map as more GAN territories are found.

Notes on setting the hyperparameters

  • gamma: this is the coefficient for loss-minimization term (the first term in the objective for optimizing L_\theta).
  • lambda: the scale of margin. This controls the desired margins between real and fake samples. However, we found this is a hyperparameter not very sensitive to the generation performance.

LS-GAN (without conditions)

For celebA dataset

1.Setup and download dataset

`mkdir celebA; cd celebA

Download from under the link "Align&Cropped Images".

unzip; cd ..
DATA_ROOT=celebA th data/crop_celebA.lua

2.Training the LS-GAN

DATA_ROOT=celebA dataset=folder th lsgan.lua

For LSUN bedroom dataset

  1. Please download bedroomtrainlmdb from

  2. Prepare the dataset following the instructions below

    1. Install LMDB in your system:
      sudo apt-get install liblmdb-dev
    2. Install torch packages:
      luarocks install lmdb.torch
      luarocks install tds
    3. Once downloading bedroomtrainlmdb, unzip the dataset and put it in a directory
    4. Create an index file : Copy lsunindexgenerator.lua to lsun/train, and run
      cd lsun/train
      DATA_ROOT=. th lsun_index_generator.lua
      Now you should have bedroomtrainlmdbhasheschartensor.t7 in lsun/train
    5. Now return to the parent direcotry of lsun, and you should be ready to run lsgan.lua:
      DATA_ROOT=lsun th lsgan.lua

How to display the generated images

To display images during training and generation, we will use the display package.

  • Install it with:
    luarocks install
  • Then start the server with:
    th -ldisplay.start
  • Open this URL in your browser: http://localhost:8000

For Conditional LS-GAN (CLS-GAN)

  1. Download and prepare datasets

    1. MNIST: run
      torch-rocks install
    2. CIFAR10: run
      th ./data/Cifar10bintoTensor.lua
    3. SVHN: run
      th ./data/svhn.lua
  2. Now you should be able to run clsgan.lua now. Select the dataset you want to use. For example you want to run MNIST to generate handwritten digits according to ten digit classes. Then you can run the following command line

    dataset=mnist th clsgan.lua

For the other parameters you can set, please refer to the script in clsgan.lua.


  1. parts of codes are reused from DCGAN at

  2. the code downloading cifar10 is available at

  3. the code downloading SVHN:

We use cookies. If you continue to browse the site, you agree to the use of cookies. For more information on our use of cookies please see our Privacy Policy.