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

About the developer

skaae
203 Stars 48 Forks BSD 3-Clause "New" or "Revised" License 8 Commits 3 Opened issues

Services available

!
?

Need anything else?

Contributors list

# 41,457
Lua
Jupyter...
Shell
deep-le...
8 commits

Torch convolutional GAN

To run the code clone the repository

git clone https://github.com/skaae/torch-gan.git

cd
to the
datasets
subfolder and run
create_dataset.py
. This will create the labeled faces in the wildt dataset. This may take a while depending on your internet connection etc.

Then run

th train_lfw.lua -g 0

where

-g 0
specifies the GPU you want to use. The code will only run on GPU, but you can easily modify to run on CPU by removing the cudnn dependencies.

The code will plot 100 generated images after each epoch. After a 5-10 epochs you should see something that looks like a face.

The code was written by Anders Boesen Lindbo Larsen and Søren Kaae Sønderby. Our code is based on code released with the LAPGAN paper.

Move in latent space

Long latent space movie

Faces

100 epochs

Faces

Dependencies

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.