• September 12, 2012 luaT API has been greatly simplified and updated. Please, recompile all your installed packages and read the new luaT documentation.
• Mars 5, 2012 FreeBSD is now supported. GotoBlas/OpenBlas support added. Torch7 discussion group created.
• February 23, 2012 the website is now up! First official release coming soon (we already use Torch7 in real-life applications!).

Torch7 provides a Matlab-like environment for state-of-the-art machine learning algorithms. It is easy to use and provides a very efficient implementation, thanks to an easy and fast scripting language (Lua) and a underlying C implementation.

Torch7 is the official successor of Torch3 and Torch5.

Torch7 is developed at Idiap Research Institute, New York University and NEC Laboratories America.

At this time, Torch7 is maintained by Ronan Collobert (the dictator), Clement Farabet (the chemist) and Koray Kavukcuoglu (Dr Jekyll). We also use the great Qt Lua interface from Leon Bottou.

If you use Torch7 for your work, please cite the following paper:

R. Collobert, K. Kavukcuoglu and C. Farabet. Torch7: A Matlab-like Environment for Machine Learning. In BigLearn, NIPS Workshop, 2011.

Torch7 is distributed under a BSD license. It runs on Linux, FreeBSD and MacOS X. With few tweaks, it should also run on Windows (we plan to support it soon).

See the installation manual for download and installation details. We do not provide binaries yet.

Torch7 git repository is hosted on Github.

We are pleased to share useful contributor packages through the torch-pkg system. See the list of available packages for more details.

Torch7 manual is available online, but also in the source packages.

Also, we have a Torch7 discussion group where users can post questions and share ideas.