Torch Manual

Torch has lots of libraries and corresponding reference documentation. But this is where you can try to make some sense out of all of it.

The Basics

• Introduction
• Getting Started
• Torch Ecosystem
• Programming in Lua
• Tensors
• Object-Oriented Programming
• Foreign Function Interface

Neural Network Training

• Datasets
• Building your own dataset
• Modules
• Writing your own torch module
• Criterions
• Optimization

Library Reference

• Torch : tensors, class factory, serialization, BLAS ;
• nn : neural network Modules and Criterions;
• optim : SGD, LBFGS and other optimization functions ;
• gnuplot : ploting and data visualization ;
• paths : make directories, concatenate file paths, and other filesystem utilities ;
• image : save, load, crop, scale, warp, translate images and such ;
• trepl : the torch LuaJIT interpreter ;
• cwrap : used for wrapping C/CUDA functions in Lua ;