A Neural Turing Machine in Torch
A Torch implementation of the Neural Turing Machine model described in this paper by Alex Graves, Greg Wayne and Ivo Danihelka.
This implementation uses an LSTM controller. NTM models with multiple read/write heads are supported.
Requirements
Torch7 (of course), as well as the following libraries:
penlight
nn
optim
nngraph
All the above dependencies can be installed using luarocks. For example:
luarocks install nngraph
Usage
For the copy task:
th tasks/copy.lua
For the associative recall task:
th tasks/recall.lua