This is my fork of the cuda-convnet convolutional neural network
implementation written by Alex Krizhevsky.
cuda-convnet has quite extensive documentation itself. Find the
MAIN DOCUMENTATION HERE <http://code.google.com/p/cuda-convnet/>_.
Update: A newer version, cuda-convnet 2 <https://code.google.com/p/cuda-convnet2/>_, has been released by
Alex. This fork is still based on the original cuda-convnet.
This document will only describe the small differences between
cuda-convnet as hosted on Google Code and this version.
Dropout is a relatively new regularization technique for neural
networks. See the Improving neural networks by preventing co-adaptation of feature detectors <http://arxiv.org/abs/1207.0580>_
and Improving Neural Networks with Dropout <http://www.cs.toronto.edu/~nitish/msc_thesis.pdf>_ papers for
details.
To set a dropout rate for one of our layers, we use the dropout
parameter in our model's layer-params configuration file. For
example, we could use dropout for the last layer in the CIFAR example
by modifying the section for the fc10 layer to look like so::
[fc10] epsW=0.001 epsB=0.002
dropout=0.5
In practice, you'll probably also want to double the number of
outputs in that layer.
An environment variable CONVNET_RANDOM_SEED, if set, will be used
to set the CURAND library's random seed. This is important in order
to get reproducable results.
The build configuration and code has been updated to work with CUDA
via CMake. Run cmake . and then make. If you have an alternative
BLAS library just set it with for example cmake -DBLAS_LIBRARIES=/usr/lib/libcblas.so ..