2D-learn is a library for machine learning research.
classes and functions for being used on tensorflow
- cuda: GPU implementation of common parallel patterns
-
required packages for tf-lib: Anaconda 3, TensorFlow, CUDA 7.0, Eigen 3.2
-
clone the repository:
mkdir ~/libraries
cd ~/libraries
git clone https://github.com/danmar3/2dlearn-lib.git twodlearn -
add the following environmental variables to your bashrc:
export CUDA_HOME="path to cuda toolkit"
export EIGEN_HOME="path to eigen installation"
export TWODLEARN_HOME="/home/username/libraries/" -
add TWODLEARN_HOME to the python path:
export PYTHONPATH="$TWODLEARN_HOME:$PYTHONPATH" -
Compile the tf-lib operations (Not needed for mat-lib)
cd tf_lib/ops/kernels
make
Daniel L. Marino (marinodl@vcu.edu)
Modern Heuristics Research Group (MHRG)
Virginia Commonwealth University (VCU), Richmond, VA
http://www.people.vcu.edu/~mmanic/