Jupyter Notebook Setup

# Clone the repogit clone https://github.com/dennybritz/rnn-tutorial-lstmcd rnn-tutorial-lstm# Create a new virtual environment (optional, but recommended)virtualenv venvsource venv/bin/activate# Install requirementspip install -r requirements.txt# Start the notebook serverjupyter notebook

Setting up a CUDA-enabled GPU instance on EC2:

# Install build toolssudo apt-get updatesudo apt-get install -y build-essential git python-pip libfreetype6-dev libxft-dev libncurses-dev libopenblas-dev  gfortran python-matplotlib libblas-dev liblapack-dev libatlas-base-dev python-dev python-pydot linux-headers-generic linux-image-extra-virtualsudo pip install -U pip# Install CUDA 7wget http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1410/x86_64/cuda-repo-ubuntu1410_7.0-28_amd64.debsudo dpkg -i cuda-repo-ubuntu1410_7.0-28_amd64.debsudo apt-get updatesudo apt-get install -y cudasudo reboot# Clone the repo and install requirementsgit clone git@github.com:dennybritz/nn-theano.gitcd nn-theanosudo pip install -r requirements.txt# Set Environment variablesexport CUDA_ROOT=/usr/local/cuda-7.0export PATH=$PATH:$CUDA_ROOT/binexport LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CUDA_ROOT/lib64export THEANO_FLAGS=mode=FAST_RUN,device=gpu,floatX=float32# For profiling onlyexport CUDA_LAUNCH_BLOCKING=1# Startup jupyter noteboookjupyter notebook