Skip to content

Instantly share code, notes, and snippets.

Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save AlexJoz/e84eb137bc4cb9389b22e92ef8f03153 to your computer and use it in GitHub Desktop.
Save AlexJoz/e84eb137bc4cb9389b22e92ef8f03153 to your computer and use it in GitHub Desktop.
tensorflow 0.9 with gpu support for py3 on aws ec2
###
# !!! THIS IS NOT A BASH SCRIPT !!!
###
# named .sh just so Github does correct syntax highlighting
# Inspired by https://gist.github.com/AlexJoz/1670baf0b32573ca7923
#
# This setup is available as a public AMI in US-East(N. Virginia): ami-60da5a77
#
# Tensorflow 0.9 with gpu support (installing cuda drivers and cudnn lib) for python3 with numpy, matplotlib and sklearn installed
sudo apt-get update
sudo apt-get install -y build-essential git python-pip libfreetype6-dev libxft-dev libncurses-dev libopenblas-dev gfortran python3-matplotlib libblas-dev liblapack-dev libatlas-base-dev python3-dev linux-headers-generic linux-image-extra-virtual unzip python3-numpy swig python3-pandas python-sklearn unzip python3-pip
# CUDA
wget http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1410/x86_64/cuda-repo-ubuntu1410_7.0-28_amd64.deb
sudo dpkg -i cuda-repo-ubuntu1410_7.0-28_amd64.deb && rm cuda-repo-ubuntu1410_7.0-28_amd64.deb
sudo apt-get update
sudo apt-get install -y cuda
# cuDNN
# You get the CUDNN_URL by logging into your nivida account and downloading cuDNN
# https://developer.nvidia.com/rdp/cudnn-archive (cudnn 7.0)
# upload it to instance
tar -zxf cudnn-7.0-linux-x64-v4.0-prod.tgz && rm cudnn-7.0-linux-x64-v4.0-prod.tgz
sudo cp ./cuda/lib64/* /usr/local/cuda/lib64/
sudo cp ./cuda/include/* /usr/local/cuda/include/
sudo reboot
echo "export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib64" >> ~/.bashrc
echo "export CUDA_HOME=/usr/local/cuda" >> ~/.bashrc
echo "export PATH=\$PATH:/usr/local/cuda/bin" >> ~/.bashrc
# BAZEL
# Go to https://github.com/bazelbuild/bazel/releases and download the latest bazel. I use 0.3.0 here.
sudo add-apt-repository ppa:webupd8team/java
sudo apt-get update
sudo apt-get install oracle-java8-installer
echo "export JAVA_HOME=/usr/lib/jvm/java-8-oracle" >> ~/.bashrc
. ~/.bashrc
cd
wget https://github.com/bazelbuild/bazel/releases/download/0.3.0/bazel-0.3.0-installer-linux-x86_64.sh
chmod +x bazel-0.3.0-installer-linux-x86_64.sh
sudo ./bazel-0.3.0-installer-linux-x86_64.sh
# TENSORFLOW
git clone --recurse-submodules https://github.com/tensorflow/tensorflow
cd tensorflow
TF_UNOFFICIAL_SETTING=1 ./configure
#All of the questions can be answered with default except these trhee:
```
Python directory: /usr/bin/python3
Do you wish to build TensorFlow with GPU support? [y/N] y
Please specify a list of comma-separated Cuda compute capabilities you want to build with.
You can find the compute capability of your device at: https://developer.nvidia.com/cuda-gpus.
Please note that each additional compute capability significantly increases your build time and binary size.
[Default is: "3.5,5.2"]: 3.0
```
bazel build -c opt --config=cuda //tensorflow/tools/pip_package:build_pip_package
bazel-bin/tensorflow/tools/pip_package/build_pip_package /tmp/pip
sudo pip3 install --upgrade /tmp/pip/tensorflow-*.whl
# REMOVE installers
sudo rm -r cuda
sudo rm bazel-*
sudo rm -r tensorflow
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment