Skip to content

Latest commit

 

History

History
73 lines (53 loc) · 2.81 KB

installation.md

File metadata and controls

73 lines (53 loc) · 2.81 KB

Building Caffe2

This guide builds from source. For alternatives, refer to https://caffe2.ai/docs/getting-started.html

Get latest source from GitHub.

git clone --recursive https://github.com/caffe2/caffe2.git
cd caffe2

Note that you might need to uninstall existing Eigen and pybind11 packages due to compile-time dependencies when building from source. For this reason, Caffe2 uses git submodules to reference external packages in the third_party folder. These are downloaded with the --recursive option.

MacOS X

brew install openblas glog gtest automake protobuf leveled lmdb
mkdir build && cd build
cmake .. -DBLAS=OpenBLAS -DUSE_OPENCV=off
make

Ubuntu

Ubuntu 14.04 LTS
sudo apt-get install libprotobuf-dev protobuf-compiler libatlas-base-dev libgoogle-glog-dev libgtest-dev liblmdb-dev libleveldb-dev libsnappy-dev python-dev python-pip libiomp-dev libopencv-dev libpthread-stubs0-dev cmake
sudo pip install numpy
wget http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1404/x86_64/cuda-repo-ubuntu1404_8.0.44-1_amd64.deb
sudo dpkg -i cuda-repo-ubuntu1404_8.0.44-1_amd64.deb
sudo apt-get update
sudo apt-get install cuda
sudo apt-get install git

CUDNN_URL="http://developer.download.nvidia.com/compute/redist/cudnn/v5.1/cudnn-8.0-linux-x64-v5.1.tgz" &&
curl -fsSL ${CUDNN_URL} -O &&
sudo tar -xzf cudnn-8.0-linux-x64-v5.1.tgz -C /usr/local &&
rm cudnn-8.0-linux-x64-v5.1.tgz &&
sudo ldconfig

mkdir build && cd build
cmake ..
make
Ubuntu 16.04 LTS
sudo apt-get install libprotobuf-dev protobuf-compiler libatlas-base-dev libgoogle-glog-dev libgtest-dev liblmdb-dev libleveldb-dev libsnappy-dev python-dev python-pip libiomp-dev libopencv-dev libpthread-stubs0-dev cmake
sudo pip install numpy
wget http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/cuda-repo-ubuntu1604_8.0.61-1_amd64.deb
sudo dpkg -i cuda-repo-ubuntu1604_8.0.61-1_amd64.deb
sudo apt-get update
sudo apt-get install cuda
sudo apt-get install git

CUDNN_URL="http://developer.download.nvidia.com/compute/redist/cudnn/v5.1/cudnn-8.0-linux-x64-v5.1.tgz" &&
curl -fsSL ${CUDNN_URL} -O &&
sudo tar -xzf cudnn-8.0-linux-x64-v5.1.tgz -C /usr/local &&
rm cudnn-8.0-linux-x64-v5.1.tgz &&
sudo ldconfig

mkdir build && cd build
cmake ..
make

Python support

To use Caffe2 in Python, you need two libraries, future and six.

pip install future six

To run the tutorials, download additional source from GitHub.

git clone --recursive https://github.com/caffe2/tutorials.git caffe2_tutorials
cd caffe2_tutorials

You'll also need jupyter (formerly ipython) notebooks and matplotlib, which can be installed on MacOS X with

brew install matplotlib --with-python3
pip install jupyter