saymrwulf/pytorch
1
0
Fork 0
mirror of https://github.com/saymrwulf/pytorch.git synced 2026-05-14 20:57:59 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
pytorch/docs/installation.md
Geunsik Lim 8aa3dab959 doc: update installation.md for third_party packages 
Summary:
PR Description
-----------------

This commit informs the developers why they have to use packages of third_party
folder instead of packages in their Linux distribution.

By default, Caffe2 find installed packages in the Linux distribution. If it
cannot be found, as a next step Caffe2 uses the version bundled in third_party folder.

**Changes proposed in this PR:**
1. Added difference between Linux distro packages and third_party packages

**Self assessment:**
Checked.

Signed-off-by: Geunsik Lim <geunsik.lim@samsung.com>
Closes https://github.com/caffe2/caffe2/pull/1724

Reviewed By: pjh5

Differential Revision: D6728185

Pulled By: orionr

fbshipit-source-id: 0c596cf56faaccf947caefc49ea3c6f0a473e9bf
2018-01-25 14:33:18 -08:00

2.7 KiB
Raw Blame History

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

brew install matplotlib --with-python3
pip install jupyter