MXnet安装和编译

根据install MXNet
MXNet安装其实是从c++文件make编译成动态库到incubator-mxnet/lib/libmxnet.so，而incubator-mxnet/python其实是用python调用c++库的方法封装，然后用pip安装到python，之后example里面的文件就可以import mxnet as mx了。

Linux-Python-CPU-Build from Source on Ubuntu安装。

• 查看系统

lsb_release -a

• Minimum Requirements
  GCC版本：gcc -v
  GNU Make：已安装

• Step 1 Install build tools and git.

$ sudo apt-get update
$ sudo apt-get install -y build-essential git

• Step 2 Install OpenBLAS

$ sudo apt-get install -y libopenblas-dev liblapack-dev

• Step 3 Install OpenCV.

$ sudo apt-get install -y libopencv-dev

• Step 4 Download MXNet sources and build MXNet core shared library.

$ git clone --recursive https://github.com/apache/incubator-mxnet
$ cd incubator-mxnet
$ make -j $(nproc) USE_OPENCV=1 USE_BLAS=openblas

cd incubator-mxnet
git submodule init
git submodule update

出错：src/nnvm/tvm_bridge.cc:37:37: fatal error: tvm/runtime/packed_func.h: No such file or directory

• 查看：issues

git submodule update --init --recursive

ps：make时间很长。

• Build the MXNet Python binding

sudo pip uninstall mxnet
cd python
pip install --upgrade pip
sudo pip install -e .

---

Linux-Python-CPU-Build from Source on CentOS, Multi-Machines安装。

• 配置MXNet在各个node上

sudo yum install build-essential git
sudo yum install lapack-devel openblas-devel opencv-devel
git clone --recursive https://github.com/apache/incubator-mxnet
cd incubator-mxnet
make clean_all
make -j $(nproc) USE_OPENCV=1 USE_BLAS=openblas USE_DIST_KVSTORE=1   USE_PROFILER=1 

sudo pip uninstall mxnet
cd python
sudo pip install --upgrade pip
sudo pip install -e .

分布式安装及运行

MXNet运行（分布式+动态库）