本文简单记录一下安装MXnet的过程,这是安装Tensflow、caffe之后感觉安装最简单的开源框架。
环境:ubuntu14.04 CPU
1. 依赖库安装
2. MXnet下载
3. 编译
4. 测试mnist
1. 依赖库安装
sudo apt-get update
sudo apt-get install -y build-essential git libblas-dev libopencv-dev
2. MXnet下载
git clone --recursive https://github.com/dmlc/mxnet
3. 编译
拷贝mxnet/make/目录下的config.mk到mxnet/目录下
make -j8
4. 测试mnist
cd mxnet/example/image-classification
python train_mnist.py
说明:
(一)因为之前安装caffe时已经安装了python,所以这里就没有安装
(二)遇到问题unrecognized command line option “-std=c++11”
解决方案:http://blog.youkuaiyun.com/gqb1226/article/details/43967723
参考资料:
1. http://blog.youkuaiyun.com/a350203223/article/details/50260419
2. http://blog.youkuaiyun.com/gqb1226/article/details/43967723