ubantu16.04配置caffe

之前由于安装显卡驱动，导致系统有问题，重装系统后需重新配置caffe环境，再次搜集教程后故记录下配置过程，以防万一

一、安装相关依赖项

1 sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compiler
2 sudo apt-get install --no-install-recommends libboost-all-dev
3 sudo apt-get install libopenblas-dev liblapack-dev libatlas-base-dev
4 sudo apt-get install libgflags-dev libgoogle-glog-dev liblmdb-dev

二、配置cuda

1、下载cuda8.0——下载地址：https://developer.nvidia.com/cuda-downloads

2、三句命令

3、sudo make all

   运行完出现如下问题：

usr/bin/ld: cannot find -lnvcuvid
collect2: error: ld returned 1 exit status
Makefile:381: recipe for target 'cudaDecodeGL' failed
make[1]: *** [cudaDecodeGL] Error 1
make[1]: Leaving directory '/usr/local/cuda-8.0/samples/3_Imaging/cudaDecodeGL'
Makefile:52: recipe for target '3_Imaging/cudaDecodeGL/Makefile.ph_build' failed
make: *** [3_Imaging/cudaDecodeGL/Makefile.ph_build] Error 2(错误信息暂且忽略)

4、执行sudo ./deviceQuery，如果显示一些关于GPU的信息，则说明安装成功。
5、sudo nvidia-smi ，若列出了GPU的信息列表则表示驱动同时也安装成功

 三、配置cuDNN
cuDNN是GPU加速计算深层神经网络的库。首先去官网 https://developer.nvidia.com/rdp/cudnn-download 下载cuDNN，注意对应系统环境和版本，按照安装文件进行安装

下载cuDNN之后进行解压：$ tar -xzvf cudnn-8.0-linux-x64-v7.tgz

进入解压之后的include目录，在命令行进行如下操作：

$ sudo cp cuda/include/cudnn.h /usr/local/cuda/include

$ sudo cp cuda/lib64/libcudnn* /usr/local/cuda/lib64

$ sudo chmod a+r /usr/local/cuda/include/cudnn.h

再将进入lib64目录下的动态文件进行复制和链接：

cd /usr/local/cuda/lib64/
sudo rm -rf libcudnn.so libcudnn.so.5    #删除原有动态文件
sudo ln -s libcudnn.so.5.0.5 libcudnn.so.5  #生成软衔接
sudo ln -s libcudnn.so.5 libcudnn.so      #生成软链接

四、配置安装opencv（安装过程有点漫长）
1、下载地址：https://github.com/jayrambhia/install-opencv
2、安装命令

if your system is Ubuntu, run the commands below.
$ cd Ubuntu
$ chmod +x * 
$ ./opencv_latest.

if your system is RedHat, run the commands below.
$ cd RedHat
$ chmod +x * 
$ ./opencv_latest.sh

if your system is ArchLinux, run the commands below.
$ cd ArchLinux
$ chmod +x * 
$ ./opencv2_4_0.sh

3、安装完进行配置
五、 安装python环境
1、下载Python，下载地址：https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/ (Anaconda-2.3.0-Linux-x86_64.sh)
2、bash Anaconda-2.3.0-Linux-x86_64.sh
3、sudo gedit ~/.bashrc最后一行添加 export LD_LIBRARY_PATH="/home/th/anaconda/lib:$LD_LIBRARY_PATH"

五、安装caffe的python 依赖库
在guthub下载caffe-master包,进入Python目录执行如下代码：
for req in $(cat requirements.txt); do pip install $req; done

六、编译caffe
sudo make all -j4
sudo make test -j4
sudo make runtest -j4

七、pycaffe配置
1、sudo apt-get install python-numpy python-scipy python-matplotlib python-sklearn python-skimage python-h5py python-protobuf python-leveldb python-networkx python-nose python-pandas python-gflags Cython ipython
sudo apt-get install python-numpy python-scipy python-matplotlib python-s
提示错误：E: Unable to locate package Cython等
解决方法：python 后 import numpy等依赖库，如不报错则安装成功，报错单独安装；
2、 cd ./caffe
sudo make pycaffe 编译依赖库
显示错误：make: Nothing to be done for `pycaffe'，检查第一步是否安装完毕，然后sudo make clean后重新编译。
3、添加python path
sudo gedit /etc/profile
添加：export PYTHONPATH=/media/th/7063baa9-26bd-4f54-9d0f-17d840d49852/caffe-master/python:$PYTHONPATH
source /etc/profile
八、测试
python
import caffe
错误：ImportError: No module named caffe
解决办法：见http://blog.csdn.net/weixin_40122581/article/details/78203538