The network of the great God configuration is very cumbersome in fact, the official website
http://caffe.berkeleyvision.org/installation.html configuration and installation procedures are simple and for different environments so I suggest you do not follow the XX blog, do a reference can (including my). When you encounter an error, Google, avoid detours.
First update the Yum source
cd/etc/yum.repos.d/
mv Centos-base.repo centos-base.repo_bak
wget http://mirrors.163.com/.help/ Centos7-base-163.repo
wget-o centos-base.repo http://mirrors.aliyun.com/repo/Centos-7.repo
wget-o/etc/ Yum.repos.d/epel.repo Http://mirrors.aliyun.com/repo/epel-7.repo
yum clean all
yum Makecache
According to the official website of the installation dependent library
sudo yum install protobuf-devel leveldb-devel snappy-devel opencv-devel boost-devel hdf5-devel
sudo yum install gflags-devel glog-devel lmdb-devel
Yum Install Atlas-devel
Yum Install Python-devel
git clone Https://github.com/BVLC/caffe
Then compiles, according to the official website steps (-J8 is multi-threaded, speed up):
CP Makefile.config.example Makefile.config
# Adjust Makefile.config (for example, if using Anaconda Python, or if CUDN N is desired) make
all-j8 make
test make
runtest
There may be errors found in Cblas and Atlas two libraries during the compilation process, change the Makefile.config file, and the solution is as follows:
OK, the installation is complete, I believe that the more complex steps for the majority of people is cuda installation, but this step I have been very skilled, and our Cuda has been configured well.