Shulou(Shulou.com)06/01 Report--

This article introduces the knowledge of "how to install machine learning framework Caffe on CentOS system". In the operation of actual cases, many people will encounter such a dilemma, so let the editor lead you to learn how to deal with these situations. I hope you can read it carefully and be able to achieve something!

Note: after the system is installed, make sure that the versions of the four packages of kernel kernel-headers kernel-devel kernel-firmware are the same.

The code is as follows:

# rpm-qa | grep kernel

Note: modify yum configuration file / etc/yum.conf modify keepcache=1 first

1. Installation library

The code is as follows:

Yum-y install epel-release.noarch

(wget http://pkgs.repoforge.org/rpmforge-release/rpmforge-release-0.5.3-1.el6.rf.x86_64.rpm)

Rpm-- import http://apt.sw.be/RPM-GPG-KEY.dag.txt

Rpm-K rpmforge-release-0.5.3-1.el6.rf.*.rpm

Rpm-ivh rpmforge-release-0.5.3-1.el6.rf.*.rpm rpmforge-releaser

2. JDK installation

The code is as follows:

Tar-xf jdk-7u25-linux-x64.tar.gz & & mv jdk1.7.0_25/ jdk1.7 mv jdk1.7/ / opt

Vim / etc/profile

Export JAVA_HOME=/opt/jdk1.7

Export JAVA_BIN=/opt/jdk1.7/bin

Export PATH=$PATH:$JAVA_HOME/bin

Export CLASSPATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar

Export PATH=$JAVA_HOME/bin:$JRE_HOME/bin:$PATH

Source / etc/profile

3. Compile and install python2.7 (python2.6 is used on centOS6)

Python official website: https://www.python.org/downloads/release/python-279/

Replace python before installing boost dependencies

The code is as follows:

# wget https://www.python.org/ftp/python/2.7.9/Python-2.7.9.tgz

Compile and install python2.7 (packages for zlib and openssl must be installed before compilation)

Install packages such as gcc zlib openssl first

The code is as follows:

Tar-xvf Python-2.7.9.tgz-C / usr/src

Cd / usr/src/Python-2.7.9

. / configure-- enable-shared

Make-J12

Make altinstall

(altinstall will distinguish between existing versions when installing) (workaround for libpython2.7.so.1.0: vi / etc/ld.so.conf add / usr/local/lib, and then ldconfig)

Replace python in the system

The code is as follows:

Ls-l `which python python2 python2.6`

Rm / usr/bin/python

Ln-s-f / usr/local/bin/python2.7 / usr/bin/python

Maintain yum availability

The code is as follows:

Vim / usr/bin/yum

#! / usr/bin/python changed to #! / usr/bin/python2.6

4. Install pip

The code is as follows:

Curl-O https://bootstrap.pypa.io/get-pip.py # gets a get-pip.py

Python get-pip.py

Installing collected packages: pip, setuptools, wheel

Successfully installed pip-7.1.0 setuptools-18.0.1 wheel-0.24.0

(you must first install the packages of openssl-devel and zlib, and then perform python compilation. If a red cann't remove easy-install.pth reminder appears when executing the command, but there is no such file in the directory, you can create a new one and execute the command again. When installing the system, it is best to check the options of the development tools, and "Successfully installed pip-6.0.8 setuptools-14.3.1 is installed successfully")

5. Install cuda-6.5 and driver

The code is as follows:

Wget http://developer.download.nvidia.com/compute/cuda/repos/rhel6/x86_64/cuda-repo-rhel6-6.5-14.x86_64.rpm

Rpm-ivh cuda-repo-rhel6-6.5-14.x86_64.rpm

Yum install cuda-6-5

The installation of cuda on the GTX 660graphics card will cause the Xorg to run until the system crashes. It is necessary to change the boot level in / etc/inittab to 3.

Note: the file structure of the driver package is incorrect, so that the nvidia_uvm.ko module cannot be compiled, which needs to be solved manually.

The code is as follows:

Cd / var/lib/dkms/nvidia/346.46

Cp-rv / usr/src/nvidia-346.46 build

If you use yum to install and use the downloaded driver package to upgrade

The code is as follows:

Chmod + x NVIDIA-Linux-x86_64-346.72.run

. / NVIDIA-Linux-x86_64-346.72.run

After reboot, dkms will complete the compilation of nvidia_uvm.ko at boot time.

There are two packages under / lib/modules/ version number / extra/: nvidia.ko nvidia_uvm.ko

The code is as follows:

Lsmod | grep nvidia

Vi / etc/rc.local # edit the file

Modprobe nvidia_uvm # add this bar

5.1 run package installation method

The code is as follows:

Chmod + x cuda_6.5.19_linux_64.run

. / cuda_6.5.19_linux_64.run

6. Install blas

The code is as follows:

Yum-y install blas.x86_64 blas-devel.x86_64\

Atlas.x86_64 atlas-devel.x86_64 atlas-sse3.x86_64 atlas-sse3-devel.x86_64

7. Install opencv

The code is as follows:

Yum-y install ant.x86_64 gcc.x86_64 gcc-c++.x86_64 cmake.x86_64 git.x86_64 pkgconfig.x86_64 gtk2.x86_64 gtk2-devel.x86_64 libdc1394.x86_64 libdc1394-devel.x86_64 libjpeg-turbo.x86_64 libjpeg-turbo-devel.x86_64 libpng.x86_64 libpng-devel.x86_64 libtiff.x86_64 libtiff-devel.x86_64 jasper.x86_64 jasper-libs.x86_64 jasper-devel.x86_64 yasm.x86_64 yasm-devel.x86_64

Pip install numpy

Install ffmpeg: # this package does not need to be installed through yum, yum installation version is not correct

The code is as follows:

Tar-xf ffmpeg-2.6.1.tar.bz2-C / usr/src

Cd / usr/src/ffmpeg-2.6.1/

. / configure-- enable-shared # should be configured as a shared library, otherwise there will be an error in linking static libraries when opencv compiles

Make-J12 & & make install

Unzip opencv-2.4.9

Cd opencv-2.4.9

Mkdir release & & cd release

Modify the source file NCVPixelOperations.hpp

Replace the file with the modules/gpu/src/nvidia/core/NCVPixelOperations.hpp under the opencv path

Configure environment variables:

The code is as follows:

Vim / etc/profile.d/custom.sh configuration completed source / etc/profile.d/custom.sh

#! / bin/bash

Export PATH=/usr/local/MATLAB/R2014a/bin:/usr/local/cuda-6.5/bin:$PATH

Export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/boost-1.55.0/lib:/usr/local/cuda-6.5/lib64:/opt/caffe-master/build/lib:/usr/lib64/atlas

Export LIBRARY_PATH=$LIBRARY_PATH:/usr/local/boost-1.55.0/lib:/usr/local/cuda-6.5/lib64:/opt/caffe-master/build/lib:/usr/lib64/atlas

Export C_INCLUDE_PATH=$C_INCLUDE_PATH:/usr/local/boost-1.55.0/include:/usr/local/cuda-6.5/include:/opt/caffe-master/build/src:/opt/caffe-master/include

Export CPLUS_INCLUDE_PATH=$CPLUS_INCLUDE_PATH:/usr/local/boost-1.55.0/include:/usr/local/cuda-6.5/include:/opt/caffe-master/build/src:/opt/caffe-master/include

Export PYTHONPATH=$PYTHONPATH:/opt/caffe-master/python

Export HISTTIMEFORMAT= "F T"

Cmake-D CMAKE_BUILD_TYPE=RELEASE-D CMAKE_INSTALL_PREFIX=/usr/local..

Make-J12

Make install

8. Install boost-1.55 (1.56 incompatible)

The code is as follows:

Yum-y install libicu.x86_64 libicu-devel.x86_64 bzip2-libs.x86_64 bzip2-devel.x86_64

Tar-xf boost_1_55_0.tar.gz & & cd boost_1_55_0

. / bootstrap.sh

. / b2

. / b2 install

Run the. / b2 install command, installed by default in the / usr/local/lib directory, and the header files in the / usr/local/include/boost directory

9. Install caffe other dependencies:

The code is as follows:

Yum-y install snappy.x86_64 snappy-devel.x86_64 hdf5.x86_64 hdf5-devel.x86_64 epel-release leveldb.x86_64 leveldb-devel.x86_64 libgfortran.x86_64

-compile and install protobuf-2.5.0 protobuf-2.5.0

-tar-xvf protobuf-2.5.0.tar.gz

-cd / usr/src/protobuf-2.5.0

. / configure

Make

Make check

Make install

-export LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH

10. Compile and install other dependent packages of caffe

The code is as follows:

Glog

Wget https://google-glog.googlecode.com/files/glog-0.3.3.tar.gz

Tar zxvf glog-0.3.3.tar.gz

Cd glog-0.3.3

. / configure

Make & & make install

Gflags

Wget https://github.com/schuhschuh/gflags/archive/master.zip

Unzip gflags-master.zip

Cd gflags-master

Mkdir build & & cd build

Export CXXFLAGS= "- fPIC" & & cmake. & & make VERBOSE=1

Make & & make install

Lmdb

Git clone git://gitorious.org/mdb/mdb.git

Tar-xf lmdb.tar

Cd mdb/libraries/liblmdb

Make & & make install # if it prompts an man1 error, manually create a

Mkdir-p / usr/local/man/man1

11. Upload matlab to the server and install it graphically

Install key 12345-67890-12345-67890

Import the lic file after the installation is complete, and then replace libmwservices.so to

/ usr/local/MATLAB/R2014a/bin/glnxa64/ to overwrite and end the installation.

12. Solve the python dependency

The code is as follows:

Pip install 'six > = 1.3'

Easy_install-U distribute

Pip2.7 install PIL-allow-external PIL-allow-unverified PIL

Unpack caffe-master.zip and move the package to the opt directory

The code is as follows:

Cd / opt/caffe-master/python

For i in $(cat requirements.txt); do pip install $I; done # needs to be executed several times

Note: there will be an error. For PIL.Image > = 1.1.7, you can use the command pip install 'PIL' to install.

If you execute the above for loop statement again, you need to upgrade python to version 2.7 or above (under installation and precautions:)

13. Install caffe

Modify the caffe-master/Makefile.config file by adding the following sentences

The code is as follows:

Cp / opt/caffe-master/Makefile.config.example Makefile.config

Vim Makefile.config

MATLAB_DIR: = / usr/local/MATLAB/R2014a/

BLAS: = atlas

BLAS_LIB: = / usr/lib64/atlas

PYTHON_INCLUDE: = / usr/include/python2.7\

/ usr/local/lib/python2.7/site-packages/numpy/core/include\

/ usr/local/include/python2.7

Execute ldconfig

Make all-J12

Make-j12 pycaffe

Make-j12 matcaffe

Make test-J12

Make runtest-J12

If matlab is going to use statically compiled libprotobuf.a, modify Makefile

Just add the-static parameter to the MATLAB_CXXFLAGS entry.

However, matlab models using dynamic libraries may not be available

If everything is all right, the installation of the caffe environment is complete and ready to be tested.

The following is an optional section

Compile and install protobuf-2.5.0 protobuf-2.5.0

The code is as follows:

Tar-xvf protobuf-2.5.0.tar.gz

Cd / usr/src/protobuf-2.5.0

. / configure

Make

Make check

Make install

Export LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH

Continue to install protobuf's python module (skip this step if you don't use python)

The code is as follows:

# cd. / python

# python setup.py build

# python setup.py test

# python setup.py install

Install cudnn

LINUX

The code is as follows:

Cd

Export LD_LIBRARY_PATH= `pwd`: $LD_LIBRARY_PATH

Add to your build and link process by adding-I to your compile line and-L-lcudnn to your link line.

a. Edit to make sure Makefile.config, enable GPU "# CPU_ONLY: = 1", and set "USE_CUDNN: = 1".

b. Install cuDNN

The code is as follows:

Tar-xzvf cudnn-6.5-linux-R1.tgz

Cd cudnn-6.5-linux-R1

Cp lib* / usr/local/cuda-6.5/lib64/

Cp cudnn.h / usr/local/cuda-6.5/include/

Cd / usr/local/cuda-6.5/lib64/

Rm-rf libcudnn.so libcudnn.so.6.5

Chmod uprirwx, gendrx, libcudnn.so.6.5.18.

Ln-s libcudnn.so.6.5.18 libcudnn.so.6.5

Ln-s libcudnn.so.6.5 libcudnn.so

Ldconfig

Note 1: put the relevant header files and library files in the system path defined in profile. The corresponding libraries need to be loaded during the mex runtime of matlab

Caffe can also be modified in Makefile.config at compile time to add the path / cache/INSTALL_cuDNN/cudnn-6.5-linux-R1 of cuDNN

The code is as follows:

INCLUDE_DIRS: = $(PYTHON_INCLUDE) / usr/local/include / cache/INSTALL_cuDNN/cudnn-6.5-linux-R1

LIBRARY_DIRS: = $(PYTHON_LIB) / usr/local/lib / usr/lib / cache/INSTALL_cuDNN/cudnn-6.5-linux-R1

Note 2: when using tesla-c2050 graphics card, you need to change the following areas in Makefile.config:

The code is as follows:

PYTHON_LIB: = / usr/lib64 # was originally PYTHON_LIB: = / usr/lib

LIBRARY_DIRS: = $(PYTHON_LIB) / usr/local/lib64 / usr/lib64 # was originally as follows:

LIBRARY_DIRS: = $(PYTHON_LIB) / usr/local/lib / usr/lib

Note 3:protobuf manual installation, do not need to install the wrong version through yum,yum, make runtest will report an error, use the version of protobuf2.5, see the installation method above, install before compiling caffe and then compile.

Package: packages to be uploaded in coffee environment: gflags-master.zip, opencv-2.4.9.zip, boost_1_55_0.tar.gz, caffe-master.zip, glog-0.3.3.tar.gz, protobuf-2.5.0.tar.gz, cuda-repo-rhel6-6.514.x86_64.rpm, jdk-7u25-linux-x64.tar.gz, lmdb.tar, Python-2.7.9.tgz, ffmpeg-2.6.1.tar.bz2,

Rpmforge-release-0.5.3-1.el6.rf.x86_64.rpm, NVIDIA-Linux-x86_64-346.72.run,

NCVPixelOperations.hpp, matlab folder

This is the end of the content of "how to install machine learning framework Caffe on CentOS system". Thank you for reading. If you want to know more about the industry, you can follow the website, the editor will output more high-quality practical articles for you!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.