Ubuntu 14.04 配置caffe环境

Last editted on 2016.7.23.

原文地址：

http://blog.csdn.net/ubunfans/article/details/47724341

略有更改,更改的部分参考了http://www.cnblogs.com/platero/p/3993877.html

-----------------------------------------------------------------------------------

Caffe 安装配置步骤：

1， 安装开发所需的依赖包

sudo apt-get install build-essential  # basic requirement
sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libboost-all-dev libhdf5-serial-dev libgflags-dev libgoogle-glog-dev liblmdb-dev protobuf-compiler #required by caffe

2，安装CUDA 7.0

验证过程省略，按照官方文档自己操作吧（遇到问题首先要看官方文档啊，血泪教训）

安装CUDA有两种方法，

离线.run安装：从官网下载对应版本的.run安装包安装，安装过程挺复杂，尝试过几次没成功，遂放弃。

在离线.deb安装：deb安装分离线和在线，我都尝试过都安装成功了，官网下载地址

官网上也有很明确的安装说明，具体可参考官网。

安装之前请先进行md5校验，确保下载的安装包完整。

这里给出一个我下载好的链接：http://pan.baidu.com/s/1kVi2uFX 密码：af78

切换到下载的deb所在目录，执行下边的命令

sudo dpkg -i cuda-repo-ubuntu1404-7-5-local_7.5-18_amd64.deb  #sudo dpkg -i cuda-repo-<distro>_<version>_<architecture>.deb
sudo apt-get update
sudo apt-get install cuda

然后重启电脑：sudo reboot

NOTE:装不成功卸了多来几遍，总会成的

3，安装cuDNN

下载cudnn-6.5-linux-x64-v2.tgz，官网申请不到，网上自己找的，百度云链接 http://pan.baidu.com/s/1o734sp8

tar -zxvf cudnn-6.5-linux-x64-v2.tgz
cd cudnn-6.5-linux-x64-v2
sudo cp lib* /usr/local/cuda/lib64/
sudo cp cudnn.h /usr/local/cuda/include/

更新软连接

cd /usr/local/cuda/lib64/
sudo rm -rf libcudnn.so libcudnn.so.6.5  
chmod u=rwx,g=rx,o=rx libcudnn.so.6.5.48
sudo ln -s libcudnn.so.6.5.48 libcudnn.so.6.5
sudo ln -s libcudnn.so.6.5 libcudnn.so

上边的操作为什么这么做，不知道，原理是什么，不知道。等我知道了再来补充

4，设置环境变量

在/etc/profile中添加CUDA环境变量

sudo gedit /etc/profile

PATH=/usr/local/cuda/bin:$PATH
export PATH

保存后, 执行下列命令, 使环境变量立即生效

source /etc/profile

同时需要添加lib库路径： 在 /etc/ld.so.conf.d/加入文件 cuda.conf, 内容如下

/usr/local/cuda/lib64

保存后，执行下列命令使之立刻生效

sudo ldconfig

5，安装CUDA SAMPLE

进入/usr/local/cuda/samples, 执行下列命令来build samples

sudo make all -j4

整个过程大概10分钟左右, 全部编译完成后， 进入 samples/bin/x86_64/linux/release, 运行deviceQuery

./deviceQuery

如果出现显卡信息， 则驱动及显卡安装成功：

./deviceQuery Starting...

 CUDA Device Query (Runtime API) version (CUDART static linking)

Detected 1 CUDA Capable device(s)

Device 0: "GeForce GTX 670"
  CUDA Driver Version / Runtime Version          6.5 / 6.5
  CUDA Capability Major/Minor version number:    3.0
  Total amount of global memory:                 4095 MBytes (4294246400 bytes)
  ( 7) Multiprocessors, (192) CUDA Cores/MP:     1344 CUDA Cores
  GPU Clock rate:                                1098 MHz (1.10 GHz)
  Memory Clock rate:                             3105 Mhz
  Memory Bus Width:                              256-bit
  L2 Cache Size:                                 524288 bytes
  Maximum Texture Dimension Size (x,y,z)         1D=(65536), 2D=(65536, 65536), 3D=(4096, 4096, 4096)
  Maximum Layered 1D Texture Size, (num) layers  1D=(16384), 2048 layers
  Maximum Layered 2D Texture Size, (num) layers  2D=(16384, 16384), 2048 layers
  Total amount of constant memory:               65536 bytes
  Total amount of shared memory per block:       49152 bytes
  Total number of registers available per block: 65536
  Warp size:                                     32
  Maximum number of threads per multiprocessor:  2048
  Maximum number of threads per block:           1024
  Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
  Max dimension size of a grid size    (x,y,z): (2147483647, 65535, 65535)
  Maximum memory pitch:                          2147483647 bytes
  Texture alignment:                             512 bytes
  Concurrent copy and kernel execution:          Yes with 1 copy engine(s)
  Run time limit on kernels:                     Yes
  Integrated GPU sharing Host Memory:            No
  Support host page-locked memory mapping:       Yes
  Alignment requirement for Surfaces:            Yes
  Device has ECC support:                        Disabled
  Device supports Unified Addressing (UVA):      Yes
  Device PCI Bus ID / PCI location ID:           1 / 0
  Compute Mode:
     < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 6.5, CUDA Runtime Version = 6.5, NumDevs = 1, Device0 = GeForce GTX 670
Result = PASS

NOTE:上边的显卡信息是从别的地方拷过来的，我的GTX650显卡不是这些信息，如果没有这些信息，那肯定是安装不成功，找原因吧！

6，安装Intel MKL 或Atlas

我没有MKL（Intel数学核心函数库（MKL）是一套高度优化、线程安全的数学例程、函数，面向高性能的工程、科学与财务应用。英特尔 MKL 的集群版本包括 ScaLAPACK 与分布式内存快速傅立叶转换，并提供了线性代数 (BLAS、LAPACK 和Sparse Solver)、快速傅立叶转换、矢量数学 (Vector Math) 与随机号码生成器支持），装的Atlas（全称：Automatic Tuned Linear Algebra Software，BLAS线性算法库的优化版本）

上述两个库有什么区别联系可以看知乎的这个提问 http://www.zhihu.com/question/27872849

安装命令：

sudo apt-get install libatlas-base-dev

7，安装OpenCV

这个尽量不要手动安装， Github上有人已经写好了完整的安装脚本：https://github.com/jayrambhia/Install-OpenCV

去掉去掉bash脚本中 rm -rf OpenCV，否则卸载时候比较麻烦~github上大神给的bash脚本在应用在caffe上cmake时候会有一些问题。

这里给出一个我自己修改过的bash脚本链接：http://pan.baidu.com/s/1chyMZ0 密码：oohx

下载该脚本，进入Ubuntu/2.4 目录, 给所有shell脚本加上可执行权限

chmod +x *.sh

然后安装 （用2.4.9吧，2.4.10貌似有人说有问题 ）

sudo ./opencv2_4_9.sh

脚本会自动安装依赖项，下载安装包，编译并安装OpenCV。整个过程大概半小时左右。 

注意，中途可能会报错

opencv-2.4.9/modules/gpu/src/nvidia/core/NCVPixelOperations.hpp(51): error: a storage class is not allowed in an explicit specialization

解决方法在此：http://code.opencv.org/issues/3814  下载 NCVPixelOperations.hpp 替换掉opencv2.4.9内的文件， 重新build。

链接：http://pan.baidu.com/s/1c2DOFDY 密码：rohl   来源：http://download.csdn.net/download/solomon1558/9581114，感谢博主Solomon1588

这个是opencv的一个bug，在后续版本里被修复了。

The compilation problem can be solved by removing all the 'static' keywords in the template functions in that file.

 

可能错误：nvcc fatal : Unsupported gpu architecture 'compute_11'

解决方法：cmake -D CMAKE_BUILD_TYPE=RELEASE -D CMAKE_INSTALL_PREFIX=/usr/local -D CUDA_GENERATION=Kepler ..

error:fatal error: opencv2/cudalegacy/NCV.hpp: No such file or directory

8，安装Caffe所需要的Python环境 

按caffe官网的推荐使用Anaconda

这里给出Anaconda-2.3.0-Linux-x86_64.sh的链接，http://pan.baidu.com/s/1pKg1HFl

也可以去Anaconda官网下载安装包
切换到文件所在目录，执行

bash Anaconda-2.3.0-Linux-x86_64.sh

NOTE:后边的文件名按自己下的版本号更改，整个安装过程请选择默认

最好还是用这个版本吧，开始时候用了2.5的最新版，结果会报错，找了一天的问题，都是泪啊～换个版本就ok了

  8.1，添加Anaconda Library Path

在/etc/ld.so.conf最后加入以下路径，并没有出现重启不能进入界面的问题

 

/root/anaconda/lib

在~/.bashrc最后添加下边路径

export LD_LIBRARY_PATH="/root/anaconda/lib:$LD_LIBRARY_PATH"

9，安装python依赖库

去caffe的github下载caffe源码包

这里给出一个下载好的caffe源文件：链接：http://pan.baidu.com/s/1kUMNFCF 密码：9qoo
进入caffe-master下的python目录

执行如下命令

for req in $(cat requirements.txt); do pip install $req; done

10，安装MATLAB

Caffe提供了MATLAB接口， 有需要用MATLAB的同学可以额外安装MATLAB。 安装教程请自行搜索。 

http://blog.csdn.net/lanbing510/article/details/41698285

安装完成后添加图标 http://www.linuxidc.com/Linux/2011-01/31632.htm

sudo vi /usr/share/applications/Matlab.desktop

输入以下内容

[Desktop Entry]
Type=Application
Name=Matlab
GenericName=Matlab 2010b
Comment=Matlab:The Language of Technical Computing
Exec=sh /usr/local/MATLAB/R2010b/bin/matlab -desktop
Icon=/usr/local/MATLAB/Matlab.png
Terminal=false
Categories=Development;Matlab;

 

(I use the R2013b patched package. First you should uncompress the .iso file. Then use sudo cp to copy the patch file)

 

11，编译Caffe

终于来到这里了

这里给出一个下载好的caffe源文件：链接：http://pan.baidu.com/s/1c1LacZA 密码：4j7n  (支持cndnn v2 from http://stackoverflow.com/questions/27987691/issues-with-compiling-caffe-with-cudnn，现有的caffe不支持cudnn v2)

进入caffe-master目录，复制一份Makefile.config.examples

cp Makefile.config.example Makefile.config

修改其中的一些路径，如果前边和我说的一致，都选默认路径的话，那么配置文件应该张这个样子

去掉前面的#号即去掉注释。

## Refer to http://caffe.berkeleyvision.org/installation.html
# Contributions simplifying and improving our build system are welcome!

# cuDNN acceleration switch (uncomment to build with cuDNN).
# USE_CUDNN := 1

# CPU-only switch (uncomment to build without GPU support).
# CPU_ONLY := 1

# To customize your choice of compiler, uncomment and set the following.
# N.B. the default for Linux is g++ and the default for OSX is clang++
# CUSTOM_CXX := g++

# CUDA directory contains bin/ and lib/ directories that we need.
CUDA_DIR := /usr/local/cuda
# On Ubuntu 14.04, if cuda tools are installed via
# "sudo apt-get install nvidia-cuda-toolkit" then use this instead:
# CUDA_DIR := /usr

# CUDA architecture setting: going with all of them.
# For CUDA < 6.0, comment the *_50 lines for compatibility.
CUDA_ARCH := -gencode arch=compute_20,code=sm_20 \
        -gencode arch=compute_20,code=sm_21 \
        -gencode arch=compute_30,code=sm_30 \
        -gencode arch=compute_35,code=sm_35 \
        -gencode arch=compute_50,code=sm_50 \
        -gencode arch=compute_50,code=compute_50

# BLAS choice:
# atlas for ATLAS (default)
# mkl for MKL
# open for OpenBlas
BLAS := atlas
# Custom (MKL/ATLAS/OpenBLAS) include and lib directories.
# Leave commented to accept the defaults for your choice of BLAS
# (which should work)!
# BLAS_INCLUDE := /path/to/your/blas
# BLAS_LIB := /path/to/your/blas

# Homebrew puts openblas in a directory that is not on the standard search path
# BLAS_INCLUDE := $(shell brew --prefix openblas)/include
# BLAS_LIB := $(shell brew --prefix openblas)/lib

# This is required only if you will compile the matlab interface.
# MATLAB directory should contain the mex binary in /bin.
# MATLAB_DIR := /usr/local
# MATLAB_DIR := /Applications/MATLAB_R2012b.app

# NOTE: this is required only if you will compile the python interface.
# We need to be able to find Python.h and numpy/arrayobject.h.
#PYTHON_INCLUDE := /usr/include/python2.7 \
        /usr/lib/python2.7/dist-packages/numpy/core/include
# Anaconda Python distribution is quite popular. Include path:
# Verify anaconda location, sometimes it's in root.
 ANACONDA_HOME := $(HOME)/anaconda
 PYTHON_INCLUDE := $(ANACONDA_HOME)/include \
         $(ANACONDA_HOME)/include/python2.7 \
         $(ANACONDA_HOME)/lib/python2.7/site-packages/numpy/core/include \

# We need to be able to find libpythonX.X.so or .dylib.
#PYTHON_LIB := /usr/lib
PYTHON_LIB := $(ANACONDA_HOME)/lib

# Homebrew installs numpy in a non standard path (keg only)
# PYTHON_INCLUDE += $(dir $(shell python -c 'import numpy.core; print(numpy.core.__file__)'))/include
# PYTHON_LIB += $(shell brew --prefix numpy)/lib

# Uncomment to support layers written in Python (will link against Python libs)
# WITH_PYTHON_LAYER := 1

# Whatever else you find you need goes here.
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib

# If Homebrew is installed at a non standard location (for example your home directory) and you use it for general dependencies
# INCLUDE_DIRS += $(shell brew --prefix)/include
# LIBRARY_DIRS += $(shell brew --prefix)/lib

# Uncomment to use `pkg-config` to specify OpenCV library paths.
# (Usually not necessary -- OpenCV libraries are normally installed in one of the above $LIBRARY_DIRS.)
# USE_PKG_CONFIG := 1

BUILD_DIR := build
DISTRIBUTE_DIR := distribute

# Uncomment for debugging. Does not work on OSX due to https://github.com/BVLC/caffe/issues/171
# DEBUG := 1

# The ID of the GPU that 'make runtest' will use to run unit tests.
TEST_GPUID := 0

# enable pretty build (comment to see full commands)
Q ?= @

保存退出

编译

./include/caffe/layers/cudnn_conv_layer.hpp:52:3: error: ‘cudnnConvolutionBwdFilterAlgo_t’ does not name a type
   cudnnConvolutionBwdFilterAlgo_t *bwd_filter_algo_;

make all -j4
make test
make runtest

遇到的报错：

//usr/lib/x86_64-linux-gnu/libunwind.so.8：对‘lzma_index_buffer_decode@XZ_5.0’ 未定义的引用

解决方法：

直接换Anaconda版本，前面说anaconda安装过程时候有提到。

如果还不行，在～/.bashrc的末尾加上export LD_LIBRARY_PATH="/lib/x86_64-linux-gnu:$LD_LIBRARY_PATH"，然后重启.

还不行，参考https://github.com/BVLC/caffe/issues/985

总之最大可能就是anaconda的问题。

 

make runtest-----> YOU HAVE 2 DISABLED TESTS.  

Just ignore it.  can refer to github.com/BVLC/caffe/issues/982

 

12.1. 编译Matlab wrapper

执行如下命令

make matcaffe

然后就可以跑官方的matlab demo啦。

12.2. 编译Python wrapper

 make pycaffe 

到这里就基本结束了，可以跑个MNIST尝试一下，可参考http://www.cnblogs.com/denny402/p/5075490.html, run MNIST. my GPU is GTX 760. finished in less than 1 minute with cudnn acceleration. (my ubuntu is new. i have not installed Chinese language support. it does not have Chinese input method. orz)

(HAVE FUN!)