• 上午把tensorflow安装好了，下午和晚上装caffe的确很费劲。
• 默认CUDA,cuDNN可以用了
• 有些安装顺序自己也不清楚，简直就是碰运气

1. 安装之前依赖项

General dependencies

sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compilersudo apt-get install --no-install-recommends libboost-all-dev

安装matlab见后面：

为什么需要安装Matlab？caffe有Matlab的接口，因此如果需要使用Matlab调用caffe，进行编程，就需要安装Matlab。如果你觉得使用C或Python编程比较难，就请安装Matlab。当然如果不需要，并且后面不会编译caffe生成Matlab的接口，就不需要安装Matlab了。这个纯粹根据个人需求来定。为什么需要安装OpenCV？caffe是用来做深度学习的，深度学习的一大应用对象就是图像和视频。而OpenCV是目前最火的开源计算机视觉库，非常多的项目多用到了OpenCV，当然caffe也依赖OpenCV。所以，需要安装OpenCV，否则无法使用caffe哦OpenCV的版本和cuda的版本最好匹配。这样子安排的目的是为了减少错误出现的概率

 

2.OpeCV安装

从官网()下载Opencv,并将其解压到你要安装的位置，假设解压到了/home/opencv。 安装前准备，创建编译文件夹：

cd ~/opencvmkdir buildcd build

配置：

cmake -D CMAKE_BUILD_TYPE=Release -D CMAKE_INSTALL_PREFIX=/usr/local ..

编译：

make -j8  #-j8表示并行计算，根据自己电脑的配置进行设置，配置比较低的电脑可以将数字改小或不使用，直接输make。

以上只是将opencv编译成功，还没将opencv安装，需要运行下面指令进行安装：

sudo make install

问题：由于CUDA 8.0不支持OpenCV的 GraphCut 算法，可能出现以下错误：

 

/home/dsp/opencv-3.1.0/modules/core/include/opencv2/core/private.cuda.hpp:165:52: note: in definition of macro ‘nppSafeCall’ #define nppSafeCall(expr)  cv::cuda::checkNppError(expr, __FILE__, __LINE__, CV_Func)                                                    ^modules/cudalegacy/CMakeFiles/opencv_cudalegacy.dir/build.make:146: recipe for target 'modules/cudalegacy/CMakeFiles/opencv_cudalegacy.dir/src/graphcuts.cpp.o' failedmake[2]: *** [modules/cudalegacy/CMakeFiles/opencv_cudalegacy.dir/src/graphcuts.cpp.o] Error 1make[2]: *** 正在等待未完成的任务....CMakeFiles/Makefile2:9226: recipe for target 'modules/cudalegacy/CMakeFiles/opencv_cudalegacy.dir/all' failedmake[1]: *** [modules/cudalegacy/CMakeFiles/opencv_cudalegacy.dir/all] Error 2make[1]: *** 正在等待未完成的任务....[ 92%] Linking CXX shared library ../../lib/libopencv_photo.so[ 92%] Built target opencv_photoMakefile:160: recipe for target 'all' failedmake: *** [all] Error 2

 

进入opencv-3.1.0/modules/cudalegacy/src/目录，修改graphcuts.cpp文件，将：

#include "precomp.hpp"#if !defined (HAVE_CUDA) || defined (CUDA_DISABLER)改为#include "precomp.hpp"#if !defined (HAVE_CUDA) || defined (CUDA_DISABLER) || (CUDART_VERSION >= 8000)然后make编译就可以了

- 编译和安装完成

 

装BLAS

这里可以选择（ATLAS，MKL或者OpenBLAS）：不知道这个，下载有问题，所以就没有搞这个，但是makefile.config文件里面有配置

MKL首先下载并安装英特尔® 数学内核库 Linux* 版MKL(Intel(R) Parallel Studio XE Cluster Edition for Linux 2016)，下载链接是：https://software.intel.com/en-us/qualify-for-free-software/student， 使用学生身份（邮件 + 学校）下载Student版，填好各种信息，可以直接下载，同时会给你一个邮件告知序列号。

后面就直接：sudo apt-get install libatlas-base-dev -y  

sudo apt-get install libatlas-base-dev

 3.MATLAB2017a安装

• utuntu2017a安装:
• 安装过程很详细，按照流程就可以按照成功。

4.安装caffe

（1）将终端cd到要安装caffe的位置。（2）从github上获取caffe：git clone https://github.com/BVLC/caffe.git注意：若没有安装Git，需要先安装Git：sudo apt-get install git（3）因为make指令只能make Makefile.config文件，而Makefile.config.example是caffe给出的makefile例子，因此，首先将Makefile.config.example的内容复制到Makefile.config：sudo cp Makefile.config.example Makefile.config（4）打开并修改配置文件：sudo gedit Makefile.config #打开Makefile.config文件

• 自己按照改了一下，里面有写需要注意

## 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# uncomment to disable IO dependencies and corresponding data layers  USE_OPENCV := 1# USE_LEVELDB := 0# USE_LMDB := 0# uncomment to allow MDB_NOLOCK when reading LMDB files (only if necessary)#    You should not set this flag if you will be reading LMDBs with any#    possibility of simultaneous read and write# ALLOW_LMDB_NOLOCK := 1# Uncomment if you're using OpenCV 3OPENCV_VERSION := 3# 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 through *_61 lines for compatibility.# For CUDA < 8.0, comment the *_60 and *_61 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_52,code=sm_52 \        -gencode arch=compute_60,code=sm_60 \        -gencode arch=compute_61,code=sm_61 \        -gencode arch=compute_61,code=compute_61# BLAS choice:# atlas for ATLAS (default)# mkl for MKL# open for OpenBlasBLAS := 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 := /home/dsp#MATLAB_DIR := /home/dsp/bin# 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/dsp/anaconda2PYTHON_INCLUDE := $(ANACONDA_HOME)/include \          $(ANACONDA_HOME)/include/python2.7 \          $(ANACONDA_HOME)/lib/python2.7/site-packages/numpy/core/include# Uncomment to use Python 3 (default is Python 2)# PYTHON_LIBRARIES := boost_python3 python3.5m# PYTHON_INCLUDE := /usr/include/python3.5m \#                 /usr/lib/python3.5/dist-packages/numpy/core/include# We need to be able to find libpythonX.X.so or .dylib.#  PYTHON_LIB := /usr/libPYTHON_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 /usr/include/hdf5/serial/#LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/libINCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include /usr/lib/x86_64-linux-gnu/hdf5/serial/includeLIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu/hdf5/serial# 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# NCCL acceleration switch (uncomment to build with NCCL)# https://github.com/NVIDIA/nccl (last tested version: v1.2.3-1+cuda8.0)# USE_NCCL := 1# Uncomment to use `pkg-config` to specify OpenCV library paths.# (Usually not necessary -- OpenCV libraries are normally installed in one of the above $LIBRARY_DIRS.)# USE_PKG_CONFIG := 1# N.B. both build and distribute dirs are cleared on `make clean`BUILD_DIR := buildDISTRIBUTE_DIR := distribute# Uncomment for debugging. Does not work on OSX due to https://github.com/BVLC/caffe/issues/171# DEBUG := 1# The ID of the GPU that 'make runtest' will use to run unit tests.TEST_GPUID := 0# enable pretty build (comment to see full commands)Q ?= @LINKFLAGS := -Wl,-rpath,$(HOME)/anaconda2/lib

 

## 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# uncomment to disable IO dependencies and corresponding data layersUSE_OPENCV := 0# USE_LEVELDB := 0# USE_LMDB := 0# uncomment to allow MDB_NOLOCK when reading LMDB files (only if necessary)#    You should not set this flag if you will be reading LMDBs with any#    possibility of simultaneous read and write# ALLOW_LMDB_NOLOCK := 1# Uncomment if you're using OpenCV 3OPENCV_VERSION := 3# 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 through *_61 lines for compatibility.# For CUDA < 8.0, comment the *_60 and *_61 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_52,code=sm_52 \        -gencode arch=compute_60,code=sm_60 \        -gencode arch=compute_61,code=sm_61 \        -gencode arch=compute_61,code=compute_61# BLAS choice:# atlas for ATLAS (default)# mkl for MKL# open for OpenBlasBLAS := 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/dsp/anaconda2PYTHON_INCLUDE := $(ANACONDA_HOME)/include \          $(ANACONDA_HOME)/include/python2.7 \          $(ANACONDA_HOME)/lib/python2.7/site-packages/numpy/core/include# Uncomment to use Python 3 (default is Python 2)# PYTHON_LIBRARIES := boost_python3 python3.5m# PYTHON_INCLUDE := /usr/include/python3.5m \#                 /usr/lib/python3.5/dist-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 /usr/include/hdf5/serial/#LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/libINCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include /usr/lib/x86_64-linux-gnu/hdf5/serial/includeLIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu/hdf5/serial# 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# NCCL acceleration switch (uncomment to build with NCCL)# https://github.com/NVIDIA/nccl (last tested version: v1.2.3-1+cuda8.0)# USE_NCCL := 1# 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# N.B. both build and distribute dirs are cleared on `make clean`BUILD_DIR := buildDISTRIBUTE_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 ?=

问题：

第一次编译：出错

sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compilersudo apt-get install --no-install-recommends libboost-all-devAR -o .build_release/lib/libcaffe.aLD -o .build_release/lib/libcaffe.so.1.0.0/usr/bin/ld: 找不到 -lhdf5_hl/usr/bin/ld: 找不到 -lhdf5/usr/bin/ld: 找不到 -lcudnncollect2: error: ld returned 1 exit statusMakefile:572: recipe for target '.build_release/lib/libcaffe.so.1.0.0' failedmake: *** [.build_release/lib/libcaffe.so.1.0.0] Error 1

- hdf5的问题，通过修改Makefile.config文件

在文件里面添加文本由于hdf5库目录更改，所以需要单独添加：#INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial/#LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/libINCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include /usr/lib/x86_64-linux-gnu/hdf5/serial/includeLIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu/hdf5/serial

- 然后再次编译有一个问题：

dsp@dsp:~/caffe$ make all -j16LD -o .build_release/lib/libcaffe.so.1.0.0/usr/bin/ld: 找不到 -lcudnncollect2: error: ld returned 1 exit statusMakefile:572: recipe for target '.build_release/lib/libcaffe.so.1.0.0' failedmake: *** [.build_release/lib/libcaffe.so.1.0.0] Error 1

• i found that in the path "/usr/local/cuda/lib64/" don't have the file liblcudnn.so

该问题还在有待解决。

- 这个问题其实挺简单的：后面自己想清楚了：就是cudnn的链接问题，重新拷贝cudnn文件；然后链接了一遍，后面就不报这个错了

- 继续编译错误：

//home/dsp/anaconda2/lib/libpng16.so.16：对‘inflateValidate@ZLIB_1.2.9’未定义的引用//homecollect2: error: ld returned 1 exit status/dsp/anaconda2/lib/libpng16.so.16：对‘inflateValidate@ZLIB_1.2.9’�Makefile:625: recipe for target '.build_release/tools/upgrade_net_proto_text.bin' failedmake: *** [.build_release/tools/upgrade_net_proto_text.bin] Error 1�make: *** 正在等待未完成的任务....�定义的引用collect2: error: ld returned 1 exit statusMakefile:625: recipe for target '.build_release/tools/upgrade_net_proto_binary.bin' failedmake: *** [.build_release/tools/upgrade_net_proto_binary.bin] Error 1//home/dsp/anaconda2/lib/libpng16.so.16：对‘inflateValidate@ZLIB_1.2.9’未定义的引用collect2: error: ld returned 1 exit statusMakefile:625: recipe for target '.build_release/tools/upgrade_solver_proto_text.bin' failedmake: *** [.build_release/tools/upgrade_solver_proto_text.bin] Error 1//home/dsp/anaconda2/lib/libpng16.so.16：对‘inflateValidate@ZLIB_1.2.9’未定义的引用collect2: error: ld returned 1 exit status

- 然后我添加链接：sudo ln -s /home/username/anaconda2/lib/libpng16.so.16 libpng16.so.16 （方法不行）报另外的错：

/usr/local/cuda-8.0/lib64/libpng16.so.16：对‘inflateValidate@ZLIB_1.2.9’未定义的引用collect2: error: ld returned 1 exit statusMakefile:625: recipe for target '.build_release/tools/upgrade_net_proto_binary.bin' failedmake: *** [.build_release/tools/upgrade_net_proto_binary.bin] Error 1make: *** 正在等待未完成的任务..../usr/local/cuda-8.0/lib64/libpng16.so.16：对‘inflateValidate@ZLIB_1.2.9’未定��/usr�的引用/localcollect2: error: ld returned 1 exit status/cuda-8.0/lib64/libpng16.so.16：对‘inflateValidate@ZLIB_1.2.9’未定义/的引用collect2: error: ld returned 1 exit statususr/local/cudaMakefile:630: recipe for target '.build_release/examples/siamese/convert_mnist_siamese_data.bin' failed-make: *** [.build_release/examples/siamese/convert_mnist_siamese_data.bin] Error 18.0/lib64/libpng16.so.16：对‘inflateValidate@ZLIB_1.2.9’未定义的引用collect2: error: ld returned 1 exit status

- 最后非常感谢：Caffe 编译错误记录：

• 里面的几个错误有代表性，按照下面的方法就没有报这个错了

 

在 Makefile.config 中，加入下一句LINKFLAGS := -Wl,-rpath,$(HOME)/anaconda2/lib

- 然后执行：make all  报错：

dsp@dsp:~/caffe$ make all -j16make: Nothing to be done for 'all'

• 解决方法很简单：
• make: Nothing to be done for `all' 解决方法1.这句提示是说明你已经编译好了，而且没有对代码进行任何改动。若想重新编译，可以先删除以前编译产生的目标文件：    make clean     make

 

5. 黎明的曙光

• 按照如下编译顺序

make all -j16make runtest -j16make pycaffe -j16make matcaffe -j16

- 其中make all 和make runtest时间比较长；make pycaffe 很顺利

[----------] Global test environment tear-down[==========] 2123 tests from 281 test cases ran. (285688 ms total)[  PASSED  ] 2123 tests.dsp@dsp:~/caffe$ make pycaffe -j16touch python/caffe/proto/__init__.pyCXX/LD -o python/caffe/_caffe.so python/caffe/_caffe.cppPROTOC (python) src/caffe/proto/caffe.proto

- 实际使用pycaffe,出错：

dsp@dsp:~/caffe$ pythonPython 2.7.14 |Anaconda, Inc.| (default, Oct 16 2017, 17:29:19) [GCC 7.2.0] on linux2Type "help", "copyright", "credits" or "license" for more information.>>> import sys>>> caffe_root="/home/dsp/caffe/">>> sys.path.insert(0,caffe_root+'python')>>> import caffeTraceback (most recent call last):  File "
     
      ", line 1, in 
      
         File "/home/dsp/caffe/python/caffe/__init__.py", line 1, in 
       
            from .pycaffe import Net, SGDSolver, NesterovSolver, AdaGradSolver, RMSPropSolver, AdaDeltaSolver, AdamSolver, NCCL, Timer  File "/home/dsp/caffe/python/caffe/pycaffe.py", line 15, in 
        
             import caffe.io  File "/home/dsp/caffe/python/caffe/io.py", line 8, in 
         
              from caffe.proto import caffe_pb2  File "/home/dsp/caffe/python/caffe/proto/caffe_pb2.py", line 6, in 
          
            from google.protobuf.internal import enum_type_wrapperImportError: No module named google.protobuf.internal
          
         
        
       
      
     

• 通过conda下安装protobuf即可
• python caffe报错：No module named google.protobuf.internal我装的是anaconda2, 解决方法是在其中安装protobuf最新版本conda install protobuf

6. MNIST数据集测试

配置caffe完成后，我们可以利用MNIST数据集对caffe进行测试，过程如下：

1.将终端定位到Caffe根目录

cd ~/caffe

2.下载MNIST数据库并解压缩

./data/mnist/get_mnist.sh

3.将其转换成Lmdb数据库格式

./examples/mnist/create_mnist.sh

4.训练网络

./examples/mnist/train_lenet.sh

训练的时候可以看到损失与精度数值，如下图：

- make matcaffe 有gcc版本问题

dsp@dsp:~/caffe$ make matcaffe -j16MEX matlab/+caffe/private/caffe_.cpp使用 'g++' 编译。警告: 您使用的 gcc 版本为 '5.4.0'。不支持该版本的 gcc。MEX 当前支持的版本为 '4.9.x'。有关当前支持的编译器列表，请参阅: http://www.mathworks.com/support/compilers/current_release。MEX 已成功完成。

• 解决办法是： 在Makefile里面，大约第410行那一句话CXXFLAGS += -MMD -MP下面添加CXXFLAGS += -std=c++11，最后是这样 CXXFLAGS += -MMD -MP CXXFLAGS += -std=c++11然后在caffe根目录下make clean,make all

- 执行 make mattest的时候，报错：

.......b/+caffe/private/caffe_.mexa64'需要的符号 '_ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEED1Ev'缺少'/usr/lib/x86_64-linux-gnu/libboost_python-py27.so.1.58.0->/home/dsp/caffe/matlab/+caffe/private/caffe_.mexa64'需要的符号'_ZNSt7__cxx1112basic_stringIwSt11char_traitsIwESaIwEE12_M_constructEmw'缺少'/usr/lib/x86_64-linux-gnu/libboost_python-py27.so.1.58.0->/home/dsp/caffe/matlab/+caffe/private/caffe_.mexa64'需要的符号'_ZNSt7__cxx1112basic_stringIwSt11char_traitsIwESaIwEE9_M_createERmm'。出错 caffe.set_mode_cpu (line 5)caffe_('set_mode_cpu');出错 caffe.run_tests (line 6)caffe.set_mode_cpu();

- 参考：

•  最后设置调用caffe/python的路径，可以在任意路径终端下导入caffe
• 经过差不多两天的时间，安装了很多东西，情形庆幸没有重装系统，具体的内容如下：

cuda: /usr/local/opencv_3.1: /usr/local/anaconda2,caffe: /home/dsp/python系统默认：2.7anaconda：2.7 ；虚拟环境下tensorflow_py3.5matlab2017a: /home/dsp/bin/matlabcaffe: /home/dsp/caffe使用方法：------matlab2017a： 终端输入： matlab即可，界面有问题，待解决------默认终端python:dsp@dsp:~$ pythonPython 2.7.14 |Anaconda, Inc.| (default, Oct 16 2017, 17:29:19) [GCC 7.2.0] on linux2------终端输入：spyderpython为：anaconda自带的python2.7------tensorflow1.4 + python3.5使用：dsp@dsp:~$ source activate tensorflow_py3.5(tensorflow_py3.5) dsp@dsp:~$ spyder注： 1. 需要不同的python环境，需要自己创建虚拟环境     2. 安装依赖项时注意，安装的位置     3. 也可以通过：(tensorflow_py3.5) dsp@dsp:~$ anaconda-navigator 来安装和启动spyder------pycharm 使用：1. 解压安装包可直接使用2. 运行：(tensorflow_py3.5) dsp@dsp:~$ sh ./pycharm/bin/pycharm.sh ；只要路径对即可3. 设置解释器为：python2.7 或者tensorflow_py3.5------caffe 使用：1. 使用anaconda自带的python2.7即可2. 添加caffe的路径，再使用3. 本机可以在任意路径终端下：输入：python； 然后：import caffe

 

Reference：

http://blog.csdn.net/shiorioxy/article/details/52652831

http://blog.csdn.net/u012841667/article/details/53572431(makefile.config各代码配置说明)