ubuntu14.04 安装 caffe

组件&版本

• Ubuntu

  • 14.04

• CUDA

  • 8.0
• cuDNN
  • cudnn-7.0-linux-x64-v3.0-prod.tgz
• opencv
  • 3.2
• python
  • Python 2.7.6
• caffe
  • release candidate 3

安装过程

安装开发所需依赖包

sudo apt-get install build-essential  
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

安装CUDA 8.0

• 确定 GPU 支持 CUDA

  lspci | grep -i nvidia

  显示

  02:00.0 VGA compatible controller: NVIDIA Corporation Device 1b06 (rev a1)
  02:00.1 Audio device: NVIDIA Corporation Device 10ef (rev a1)
  03:00.0 VGA compatible controller: NVIDIA Corporation Device 1b06 (rev a1)
  03:00.1 Audio device: NVIDIA Corporation Device 10ef (rev a1)
  82:00.0 VGA compatible controller: NVIDIA Corporation Device 1b06 (rev a1)
  82:00.1 Audio device: NVIDIA Corporation Device 10ef (rev a1)
  83:00.0 VGA compatible controller: NVIDIA Corporation Device 1b06 (rev a1)
  83:00.1 Audio device: NVIDIA Corporation Device 10ef (rev a1)

  根据 http://developer.nvidia.com/cuda-gpus 去验证，发现支持 CUDA

• 确定 linux 版本支持 CUDA

  uname -m && cat /etc/*release

  显示

  x86_64
  DISTRIB_ID=Ubuntu
  DISTRIB_RELEASE=14.04
  DISTRIB_CODENAME=trusty
  DISTRIB_DESCRIPTION="Ubuntu 14.04.5 LTS"
  NAME="Ubuntu"
  VERSION="14.04.5 LTS, Trusty Tahr"
  ID=ubuntu
  ID_LIKE=debian
  PRETTY_NAME="Ubuntu 14.04.5 LTS"
  VERSION_ID="14.04"
  HOME_URL="http://www.ubuntu.com/"
  SUPPORT_URL="http://help.ubuntu.com/"
  BUG_REPORT_URL="http://bugs.launchpad.net/ubuntu/"

• 确定系统已经安装了 gcc

  gcc --version

  显示

  gcc (Ubuntu 4.8.5-2ubuntu1~14.04.1) 4.8.5
  Copyright (C) 2015 Free Software Foundation, Inc.
  This is free software; see the source for copying conditions.  There is NO
  warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

• 确定系统已经安装了正确的 Kernel Headers 和开发包

  • 查看系统正在运行的 kernel 版本

    uname -r

    显示

    3.13.0-142-generic

  • 安装对应的 kernels header 和开发包

    sudo apt-get install linux-headers-$(uname -r)

• 安装 CUDA

  • 下载对应版本的cuda，从 https://developer.nvidia.com/cuda-downloads 下载 deb(network) 版

• 进行 md5 校验

  md5sum cuda-repo-ubuntu1404_8.0.61-1_amd64.deb

• 对比相同，然后使用 deb 文件进行安装

  sudo dpkg -i cuda-repo-ubuntu1404_8.0.61-1_amd64.deb
  sudo apt-get update
  sudo apt-get install cuda

• 重启，完成 cuda 安装

安装 cuDNN

• 从 https://developer.nvidia.com/rdp/cudnn-download 下载 cuDNN ，版本为 cudnn-7.0-linux-x64-v3.0-prod.tgz

• 安装

  tar -zxvf cudnn-7.0-linux-x64-v3.0-prod.tgz
  cd cuda
  sudo cp lib64/* /usr/local/cuda/lib64/
  sudo cp include/cudnn.h /usr/local/cuda/include/

• 更新软链接

  cd /usr/local/cuda/lib64
  sudo rm -rf libcudnn.so libcudnn.so.7.0
  sudo ln -s libcudnn.so.7.0.64 libcudnn.so.7.0
  sudo ln -s libcudnn.so.7.0 libcudnn.so

设置 CUDA 环境变量

• VERSION1

  • 在 /etc/profile 中添加 CUDA 环境变量

  • 执行

    sudo vim /etc/profile

    在打开的文件中加入如下两句话

    export PATH=/usr/local/cuda/bin:$PATH
    export LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH

  • 保存后，使环境变量立即生效，执行

    source /etc/profile

• VERSION2

  • 在 ~/.bashrc 中添加 CUDA 环境变量

  • 执行

    sudo vim ~/.bashrc

    在打开的文件中加入如下两句话

    export PATH=/usr/local/cuda/bin:$PATH
    export LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH

  • 保存后，使环境变量立即生效，执行

• 我用的 VERSION2

安装cuda samples

• 进入 /usr/local/cuda/samples

  cd /usr/local/cuda/samples

• 执行下面的命令来 build samples

  sudo make all -j4

• 全部编译完成后，进入 samples/bin/x86_64/linux/release ，运行 deviceQuery

  cd /usr/local/cuda/samples/bin/x86_64/linux/release
  ./deviceQuery

  如果出现显卡信息，则驱动及显卡安装成功，结果如下

  ./deviceQuery Starting...
  
   CUDA Device Query (Runtime API) version (CUDART static linking)
  
  Detected 4 CUDA Capable device(s)
  
  Device 0: "GeForce GTX 1080 Ti"
  	.
  	.
  	.
  Device 3: "GeForce GTX 1080 Ti"
    CUDA Driver Version / Runtime Version          9.0 / 8.0
    CUDA Capability Major/Minor version number:    6.1
    Total amount of global memory:                 11172 MBytes (11715084288 bytes)
    (28) Multiprocessors, (128) CUDA Cores/MP:     3584 CUDA Cores
    GPU Max Clock rate:                            1582 MHz (1.58 GHz)
    Memory Clock rate:                             5505 Mhz
    Memory Bus Width:                              352-bit
    L2 Cache Size:                                 2883584 bytes
    Maximum Texture Dimension Size (x,y,z)         1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
    Maximum Layered 1D Texture Size, (num) layers  1D=(32768), 2048 layers
    Maximum Layered 2D Texture Size, (num) layers  2D=(32768, 32768), 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 2 copy engine(s)
    Run time limit on kernels:                     No
    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 Domain ID / Bus ID / location ID:   0 / 131 / 0
    Compute Mode:
       < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
  > Peer access from GeForce GTX 1080 Ti (GPU0) -> GeForce GTX 1080 Ti (GPU1) : Yes
  > Peer access from GeForce GTX 1080 Ti (GPU0) -> GeForce GTX 1080 Ti (GPU2) : No
  > Peer access from GeForce GTX 1080 Ti (GPU0) -> GeForce GTX 1080 Ti (GPU3) : No
  > Peer access from GeForce GTX 1080 Ti (GPU1) -> GeForce GTX 1080 Ti (GPU0) : Yes
  > Peer access from GeForce GTX 1080 Ti (GPU1) -> GeForce GTX 1080 Ti (GPU2) : No
  > Peer access from GeForce GTX 1080 Ti (GPU1) -> GeForce GTX 1080 Ti (GPU3) : No
  > Peer access from GeForce GTX 1080 Ti (GPU2) -> GeForce GTX 1080 Ti (GPU0) : No
  > Peer access from GeForce GTX 1080 Ti (GPU2) -> GeForce GTX 1080 Ti (GPU1) : No
  > Peer access from GeForce GTX 1080 Ti (GPU2) -> GeForce GTX 1080 Ti (GPU3) : Yes
  > Peer access from GeForce GTX 1080 Ti (GPU3) -> GeForce GTX 1080 Ti (GPU0) : No
  > Peer access from GeForce GTX 1080 Ti (GPU3) -> GeForce GTX 1080 Ti (GPU1) : No
  > Peer access from GeForce GTX 1080 Ti (GPU3) -> GeForce GTX 1080 Ti (GPU2) : Yes
  
  deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 9.0, CUDA Runtime Version = 8.0, NumDevs = 4, Device0 = GeForce GTX 1080 Ti, Device1 = GeForce GTX 1080 Ti, Device2 = GeForce GTX 1080 Ti, Device3 = GeForce GTX 1080 Ti
  Result = PASS

  因为我是 4 个 GPU ，所以会有多个 Device 信息

  安装成功！

安装Intel MKL，openBlas 或Atlas

我选择的是 Atlas ，为 caffe 默认使用的，不要额外配置，安装命令

sudo apt-get install libatlas-base-dev

安装opencv

• 首先安装必须的包

  sudo apt-get install cmake git libgtk2.0-dev pkg-config libavcodec-dev libavformat-dev libswscale-dev

• 从 opencv 官网下载最新版本 opencv ，3.2.0.zip

  解压

  unzip 3.2.0.zip

• 编译

  cd  opencv-3.2.0
  mkdir release
  cd release
  cmake -D CMAKE_BUILD_TYPE=RELEASE -D CMAKE_INSTALL_PREFIX=/usr/local .. -DWITH_LAPACK=OFF
  make -j4
  sudo make install

• 测试

  因为我的 GPU 服务器走的别人的网关，这里显示不出来图片，也就没有测试。

  但是是需要测试的，自行 Google 吧

安装python依赖库

• 下载 caffe 源码，解压后进入 caffe-master 下的 python 目录

  这里一定要注意当前下载的 caffe 版本是否支持自己安装的 cuDNN 版本

• 安装 python-pip

  sudo apt-get install python-pip

• 执行如下命令安装相关依赖

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

编译Caffe

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

  cp Makefile.config.example Makefile.config

• 修改 Makefile.config

  开启对 CuDNN 和 Opencv3 的支持

  ## 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 := 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 3
  OPENCV_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 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 \
  
  # 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
  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

  make all -j4
  make test -j4
  make runtest -j4

• 结果

  [ FAILED ] 5 tests, listed below:
  [ FAILED ] CuDNNConvolutionLayerTest/0.TestSimpleConvolutionCuDNN, where TypeParam = float
  [ FAILED ] CuDNNConvolutionLayerTest/0.TestGradientCuDNN, where TypeParam = float
  [ FAILED ] CuDNNConvolutionLayerTest/0.TestGradientGroupCuDNN, where TypeParam = float
  [ FAILED ] CuDNNConvolutionLayerTest/0.TestSobelConvolutionCuDNN, where TypeParam = float
  [ FAILED ] CuDNNConvolutionLayerTest/0.TestSimpleConvolutionGroupCuDNN, where TypeParam = float

  这个是因为我们使用的是 cuDNN v3 版本，而这部分只在 cuDNN v4 下 work

更换cuDNN与caffe版本

为啥要换

我当然不想换！因为我老大不太满意！也没有说不满意，就是素质三连了一下

都行

随你

都可以

。。。

确定依赖关系

• CuDNN 是专门针对 Deep Learning 框架设计的一套 GPU 计算加速方案，目前支持的 DL 库包括 Caffe ，ConvNet ，Torch7 等

  基本原理是把 lib 文件加入到系统能找到的 lib 文件夹里， 把头文件加到系统能找到的 include 文件夹里就可以。这里把他们加到 CUDA 的文件夹下

• CUDA 是直接安装的、OpenCV 安装的时候是依赖于 CUDA 的、cuDNN 的安装就是将其 lib文件加入到系统能找到的lib文件夹里， 把头文件加到系统能找到的include文件夹里，这里把他们加到CUDA的文件夹下

• 所以说 cuDNN 无关于 CUDA ，无关于 OpenCV ，它只和 caffe 有关系，因为 caffe 安装的时候要依赖它。

重装过程

• cuDNN 文件结构

  root@Jedy:/home/jedy/deeplearning/cuda# tree -N
  .
  ├── include
  │   └── cudnn.h
  └── lib64
      ├── libcudnn.so
      ├── libcudnn.so.5
      ├── libcudnn.so.5.0.5
      └── libcudnn_static.a
  
  2 directories, 5 files

  而安装 cuDNN 的命令是

  tar -zxvf cudnn-7.0-linux-x64-v3.0-prod.tgz 
  cd cuda
  sudo cp lib64/* /usr/local/cuda/lib64/
  sudo cp include/cudnn.h /usr/local/cuda/include/
  
  #更新软链接
  cd /usr/local/cuda/lib64
  sudo rm -rf libcudnn.so libcudnn.so.7.0
  sudo ln -s libcudnn.so.7.0.64 libcudnn.so.7.0
  sudo ln -s libcudnn.so.7.0 libcudnn.so

• 重装 cuDNN

  • 因为 cuDNN 只是将文件拷贝进去，所以我们只需要下载更新版本的 cuDNN 然后覆盖安装即可。我是打算一步到位，使用最新的 caffe ，又怕 cuDNN 向下兼容不好，所以下载了 v5 版本，版本为 cudnn-8.0-linux-x64-v5.0-ga.tgz

  • 进入到 /usr/local/cuda/lib64 中，删除旧版本的 cuDNN

    cd /usr/local/cuda/lib64
    sudo rm -rf libcudnn.so libcudnn.so.7.0 libcudnn.so.7.0.64

  • 解压新版本的 cuDNN 并拷贝

    tar -zxvf cudnn-8.0-linux-x64-v5.0-ga.tgz
    cd cuda
    sudo cp lib64/* /usr/local/cuda/lib64/
    sudo cp include/cudnn.h /usr/local/cuda/include/

    这里的 include/cudnn.h 我们并没有删除旧版本的，而是采用了覆盖安装的方式

  • 更新软链接

    cd /usr/local/cuda/lib64
    sudo rm -rf libcudnn.so libcudnn.so.5
    sudo ln -s libcudnn.so.5.0.5 libcudnn.so.5
    sudo ln -s libcudnn.so.5 libcudnn.so

• 重装 caffe

  • 在 caffe 的 root 目录执行

    make clean

  • 删除旧版本的 caffe

  • 下载新版本的 caffe ，我这里直接下载了 1.0 版本的 caffe

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

    cp Makefile.config.example Makefile.config

  • 修改 Makefile.config ，开启 CuDNN 和 Opencv3 的支持即可，修改好的如下

    ## 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 := 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 3
    OPENCV_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 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
    
    # 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
    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
    
    # 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 := 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

    make all -j4
    make test -j4
    make runtest -j4

  • 没有报错，成功

    [----------] Global test environment tear-down
    [==========] 2101 tests from 277 test cases ran. (371248 ms total)
    [  PASSED  ] 2101 tests.

修改后的组件版本

• Ubuntu
  • 14.04

• CUDA
  • 8.0
• cuDNN
  • cudnn-8.0-linux-x64-v5.0-ga.tgz
• opencv
  • 3.2
• python
  • Python 2.7.6
• caffe
  • 1.0

一点心得

• make 编译的时候可以通过

  make -j4

  指令指定多线程编译，可以很大程度加快编译速度。具体可以开多少看自己机器配置

• 看好OpenCV支持的CUDA版本

• 看好caffe支持的cuDNN版本

• 每个人硬件、环境的不同都会导致踩到不同的坑。这个就多 Google 吧！

部分参考 Ubuntu 14.04上安装caffe