Ubunt16.04 CUDA8.0+caffe+tensorflow+gpu

我的电脑配置比较垃圾：3代i5+GT750m

装过不下50次caffe和ubuntu的我建议先装tensorflow-gpu（cuda9.1不支持tensorflow-gpu）再来caffe-gpu

ubuntu16.04自带python2.7和3.5

tensorflow-gpu安装非常简单

pip install tensorflow-gpu == 1.x.x #也可以前面加sudo,后面==是要安装的版本，按错了直接换就行

pip install tensorflow == 1.x.x  #这是cpu的

配套tensorflow的肯定就是Keras和theano

pip install keras == 2.1.x
pip install theano

注意：有些显卡驱动版本不兼容，必须要配套的tensorflow，所以省时间报错就换版本 我的是tensorflow1.2 keras2.1.3

CPU的随便按 pip install tensorflow keras 这样可以安装俩的最新版

当然，python里面一堆乱七八糟的包要按好，比如numpy之类的。。。

而且ubuntu17.04安装caffe可以像上面按tensorflow一样。

一、安装nVidia驱动，配置cuda，cudnn运行环境

1.安装相关依赖项

sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev> libopencv-dev libhdf5-serial-dev protobuf-compiler  
sudo apt-get install --no-install-recommends libboost-all-dev  
sudo apt-get install libopenblas-dev liblapack-dev libatlas-base-dev  
sudo apt-get install libgflags-dev libgoogle-glog-dev liblmdb-dev

扫描二维码关注公众号，回复： 1569522 查看本文章

2.安装NVIDIA驱动

  驱动最简单的方法是在Ubuntu系统自带的设置里更改，去官网下载的驱动会比较崩还麻烦。

3.安装CUDA

首先在官网上( https://developer.nvidia.com/cuda-downloads )下载CUDA，选择自己合适的版本。 

该链接界面只显示最新版本。若需要下载以前的版本，可在页面下方点击Legacy Releases，选择自己需要的其他版本。

下载完成后执行以下命令：

sudo chmod 777 cuda_8.0.44_linux.run 
sudo ./cuda_8.0.44_linux.run

安装时首先会有一系列提示让你确认到100%，但是注意，有个让你选择是否安装nvidia361驱动时，一定要选择否：

Install NVIDIA Accelerated Graphics Driver for Linux-x86_64 361.62?

这个东西是CUDA自带的驱动，可能和你的显卡不兼容。上面装驱动的方法是我觉得最简单不烦人的方法。

安装cuda时可能有下面的信息

Installing the CUDA Toolkit in /usr/local/cuda-8.0 … 
Missing recommended library: libGLU.so 
Missing recommended library: libX11.so 
Missing recommended library: libXi.so 
Missing recommended library: libXmu.so

原因是缺少相关的依赖库,安装相应库就解决了：

sudo apt-get install freeglut3-dev build-essential libx11-dev libxmu-dev libxi-dev libgl1-mesa-glx libglu1-mesa libglu1-mesa-dev

环境变量配置：

打开~/.bashrc文件： sudo gedit ~/.bashrc 将以下内容写入到~/.bashrc尾部：

export PATH=/usr/local/cuda-8.0/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda-8.0/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

测试CUDA的samples：

cd /usr/local/cuda-8.0/samples/1_Utilities/deviceQuery 
make 
sudo ./deviceQuery

如果显示一些乱七八糟的关于GPU的信息，则说明安装成功。

4.配置cuDNN

cuDNN是GPU加速计算深层神经网络的库。 首先去官网https://developer.nvidia.com/rdp/cudnn-download 下载cuDNN，需要注册一个账号才能下载。

下载cuDNN5.1 之后进行解压：

cd cuda/include
sudo cp cudnn.h /usr/local/cuda/include  

再进入lib64目录下的动态文件进行复制和链接：

cd ..
cd lib64 
sudo cp lib* /usr/local/cuda/lib64/

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

sudo ln -s libcudnn.so.5.1.10 libcudnn.so.5  #生成软衔接（注意这里要和自己下载的cudnn版本对应，可以在/usr/local/cuda/lib64下查看自己libcudnn的版本）
sudo ln -s libcudnn.so.5 libcudnn.so   #生成软链接

执行

sudo ldconfig -v  #必须在/usr/local/cuda/lib64/目录下执行，否则可能会报libcudnn.so.5: cannot open shared object file: No such file or directory

或者

sudo ldconfig /usr/local/cuda/lib64 #cudnn就配置完成了。

三、配置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

根据个人情况修改文件： 
a.若使用cudnn，则

将
#USE_CUDNN := 1
修改成： 
USE_CUDNN := 1

b.若使用的opencv版本是3的，则

将
#OPENCV_VERSION := 3 
修改为： 
OPENCV_VERSION := 3

c.若要使用python来编写layer，则

将
#WITH_PYTHON_LAYER := 1 
修改为 
WITH_PYTHON_LAYER := 1

d. 重要的一项 : 
将# Whatever else you find you need goes here.下面的

INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib 
修改为： 
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu /usr/lib/x86_64-linux-gnu/hdf5/serial    

这是因为ubuntu16.04的文件包含位置发生了变化，尤其是需要用到的hdf5的位置，所以需要更改这一路径。 
这里贴出我修改后的Makefile.config文件：

## 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  #使用cuDNN

# 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  #使用openCV3

# 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  #CUDA路径
# 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.
# For CUDA >= 9.0, comment the *_20 and *_21 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  #计算能力

# BLAS choice:
# atlas for ATLAS (default)
# mkl for MKL
# open for OpenBlas
BLAS := atlas  #使用哪种BLAS：atlas or mkl or openblas
# 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)/anaconda2  #使用anaconda，anaconda的路径
PYTHON_INCLUDE := $(ANACONDA_HOME)/include \
        $(ANACONDA_HOME)/include/python2.7 \
        $(ANACONDA_HOME)/lib/python2.7/site-packages/numpy/core/include  #anaconda的python头文件路径

# 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  #anaconda的python库目录

# 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  #使用python层

# Whatever else you find you need goes here.
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial  #头文件路径
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu /usr/lib/x86_64-linux-gnu/hdf5/serial  #库路径

# 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  #GPU编号

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

（5）修改makefile文件 
打开makefile文件，做如下修改：

将：
NVCCFLAGS +=-ccbin=$(CXX) -Xcompiler-fPIC $(COMMON_FLAGS)
替换为：
NVCCFLAGS += -D_FORCE_INLINES -ccbin=$(CXX) -Xcompiler -fPIC $(COMMON_FLAGS)

（6）编辑/usr/local/cuda/include/host_config.h 
将其中的第115行注释掉：

将
#error-- unsupported GNU version! gcc versions later than 4.9 are not supported!
改为
//#error-- unsupported GNU version! gcc versions later than 4.9 are not supported!

（7）编译

make all -j4 #-j根据自己电脑配置决定 -j2代表2线程 4为4线程 现在8代i7可以12线程
出错就make clean 重来

编译过程中可能会出现如下错误： 
错误内容1：

"fatal error: hdf5.h: 没有那个文件或目录"

解决办法： 
step1:在Makefile.config文件的第85行，添加/usr/include/hdf5/serial/ 到 INCLUDE_DIRS，也就是把下面第一行代码改为第二行代码。

将：
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include
替换为：
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial/

stept2:在Makefile文件的第173行，把 hdf5_hl 和hdf5修改为hdf5_serial_hl 和 hdf5_serial，也就是把下面第一行代码改为第二行代码。

将：
LIBRARIES += glog gflags protobuf boost_system boost_filesystem m hdf5_hl hdf5
改为：
LIBRARIES += glog gflags protobuf boost_system boost_filesystem m hdf5_serial_hl hdf5_serial

错误内容2：

"libcudart.so.8.0 cannot open shared object file: No such file or directory"

解决办法是将一些文件复制到/usr/local/lib文件夹下：

#注意自己CUDA的版本号！
sudo cp /usr/local/cuda-8.0/lib64/libcudart.so.8.0 /usr/local/lib/libcudart.so.8.0 && sudo ldconfig
sudo cp /usr/local/cuda-8.0/lib64/libcublas.so.8.0 /usr/local/lib/libcublas.so.8.0 && sudo ldconfig
sudo cp /usr/local/cuda-8.0/lib64/libcurand.so.8.0 /usr/local/lib/libcurand.so.8.0 && sudo ldconfig

编译caffe问题主要是cudnn的问题，一定要按相应的版本来匹配！！！

8）测试

make runtest -j4

这个4线程容易出错。

显示PASSED就ojbk了。到此caffe配置完毕！

MNIST数据集测试

cd ~/caffe #caffe的根目录

./data/mnist/get_mnist.sh   #下载MNIST数据库并解压缩./examples/mnist/create_mnist.sh   #将其转换成Lmdb数据库格式./examples/mnist/train_lenet.sh  #训练网络

这时候不出问题你就可以看到乱七八糟的代码往天上飞，显卡越好飞的越快。最后显示精度和loss。（GTX1050两分钟左右，我的GT 750m 大概18分钟，GTX1080大概就几十秒吧）

训练完成，caffe也就ok了！文章如有错我改。。。