Ubuntu16.04 + gtx1060 + cuda8.0 + cudnn5.1 + caffe + Theano + Tensorflow

参考 ubuntu16.04+gtx1060+cuda8.0+caffe安装、测试经历 ，细节处有差异。

首先说明，这是在台式机上的安装测试经历，首先安装的win10，然后安装ubuntu16.04双系统，显卡为GTX1060 
台式机显示器接的是GTX1060 HDMI口，win10上首先安装了最新的GTX1060驱动375

---

废话不多说，上车吧，少年

一、首先安装nvidia显卡驱动

1. 我是1080P的显示器，在没有安装显卡驱动前，ubuntu分辨率很低，可以手动修改一下grub文件，提高分辨率，在终端输入

   sudo vim /etc/default/grub
   找到以下行 
   # The resolution used on graphical terminal
# note that you can use only modes which your graphic card supports via VBE
# you can see them in real GRUB with the command 'vbeinfo'
# GRUB_GFXMODE=640×480
   按a进入插入模式，增加下面一行 
   GRUB_GFXMODE=1920×1080 #这里分辨率自行设置 
   按esc退出插入模式，按:wq保存退出 
   在终端编辑 
   sudo update-grub
   更新grub 
   重新启动ubuntu使之生效

2. 进入ubuntu系统设置-软件与更新-附加驱动

   
   安装之后重启系统让GTX1060显卡驱动生效

3. 测试

   终端输入 
   nvidia-smi 
   显示效果如下图表示安装成功 
   

---

二、cuda安装

1. 下载cuda_8.0.61_375.26_linux.run 和 cudnn-8.0-linux-x64-v5.1.tgz

   这里我提供了百度网盘，这两个文件我先在win10下下载好，并用u盘拷贝到ubuntu的下载目录下

2. 安装cuda8.0

   终端输入 
   cd 下载/ 
sh cuda_8.0.27_linux.run --override 
   启动安装程序，一直按空格到最后，输入accept接受条款 (或者按 Q)
   输入n不安装nvidia图像驱动，之前已经安装过了 
   输入y安装cuda 8.0工具 
   回车确认cuda默认安装路径：/usr/local/cuda-8.0 
   输入y用sudo权限运行安装，输入密码 
   输入y或者n安装或者不安装指向/usr/local/cuda的符号链接 
   输入y安装CUDA 8.0 Samples，以便后面测试 
   回车确认CUDA 8.0 Samples默认安装路径：/home/yt（yt是我的用户名），该安装路径测试完可以删除

3. 安装cudnn v5.1

   终端输入 
   

   cd 下载/
   tar zxvf cudnn-8.0-linux-x64-v5..tgz

   解压在下载目录下产生一个cuda目录

   cd cuda
   sudo cp lib64/* /usr/local/cuda/lib64/　　　　　    #复制头文件
   sudo cp include/cudnn.h /usr/local/cuda/include/ 　#复制动态链接库
   
   sudo chmod a+r /usr/local/cuda/include/cudnn.h
   sudo chmod a+r /usr/local/cuda/lib64
   sudo chmod a+r /usr/local/cuda/lib64/libcudnn*   #给所有用户增加这些文件的读权限  

4. 建立软链接

   终端输入

   cd /usr/local/cuda/lib64/
   sudo rm -rf libcudnn.so libcudnn.so.
   sudo ln -s libcudnn.so.5.1. libcudnn.so.   #具体看版本
   sudo ln -s libcudnn.so. libcudnn.so 

   设置环境变量，终端输入

   sudo gedit /etc/profile 

   在末尾加入

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

   保存后，创建链接文件

   sudo vim /etc/ld.so.conf.d/cuda.conf 

   按a进入插入模式，增加下面一行

   /usr/local/cuda/lib64 

   按esc退出插入模式，按:wq保存退出 
   最后在终端输入sudo ldconfig使链接生效

5. cuda Samples测试

   打开CUDA 8.0 Samples默认安装路径，终端输入 
   cd /home/yt/NVIDIA_CUDA-8.0_Samples （yt是我的用户名） 
   sudo make all -j4 （4核） 
   出现“unsupported GNU version! gcc versions later than 5.3 are not supported!”的错误，这是由于GCC版本过高，在终端输入 
   cd /usr/local/cuda-8.0/include 
sudo cp host_config.h host_config.h.bak 
sudo gedit host_config.h 
   ctrl+f寻找有“5.3”的地方，只有一处，如下 
   # if __GNUC__ > 5 || (__GNUC__ == 5 && __GNUC_MINOR__ > 3) 
#error -- unsupported GNU version! gcc versions later than 5.3 are not supported! 
   将两个5改成6，即 
   #if __GNUC__ > 6 || (__GNUC__ == 6 && __GNUC_MINOR__ > 3) 
   保存退出，继续在终端输入 
   cd /home/yt/NVIDIA_CUDA-8.0_Samples （yt是我的用户名） 
   sudo make all -j4 （4核） 
   完成后继续向终端输入 
   cd bin/x86_64/linux/release 
./deviceQuery 
   完成之后出现如下图所示，表示成功安装cuda 
   

---

三、依赖包安装

1. sudo apt-get install build-essential #必要的编译工具依赖

2. sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compiler

3. sudo apt-get install --no-install-recommends libboost-all-dev

4. sudo apt-get install libatlas-base-dev

5. sudo apt-get install libgflags-dev libgoogle-glog-dev liblmdb-dev

---

四

1. 安装python的pip和easy_install，方便安装软件包

   终端输入 
   cd 
wget --no-check-certificate https://bootstrap.pypa.io/ez_setup.py 
sudo python ez_setup.py --insecure 
wget https://bootstrap.pypa.io/get-pip.py 
sudo python get-pip.py

---

五

1. 安装科学计算和python所需的部分库

   终端输入 
   sudo apt-get install libblas-dev liblapack-dev libatlas-base-dev gfortran python-numpy

---

六

1. 安装git，拉取源码

   终端输入 
   sudo apt-get install git 
git clone https://github.com/BVLC/caffe.git

---

七

1. 安装python依赖

   终端输入 
   sudo apt-get install python-pip 安装pip 
   cd /home/yt/caffe/python
   sudo su 
for req in $(cat "requirements.txt"); do pip install -i https://pypi.tuna.tsinghua.edu.cn/simple $req; done
   按Ctrl+D退出sudo su模式

---

八、编译caffe(暂不对matlab说明)

1. 终端输入 
   cd /home/yt/caffe 
cp Makefile.config.example Makefile.config 
gedit Makefile.config

   ①将USE_CUDNN := 1取消注释，

   ②INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include后面打上一个空格 然后添加/usr/include/hdf5/serial如果没有这一句可能会报一个找不到hdf5.h的错误

2. 终端输入 
   make all -j4 
   make过程中出现找不到lhdf5_hl和lhdf5的错误， 
   解决方案： 
   在计算机中搜索libhdf5_serial.so.10.1.0，找到后右键点击打开项目位置 
   该目录下空白处右键点击在终端打开，打开新终端输入 
   sudo ln libhdf5_serial.so.10.1.0 libhdf5.so 
sudo ln libhdf5_serial_hl.so.10.0.2 libhdf5_hl.so 
   最后在终端输入sudo ldconfig使链接生效 
   原终端中输入make clean清除第一次编译结果 
   再次输入make all -j4重新编译

3. 终端输入

   make test -j4
   make runtest -j4
   make pycaffe -j4
   make distribute   #生成发布安装包

4. 测试python，终端输入 
   pip install protobuf  -i https://pypi.tuna.tsinghua.edu.cn/simple pyspider
   cd /home/yt/caffe/python 
python 
import caffe 
   如果不报错就说明编译成功

---

九、mnist测试

1. 下载mnist数据集，终端输入 
   cd /home/yt/caffe/data/mnist/
./get_mnist.sh 获取mnist数据集 
   在/home/yt/caffe/data/mnist/目录下会多出训练集图片、训练集标签、测试集图片和测试集标签等4个文件

2. mnist数据格式转换，终端输入 
   cd /home/yt/caffe/
./examples/mnist/create_mnist.sh
   必须要在第一行之后运行第二行，即必须要在caffe根目录下运行create_mnist.sh 
   此时在/caffe/examples/mnist/目录下生成mnist_test_lmdb和mnist_train_lmdb两个LMDB格式的训练集和测试集

3. LeNet-5模型描述在/caffe/examples/mnist/lenet_train_test.prototxt

4. Solver配置文件在/caffe/examples/mnist/lenet_solver.prototxt

5. 训练mnist，执行文件在/caffe/examples/mnist/train_lenet.sh
   终端输入 
   cd /home/yt/caffe/
./examples/mnist/train_lenet.sh
   测试结果如下 
   

十、安装theano

1、直接输入命令：

sudo pip install theano 

2、配置参数文件：.theanorc

sudo gedit ~/.theanorc

[global]
floatX=float32
device=gpu
base_compiledir=~/external/.theano/
allow_gc=False
warn_float64=warn
[mode]=FAST_RUN  

[nvcc]
fastmath=True  

[cuda]
root=/usr/local/cuda 

3、运行测试例子：

from theano import function, config, shared, sandbox
import theano.tensor as T
import numpy
import time  

vlen = 10 * 30 * 768  # 10 x #cores x # threads per core
iters = 1000  

rng = numpy.random.RandomState(22)
x = shared(numpy.asarray(rng.rand(vlen), config.floatX))
f = function([], T.exp(x))
print(f.maker.fgraph.toposort())
t0 = time.time()
for i in range(iters):
    r = f()
t1 = time.time()
print("Looping %d times took %f seconds" % (iters, t1 - t0))
print("Result is %s" % (r,))
if numpy.any([isinstance(x.op, T.Elemwise) for x in f.maker.fgraph.toposort()]):
    print('Used the cpu')
else:
    print('Used the gpu') 

十、安装tensosrflow

1、安装编译工具Bazel

echo "deb [arch=amd64] http://storage.googleapis.com/bazel-apt stable jdk1.8" | sudo tee /etc/apt/sources.list.d/bazel.list 

curl https://storage.googleapis.com/bazel-apt/doc/apt-key.pub.gpg | sudo apt-key add -  

sudo apt-get update && sudo apt-get install bazel  

sudo apt-get upgrade bazel

2、下载tensorflow并编译

git clone https://github.com/tensorflow/tensorflow
cd tensorflow
git checkout Branch # where Branch is the desired branch
git checkout r1.

sudo apt-get install python-numpy python-dev python-pip python-wheel
sudo apt-get install python3-numpy python3-dev python3-pip python3-wheel

sudo apt-get install libcupti-dev 

./configure  

$ ./configure # 以下是一个例子
Please specify the location of python. [Default is /usr/bin/python]: y
Invalid python path. y cannot be found
Please specify the location of python. [Default is /usr/bin/python]:
Please specify optimization flags to use during compilation when bazel option "--config=opt" is specified [Default is -march=native]:
Do you wish to use jemalloc as the malloc implementation? [Y/n] y
jemalloc enabled
Do you wish to build TensorFlow with Google Cloud Platform support? [y/N] n
No Google Cloud Platform support will be enabled for TensorFlow
Do you wish to build TensorFlow with Hadoop File System support? [y/N] y
Hadoop File System support will be enabled for TensorFlow
Do you wish to build TensorFlow with the XLA just-in-time compiler (experimental)? [y/N] n
No XLA JIT support will be enabled for TensorFlow
Found possible Python library paths:
  /usr/local/lib/python2.7/dist-packages
  /usr/lib/python2.7/dist-packages
Please input the desired Python library path to use.  Default is [/usr/local/lib/python2.7/dist-packages]

Using python library path: /usr/local/lib/python2.7/dist-packages
Do you wish to build TensorFlow with OpenCL support? [y/N] n
No OpenCL support will be enabled for TensorFlow
Do you wish to build TensorFlow with CUDA support? [y/N] y
CUDA support will be enabled for TensorFlow
Please specify which gcc should be used by nvcc as the host compiler. [Default is /usr/bin/gcc]:
Please specify the CUDA SDK version you want to use, e.g. 7.0. [Leave empty to use system default]: 8.0
Please specify the location where CUDA 8.0 toolkit is installed. Refer to README.md for more details. [Default is /usr/local/cuda]:
Please specify the Cudnn version you want to use. [Leave empty to use system default]: 5
Please specify the location where cuDNN 5 library is installed. Refer to README.md for more details. [Default is /usr/local/cuda]:
Please specify a list of comma-separated Cuda compute capabilities you want to build with.
You can find the compute capability of your device at: https://developer.nvidia.com/cuda-gpus.
Please note that each additional compute capability significantly increases your build time and binary size.
[Default is: "3.5,5.2"]: 6.1
INFO: Starting clean (this may take a while). Consider using --expunge_async if the clean takes more than several minutes.
........
INFO: All external dependencies fetched successfully.
Configuration finished

编译

bazel build --config=opt --config=cuda //tensorflow/tools/pip_package:build_pip_package --cxxopt="-D_GLIBCXX_USE_CXX11_ABI=0"

bazel-bin/tensorflow/tools/pip_package/build_pip_package /tmp/tensorflow_pkg 

检查tmp文件夹下生成的whl文件名

sudo pip install /tmp/tensorflow_pkg/ tensorflow-1.0.-cp27-cp27mu-linux_x86_64.whl

3、测试

python
import tensorflow as tf
sess = tf.Session()