utuntu安装tensorflow-gpu

为什么需要先安装编译器GCC和G++，而不是先安装显卡驱动和cuda等等呢？
首先，先安装显卡驱动，后安装GCC并降级的安装顺序过程中，遇到了很多问题。比如，在安装CUDA SAMPLES的过程中，遇到了问题/usr/bin/ld: cannot find –lGL，后面通过网上教程，重新连接该库文件，仍然不能通过OpenCV的编译。但是，先安装GCC，再安装显卡驱动，就不会遇到这个问题了。

Driver:   Not Selected
Toolkit:  Installed in /usr/local/cuda-8.0
Samples:  Installed in /home/dsp, but missing recommended libraries

Please make sure that
 -   PATH includes /usr/local/cuda-8.0/bin
 -   LD_LIBRARY_PATH includes /usr/local/cuda-8.0/lib64, or, add /usr/local/cuda-8.0/lib64 to /etc/ld.so.conf and run ldconfig as root

To uninstall the CUDA Toolkit, run the uninstall script in /usr/local/cuda-8.0/bin

Please see CUDA_Installation_Guide_Linux.pdf in /usr/local/cuda-8.0/doc/pdf for detailed information on setting up CUDA.

***WARNING: Incomplete installation! This installation did not install the CUDA Driver. A driver of version at least 361.00 is required for CUDA 8.0 functionality to work.
To install the driver using this installer, run the following command, replacing <CudaInstaller> with the name of this run file:
    sudo <CudaInstaller>.run -silent -driver

Logfile is /tmp/cuda_install_3147.log

打开~/.bashrc文件：

sudo vim ~/.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}}

下载cuDNN5.1之后进行解压，cd进入cuDNN5.1解压之后的include目录，在命令行进行如下操作：

sudo cp cudnn.h /usr/local/cuda/include/    #复制头文件

再将cd进入lib64目录下的动态文件进行复制和链接：

sudo cp lib* /usr/local/cuda/lib64/    #复制动态链接库
cd /usr/local/cuda/lib64/
sudo rm -rf libcudnn.so libcudnn.so.5    #删除原有动态文件
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

tensorflow安装所遇问题：libcudnn.so.6:cannot open sharedobject file: No such file or directory

 【报错】ImportError: libcudnn.so.6: cannot open shared object file: No such file or directory

问题解决：

pip uninstall tensorflow-gpu
pip install tensorflow-gpu==1.2

下载完毕后，在terminal中到达文件所在位置

bash Ana.......(快捷键Tab) .sh

一路按回车键，中间记得输入一个yes就可以顺利完成安装。

安装完后，配置环境路径:安装过程自动选择添加，就不需要手动添加

# 将anaconda的bin目录加入PATH，根据版本不同，也可能是~/anaconda3/bin
echo ‘export PATH="~/anaconda2/bin:$PATH"‘ >> ~/.bashrc
# 更新bashrc以立即生效
source ~/.bashrc

dsp@dsp:~/下载/cuda/lib64$ python
Python 2.7.12 (default, Nov 19 2016, 06:48:10) 
[GCC 5.4.0 20160609] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> exit()
dsp@dsp:~/下载/cuda/lib64$ source /home/dsp/.bashrc
dsp@dsp:~/下载/cuda/lib64$ python
Python 2.7.14 |Anaconda, Inc.| (default, Oct 16 2017, 17:29:19) 
[GCC 7.2.0] on linux2
Type "help", "copyright", "credits" or "license" for more information.

# 创建一个名为tenseorflow的环境，指定Python版本是3.5
conda create --name tensorflow python=3.5

# 安装好后，使用activate激活某个环境
activate tensorflow  # for Windows
source activate tensorflow # for Linux & Mac
# 激活后，会发现terminal输入的地方多了tensorflow 的字样，实际上，此时系统做的事情就是把默认2.7环境从PATH中去除，再把3.5对应的命令加入PATH

# 此时，再次输入
python --version
# 可以得到`Python 3.5.3 :: Continuum Analytics, Inc.`

# 如果想返回默认的python 2.7环境，运行
deactivate tensorflow # for Windows
source deactivate tensorflow# for Linux & Mac
#可以得到`Python 2.7.13 :: Anaconda 4.3.1 (64-bit)`

# 删除一个已有的环境
conda remove --name tensorflow --all

(tensorflow_py3.5) dsp@dsp:/usr/local/cuda/lib64$ python
Python 3.5.4 |Anaconda, Inc.| (default, Nov  3 2017, 20:01:27) 
[GCC 7.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow as tf
>>>