tensorflow:ubuntu16.04 gpu版本安装

1、cuda和cudnn的区别

CUDA is NVIDIA’s language/API for programming on the graphics card. I’ve found it to be the easiest way to write really high performance programs run on the GPU.cuDNN is a library for deep neural nets built using CUDA. It provides GPU accelerated functionality for common operations in deep neural nets. You could use it directly yourself, but other libraries like TensorFlow already have built abstractions backed by cuDNN.

2、验证你的GPU能够运行CUDA

$ lspci | grep -i nvidia

3、查看显卡驱动是否安装

首先得安装mesa-utils，在终端输入命令：sudo apt-get install mesa-utils
然后再运行命令：glxinfo | grep rendering

如果结果是“yes”，证明显卡 驱动已经成功安装。

4、安装显卡驱动

虽然ubuntu安装时已经安装了显卡驱动，但是这个驱动和后面的CUDA不知道是不兼容还是版本过低（有了解的同学帮忙解答一下啊），通过nvidia-smi命令，发现显示command cannot found，所以需要重新安装驱动。

sudo apt purge nvidia-*
sudo add-apt-repository ppa:graphics-drivers/ppa
sudo apt update
sudo apt install nvidia-381

sudo reboot

重启后在通过nvidia-smi命令，可以显示NVIDIA显卡的信息

5、安装CUDA

首先去NVIDA官网下载CUDA8.0版本的runfile文件，有1.5G。


下载完成后，通过命令sudo sh cuda_8.0.27_linux.run执行安装。安装过程中有需注意Install NVIDIA Accelerated Graphics Driver for Linux-x86_64 361.62选择n，因为之前已经安装了（这里如果前面不安装显卡驱动，而在这里直接选则y不知道是否可行？没有尝试过。如果前面不装驱动，这里选n是不行的。）

Do you accept the previously read EULA?accept/decline/quit: accept

Install NVIDIA Accelerated Graphics Driver for Linux-x86_64 361.62?(y)es/(n)o/(q)uit: n

Install the CUDA 8.0 Toolkit?(y)es/(n)o/(q)uit: y

Enter Toolkit Location[ default is /usr/local/cuda-8.0 ]:

Do you want to install a symbolic link at /usr/local/cuda?(y)es/(n)o/(q)uit: y

Install the CUDA 8.0 Samples?(y)es/(n)o/(q)uit: y

Enter CUDA Samples Location[ default is /home/textminer ]:

6、安装完成后添加环境变量，将下面语句写入到 ~/.bashrc 的尾部:

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}}

7、安装cuDNN

如果要使用gpu来对tensorflow进行加速，除了安装CUDA以外，cuDNN也是必须要安装的。跟cuda一样，去nvidia的官网下载cuDNN的安装包。不过这次没法直接下载，需要先注册，然后还要做个调查问卷什么的，稍微有点麻烦。我下的是cuDNN v5.1 Library for Linux这个版本。不要下cuDNN v5.1 Developer Library for Ubuntu16.04 Power8 (Deb)这个版本,因为是给powe8处理器用的，不是amd64.

下载下来以后，发现是一个tgz的压缩包，使用tar进行解压，命令为：

tar -xvf cudnn-8.0-linux-x64-v5.1 .tgz
安装cuDNN比较简单，解压后把相应的文件拷贝到对应的CUDA目录下即可


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

sudo cp cuda/lib64/libcudnn* /usr/local/cuda/lib64/

sudo chmod a+r /usr/local/cuda/include/cudnn.h

sudo chmod a+r /usr/local/cuda/lib64/libcudnn*

最好下载低一个版本的，高版本的有时会有坑，最好下载压缩包的cuDNN v5.1 Library for Linux，而不是deb的。
8、安装tensorflow_gpu版本
如果之前已经安装了cpu-only版的tensorflow，所以现在要先把原先的tf卸载，即使以前装的是gpu版本，但是cuda、cudnn等没有装，依然装的是cpu版本，需要卸载重新装：
sudo pip uninstall tensorflow
之后从https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow_gpu-0.12.0rc1-cp35-cp35m-linux_x86_64.whl下载gpu版的tf


sudo pip3 install tensorflow_gpu-0.12 .0rc1-cp35-cp35m-linux_x86_64.whl
来测试一下


$ipython3

import

nvidia-docker

tensorflow as tf
结果如下：
Python 3.5.2 (default, Nov 17 2016, 17:05:23)
[GCC 5.4.0 20160609] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow as tf
I tensorflow/stream_executor/dso_loader.cc:128] successfully opened CUDA library libcublas.so locally
I tensorflow/stream_executor/dso_loader.cc:128] successfully opened CUDA library libcudnn.so locally
I tensorflow/stream_executor/dso_loader.cc:128] successfully opened CUDA library libcufft.so locally
I tensorflow/stream_executor/dso_loader.cc:128] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:128] successfully opened CUDA library libcurand.so locally
>>> exit()

则表示安装成功。

如果使用docker进行安装，在安装完cuda和cudnn后的步骤如下：
1、安装docker


$ sudo apt-get update

$ sudo apt-get install \
    apt-transport-https \
    ca-certificates \
    curl \
    software-properties-common

$ curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -

$ sudo apt-key fingerprint 0EBFCD88

$ sudo add-apt-repository \
   "deb [arch=amd64] https://download.docker.com/linux/ubuntu \
   $(lsb_release -cs) \
   stable"

$ sudo add-apt-repository \
   "deb [arch=armhf] https://download.docker.com/linux/ubuntu \
   $(lsb_release -cs) \
   stable"

$ sudo add-apt-repository \
   "deb [arch=s390x] https://download.docker.com/linux/ubuntu \
   $(lsb_release -cs) \
   stable"

$ sudo apt-get update

$ sudo apt-get install docker-ce

$ apt-cache madison docker-ce

$ sudo apt-get install docker-ce=<VERSION>

$ sudo docker run hello-world

2、安装nvidia-docker

下载nvidia-docker的安装包 https://github.com/NVIDIA/nvidia-docker/releases
   #deb类型: https://github.com/NVIDIA/nvidia-docker/releases/download/v1.0.0/nvidia-docker_1.0.0-1_amd64.deb
可以在上面的网址下直接下载压缩包，解压后放到/usr/bin目录下就可以了
由于我使用的gnome桌面的ubuntu系统，对于deb文件的安装总出问题，所以安装了GDebi工具用于安装deb包 

3、安装tensorflow gpu
nvidia-docker run -it gcr.io/tensorflow/tensorflow:latest-gpu bash
4、安装好后可以进行验证
docker run -it gcr.io/tensorflow/tensorflow bash
python
# Python
import tensorflow as tf
hello = tf.constant('Hello, TensorFlow!')
sess = tf.Session()
print(sess.run(hello))