Ubuntu install TensorFlow 1.10 + CUDA 9.2 + cuDNN 7.2

 

为了装TensorFlow 1.10 下面升级一下系统的软件环境

NVIDIA DRIVER

去官网下载最新的linux驱动   http://www.nvidia.com/Download/index.aspx 

 

直接运行会报错

 

sudo bash NVIDIA-Linux-x86_64-390.87.run

ERROR: You appear to be running an X server; please exit X before
         installing. For further details, please see the section INSTALLING
         THE NVIDIA DRIVER in the README available on the Linux driver
         download page at www.nvidia.com.

需要先关闭图形界面，在另一台电脑上用ssh登录这台电脑然后运行

sudo init
sudo killall Xorg

然后再运行

sudo bash NVIDIA-Linux-x86_64-390.87.run

装好后运行

nvidia-smi

出现下图结果说明成功安装

再运行下面命令恢复图形界面

sudo init 

可以重启一下确认显卡驱动是否正常

 

如果需要改gcc 或g++版本 请参考上一篇博文 https://www.cnblogs.com/jins-note/p/9597210.html

 

 

CUDA 9.2

由于TensorFlow 1.10 支持cuda 9.2 

去官网下载最新版本 

https://developer.nvidia.com/cuda-downloads?target_os=Linux&target_arch=x86_64&target_distro=Ubuntu&target_version=1710&target_type=runfilelocal

先安装一些推荐库

sudo apt-get install freeglut3-dev build-essential libx11-dev libxmu-dev libxi-dev libgl1-mesa-glx libglu1-mesa libglu1-mesa-dev libglfw3-dev libgles2-mesa-dev

这里注意：cuda里带的驱动比刚从官网下的新，那么用cuda里带的驱动（居然比官网下的显卡驱动新？？）

不然会报错 

cudaerrorinsufficientdriver

然后安装

sudo init
sudo killall Xorg
sudo bash cuda_9..148_396.37_linux.run

安装过程如下

 

Description

The NVIDIA CUDA Toolkit provides command-line and graphical
Do you accept the previously read EULA?
accept/decline/quit: accept

Install NVIDIA Accelerated Graphics Driver for Linux-x86_64 396.37?
(y)es/(n)o/(q)uit: y

Install the CUDA 9.2 Toolkit?
(y)es/(n)o/(q)uit: y

Enter Toolkit Location
 [ default is /usr/local/cuda-9.2 ]:

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

Install the CUDA 9.2 Samples?
(y)es/(n)o/(q)uit: y

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

安装 补丁

sudo bash cuda_9.2.148.1_linux.run

装好后修改环境变量 ~/.bashrc 在末尾添加

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

修改完毕之后执行一下使其生效：

source ~/.bashrc

恢复显示

sudo init 

装好后去 CUDA Samples  目录编译一些例子看看能不能运行，能运行就ok

cd ~/NVIDIA_CUDA-.2_Samples/
make -j8

编译好后去下面目录里运行

cd bin/x86_64/linux/release

cuDNN

去官网下载对应版本 https://developer.nvidia.com/rdp/cudnn-download 需要登录才能下载

 

下载后的文件后缀名应该是 *.tgz 如果 是 .solitairetheme8 那就改成 .tgz

安装很简单，解压复制到对应目录就好

tar -zxvf cudnn-9.2-linux-x64-v7.2.1..tgz
sudo cp cuda/include/cudnn.h /usr/local/cuda/include/
sudo cp cuda/lib64/libcudnn* /usr/local/cuda/lib64/ -d
sudo chmod a+r /usr/local/cuda/include/cudnn.h
sudo chmod a+r /usr/local/cuda/lib64/libcudnn*

TensorFlow 1.10

安装 Anaconda   从这里下载 https://www.anaconda.com/download/

更换为国内源 https://mirrors.ustc.edu.cn/help/anaconda.html

conda config --add channels https://mirrors.ustc.edu.cn/anaconda/pkgs/free/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/pkgs/main/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/cloud/conda-forge/
conda config --set show_channel_urls yes

然后安装

conda install tensorflow-gpu==1.10

装好后测试

import tensorflow as tf
tf.__version__

 

 

 

 

 

参考： https://cuiqingcai.com/5822.html