ubuntu20.04 搭建深度学习环境tensorflow-gpu2.*

提示：文章写完后，目录可以自动生成，如何生成可参考右边的帮助文档

ubuntu20.04安装显卡驱动，cuda , cudnn, tensorflow-gpu2.*
1、源配置文件
备份源文件
sudo cp /etc/apt/sources.list /etc/apt/sources.list.bak
修改源文件
sudo gedit /etc/apt/sources.list
163源：
···
deb http://mirrors.163.com/ubuntu/ bionic main restricted universe multiverse
deb http://mirrors.163.com/ubuntu/ bionic-security main restricted universe multiverse
deb http://mirrors.163.com/ubuntu/ bionic-updates main restricted universe multiverse
deb http://mirrors.163.com/ubuntu/ bionic-proposed main restricted universe multiverse
deb http://mirrors.163.com/ubuntu/ bionic-backports main restricted universe multiverse
deb-src http://mirrors.163.com/ubuntu/ bionic main restricted universe multiverse
deb-src http://mirrors.163.com/ubuntu/ bionic-security main restricted universe multiverse
deb-src http://mirrors.163.com/ubuntu/ bionic-updates main restricted universe multiverse
deb-src http://mirrors.163.com/ubuntu/ bionic-proposed main restricted universe multiverse
deb-src http://mirrors.163.com/ubuntu/ bionic-backports main restricted universe multiverse

···
阿里：
···
deb http://mirrors.aliyun.com/ubuntu/ bionic main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu/ bionic-security main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu/ bionic-updates main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu/ bionic-proposed main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu/ bionic-backports main restricted universe multiverse
deb-src http://mirrors.aliyun.com/ubuntu/ bionic main restricted universe multiverse
deb-src http://mirrors.aliyun.com/ubuntu/ bionic-security main restricted universe multiverse
deb-src http://mirrors.aliyun.com/ubuntu/ bionic-updates main restricted universe multiverse
deb-src http://mirrors.aliyun.com/ubuntu/ bionic-proposed main restricted universe multiverse
deb-src http://mirrors.aliyun.com/ubuntu/ bionic-backports main restricted universe multiverse

···
清华：

# 默认注释了源码镜像以提高 apt update 速度，如有需要可自行取消注释
deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic main restricted universe multiverse
# deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic main restricted universe multiverse
deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-updates main restricted universe multiverse
# deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-updates main restricted universe multiverse
deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-backports main restricted universe multiverse
# deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-backports main restricted universe multiverse
deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-security main restricted universe multiverse
# deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-security main restricted universe multiverse
# 预发布软件源，不建议启用
# deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-proposed main restricted universe multiverse
# deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-proposed main restricted universe multiverse

保存并关闭文件
终端输入：
···
sudo apt update
sudo apt upgrade
···
换源后，安装build-essential（安装完成后终端输入gcc不报错）：
sudo apt-get install build-essential

查看显卡
···
ubuntu-drivers devices
···
自动安装驱动
···
ubuntu-drivers autoinstall
···
禁用开源驱动
···
sudo gedit /etc/modprobe.d/blacklist.conf
···
在文本最后添加
···
blacklist nouveau
options nouveau modeset=0
···
生成新的内核初始化文件
···
sudo update-initramfs -u
···
此处重启，输入
···
lsmod | grep nouveau
···
若没有输出，则成功禁用开源驱动。
输入
···
nvidia-smi
···
输出如下内容说明显卡安装成功

cuda10.2依赖gcc7,安装gcc7

sudo apt install gcc-7 g++-7

此处系统中存在两个gcc，分别是9和7，此时默认为9，这里需要将默认gcc改为gcc7。终端输入：

sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-9 90 --slave /usr/bin/g++ g++ /usr/bin/g++-9 --slave /usr/bin/gcov gcov /usr/bin/gcov-9
sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-7 70 --slave /usr/bin/g++ g++ /usr/bin/g++-7 --slave /usr/bin/gcov gcov /usr/bin/gcov-7
sudo update-alternatives --config gcc

选择1回车然后再次输入gcc –v查看版本为7.*

官方下载cuda10.2
cuda10.2官方地址
在驱动文件所在路径执行如下操作（此处需要根据自己的文件名称修改）
···
chmod a+x cuda_10.2.89_440.33.01_linux.run
sudo ./cuda_10.2.89_440.33.01_linux.run
···


如果安装完成未出现以下画面，说明安装成功，直接执行下面添加环境变量操作，否则执行cuda10.2安装异常操作

添加环境变量
安装完毕后，添加环境变量：
打开.bashrc文件：
···
gedit ~/.bashrc
···
在文末添加

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-10.2/lib64
export PATH=$PATH:/usr/local/cuda-10.2/bin
export CUDA_HOME=$CUDA_HOME:/usr/local/cuda-10.2

保存并关闭文件，更新环境：
source ~/.bashrc

三、cuda10.2安装异常操作
在.bashrc中添加环境变量
···
gedit ~/.bashrc
···
在文件的末尾加入下面两行语句：

export PATH=/usr/local/cuda-10.2/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda-10.2/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

设置环境变量和动态链接库
···
sudo gedit /etc/profile
···
在末尾加入下面语句：

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

保存退出
创建链接文件：
···
sudo gedit /etc/ld.so.conf.d/cuda.conf
···
加入下面语句：

/usr/local/cuda/lib64

然后执行：
sudo ldconfig // 使文件生效
测试CUDA的Sample

cd /usr/local/cuda-10.2/samples/1_Utilities/deviceQuery
make
sudo ./deviceQuery

如果出现与下图相似的图，说明CUDA10.2已经安装成功了

去官网下载cudnn，期间需要登录才可以下载。下载与cuda对应的cudnn版本即可，此处选择了7.6.5。
终端到下载路径解压缩
‘’’
tar -xzvf cudnn-10.2-linux-x64-v7.6.5.32.tgz
‘’’
将文件复制到cuda下：

cp cuda/include/cudnn.h /usr/local/cuda/include
cp cuda/lib64/libcudnn* /usr/local/cuda/lib64
chmod a+r /usr/local/cuda/include/cudnn.h /usr/local/cuda/lib64/libcudnn*

安装完成

默认已安装anaconda3
创建虚拟环境

conda create -n python38-tensorflow python=3.8

激活虚拟环境

conda activate python38-tensorflow

安装tensorflow-gpu

pip install -i https://pypi.tuna.tsinghua.edu.cn/simple tensorflow-gpu

测试tensorflow是否可以调用gpu

python
import tensorflow as tf
print(tf.test.is_gpu_available())

输出为True说明tensorflow-gpu可以调用gpu

以上就是今天要讲的内容，本文仅仅简单介绍了如何在ubuntu20.04上安装tensorflow2.0以上的版本