Ubuntu18.04深度学习环境搭建

        安装完系统后,直接ALT+CTRL+F1,进入超级终端模式,之后先安装显卡的驱动,之后再安装cuda,然后所有的应用都可以继续安装了。

一、更换下载源:

1、sudo cp /etc/apt/sources.list /etc/apt/sources.list.copy

2、添加清华源路径:

sudo gedit /etc/apt/sources.list
删除之前内容,添加下面内容:

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

3、sudo apt-get update 

     sudo apt-get upgrade 

二、显卡安装:

sudo add-apt-repository ppa:graphics-drivers/ppa
sudo apt update
sudo apt upgrade

使用如下命令查看什么驱动会被安装
ubuntu-drivers devices

输入如下命令可以自动安装最新的驱动
sudo ubuntu-drivers autoinstall

重启后输入如下命令即可查看显卡的状态
nvidia-smi


三、CUDA安装

sudo cuda_11.2.243_418.87.00_linux.run

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

之后进行环境配置,把CUDA相关命令和库文件添加到系统目录中。不然有些情况会出现找不到命令(如nvcc)或者库的问题。
sudo gedit /etc/profile

修改profle文件,这个是对所有用户有效,如果仅对当前用户,则修改~/.bashrc文件。在最后部分加上

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

然后运行

source /etc/profile

安装CUDNN
sudo dpkg -i libcudnn8_8.1.1.33-1+cuda11.2_amd64.deb
sudo dpkg -i libcudnn8-dev_8.1.1.33-1+cuda11.2_amd64.deb
sudo dpkg -i libcudnn8-samples_8.1.1.33-1+cuda11.2_amd64.deb

四、opencv安装

经过下面的操作,opencv就可以开开心心使用了,opencv4.5.1地址:

链接: https://pan.baidu.com/s/1lKshV-FTnPbTkYYztNvUUg 提取码: n4h5

1、编译源码
cd opencv-4.5.1
mkdir build
cd build

cmake -DCMAKE_BUILD_TYPE=Release -DOPENCV_GENERATE_PKGCONFIG=ON -DWITH_FFMPEG=OFF -DINSTALL_C_EXAMPLES=OFF -DINSTALL_PYTHON_EXAMPLES=OFF -DBUILD_PERF_TESTS=OFF -DCMAKE_INSTALL_PREFIX=/soft/opencv451 -DBUILD_opencv_world=ON -DBUILD_DOCS=OFF -DBUILD_PERF_TESTS=OFF -DBUILD_TESTS=OFF -DWITH_GSTREAMER=ON -DBUILD_TIFF=ON -DOPENCV_EXTRA_MODULES_PATH=../opencv_contrib4.5.1/modules ..

make -j4
make install

2、生效步骤1
vim /etc/profile.d/pkgconfig.sh
在pkgconfig.sh文件中写入(可能是空文件):
export PKG_CONFIG_PATH=/soft/opencv451/lib/pkgconfig:$PKG_CONFIG_PATH
最后:
source /etc/profile
source /etc/profile.d/pkgconfig.sh
 
3、生效步骤2
vim /etc/ld.so.conf.d/opencv4.conf
在opencv4.conf文件中写入(可能是空文件):
使其生效
sudo ldconfig

4、生效步骤3
sudo gedit /etc/bash.bashrc
PKG_CONFIG_PATH=$PKG_CONFIG_PATH:/soft/opencv451/lib/pkgconfig
export PKG_CONFIG_PATH
source /etc/bash.bashrc

查看opencv是否成功
pkg-config --modversion opencv4

五、Anacoda

下载地址:https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/

选择:Anaconda3-2021.11-Linux-x86_64.sh

默认安装即可,完成后配置路径:

gedit .bashrc

export PATH="/home/cas/anaconda3/bin:$PATH"

source .bashrc

测试一下安装成功了没有
conda info

猜你喜欢

转载自blog.csdn.net/zhulong1984/article/details/127421564