ubuntu18配置深度学习环境

Posted 2021-05-07 dangxusheng

1 安装nvidia驱动

1.1 设置root

sudo passwd 123

1.2 检测nvidia显卡

ubuntu-drivers devices

(base) dxs@dxs-ubuntu:~$ ubuntu-drivers devices
== /sys/devices/pci0000:00/0000:00:01.0/0000:01:00.0 ==
modalias : pci:v000010DEd00002184sv00001458sd00003FC7bc03sc00i00
vendor   : NVIDIA Corporation
driver   : nvidia-driver-430 - distro non-free recommended
driver   : xserver-xorg-video-nouveau - distro free builtin

(base) dxs@dxs-ubuntu:~$ 

1.3 安装nvidia驱动

sudo apt install nvidia-driver-430

1.4 安装完成后 reboot

--------------------------------------------------------

2 安装anaconda3

2.1 下载Anaconda3-5.2.0-Linux-x86_64.sh

2.2 安装：sudo sh Anaconda3-5.2.0-Linux-x86_64.sh

2.3 设置conda镜像

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

-----------------------------------------------------------------------------

3 安装cuda cudnn

conda install cudatoolkit=10.0
conda install cudnn=7.6.0

------------------------------------------------------------------------------

4 安装tensorflow-gpu

conda install tensorflow-gpu=1.13.1

5 安装pytorch

在线安装： conda install pytorch torchvision cudatoolkit=10.0 -c pytorch
离线安装： 先去 https://download.pytorch.org/whl/cu100/torch_stable.html 下载 torch-1.2.0-cp36-cp36m-manylinux1_x86_64.whl 和 torchvision-0.4.0-cp36-cp36m-manylinux1_x86_64.whl ， 然后安装

6 安装caffe

conda install -c defaults caffe-gpu