Docker部署深度学习

1.下载Ubuntu

docker pull docker.io/library/ubuntu:18.04

2.通过镜像运行一个容器

[root@localhost ~]# docker run -dit -v /root/lgf:/root/lgf -p 5001:5000 --name=tests docker.io/ubuntu:18.04

-v挂载  -p暴露端口 主机地址:容器地址  --name取名字

-d后台启动  -it以交互方式启动容器

-dit  当ENTRYPOINTis bash或sh  docker run -d ubuntu将立即停止，原因bash是找不到要分配的任何伪终端。必须指定-it以便bash或sh可以分配给伪终端。

3.进入容器

docker exec -it tests /bin/bash

安装深度学习环境

1.下载vim（编辑配置文件）

root@a5d1c5b81c15:/# apt-get update
root@a5d1c5b81c15:/# apt-get install vim

找到配置文件

vim /etc/apt/sources.list

将/etc/apt/sources.list文件的内容清除，然后加入目的软件源的配置。比如拷贝如下内容，便是使用阿里的Ubuntu软件仓库镜像站点，这儿是支持Ubuntu 18的

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

2.安装sudo

root@a5d1c5b81c15:/# apt-get update
root@a5d1c5b81c15:/# apt-get install sudo

sudo命令功能：用于普通用户使用管理员权限执行某些操作

3.安装wget 

root@a5d1c5b81c15:/# sudo apt-get install wget

4.安装zlib

root@a5d1c5b81c15:/# sudo apt-get install zlib*

5.安装python

root@a5d1c5b81c15:/# sudo wget https://www.python.org/ftp/python/3.7.0/Python-3.7.0rc1.tgz

解压

root@a5d1c5b81c15:/# sudo tar -zxvf Python-3.7.0rc1.tgz

安装gcc和make

sudo apt-get install gcc 
sudo apt-get install make

进行配置

root@a5d1c5b81c15:/# cd Python-3.7.0rc1 
root@a5d1c5b81c15:/Python-3.7.0rc1# ./configure --prefix=/opt/ptyhon3.7

编译及安装

sudo make
sudo make install  

安装python3.7出现ModuleNotFoundError: No module named ‘_ctypes’解决办法

解决办法：

sudo apt-getupdate
sudo apt-getupgrade
sudo apt-getdist-upgrade
sudo apt-getinstall build-essential python-dev python-setuptools python-pip python-smbus
sudo apt-getinstall build-essential libncursesw5-dev libgdbm-dev libc6-dev
sudo apt-getinstall zlib1g-dev libsqlite3-dev tk-dev
sudo apt-getinstall libssl-dev openssl
sudo apt-getinstall libffi-dev

测试python是否安装成功

python3 -V

安装setuptools

​sudo wget --no-check-certificate https://pypi.python.org/packages/source/s/setuptools/setuptools-19.6.tar.gz
root@a5d1c5b81c15:/Python-3.7.0rc1# sudo tar -zxvf setuptools-19.6.tar.gz
tar -zxvf setuptools-19.6.tar.gz
cd setuptools-19.6
python3 setup.py build
python3 setup.py install

​

安装pip

Ubuntu配置pip国内镜像加速安装

root@a5d1c5b81c15:/Python-3.7.0rc1# cd ~/.pip
bash: cd: /root/.pip: No such file or directory
root@a5d1c5b81c15:/Python-3.7.0rc1# mkdir ~/.pip
root@a5d1c5b81c15:/Python-3.7.0rc1# cd ~/.pip
root@a5d1c5b81c15:~/.pip# touch pip.conf
root@a5d1c5b81c15:~/.pip# sudo gedit ~/.pip/pip.conf
sudo: gedit: command not found
root@a5d1c5b81c15:~/.pip# sudo apt-get install gedit

下载pip 

sudo apt-get install python3-pip

下载深度学习需要的python库

sudo pip install tensorflow==1.9.0
sudo pip install keras==2.3.1   

镜像保存

目前已经得到了一个可以快速部署模型的容器，接下来将其做成镜像并保存
在退出容器后将容器保存为镜像

1. # 命令和git 原理类似

2. docker commit -m="提交的描述信息" -a="作者" 容器id 目标镜像名：[TAG]

docker commit -a="lgf" -m="deep learning" d798a5946c1f tests:1.0

将该镜像保存为tar文件

docker save tests:1.0 > tests.tar