快速搭建数据科学环境之mac docker kaggle image环境配置与使用 VSCode+Docker: 打造最舒适的深度学习环境

1 安装1.1 安装dockerbrew cask install docker1.2 拉取kaggle镜像目前只有cpu版本的docker run--rm-it kaggle/python2 运行2.1 运行jupyter-notebookdocker run-v $PWD:/tmp/working-w/tmp/working-p8888:8888--rm-it kaggle/python jupyter notebook--no-browser--ip0.0.0.0--notebook-dir/tmp/working--allow-root2.2 访问jupyter-notebookhttp://0.0.0.0:8888/?token6ffe65f5860a5736f6b190917673b532e772b7caca5f3b8e2.3 直接配置到环境变量中这样直接可以用本地命令来启用了添加到~/.bash_profilekaggle_python(){docker run-v $PWD:/tmp/working-w/tmp/working-p8081:8081--rm-it kaggle/python python$}kaggle_jupyter(){docker run-v $PWD:/tmp/working-w/tmp/working-p8888:8888--rm-it kaggle/python jupyter notebook--no-browser--ip0.0.0.0--notebook-dir/tmp/working--allow-root}启用环境变量source~/.bash_profile启用docker_python或者docker_jupyter-notebookkaggle_python kaggle_jupyter3 远程环境VSCodeDocker: 打造最舒适的深度学习环境3.1 安装ssh$ apt update $ apt install-y openssh-server3.2 配置$ mkdir/var/run/sshd $ echoroot:passwd|chpasswd# 这里使用你自己想设置的用户名和密码但是一定要记住$ sed-is/PermitRootLogin prohibit-password/PermitRootLogin yes//etc/ssh/sshd_config $ sedssession\s*required\s*pam_loginuid.sosession optional pam_loginuid.sog-i/etc/pam.d/sshd $ echoexport VISIBLEnow/etc/profile3.3 运行dockerdocker run-v $PWD:/tmp/working-w/tmp/working-p8081:8081--rm-it kaggle/python3.4 激活ssh$ service ssh restart3.5 查看端口docker ps#查看容器docker port[your_container_name]8081#0.0.0.0:80223.6 登入ssh root[your_host_ip]-p80224 参考链接[1]:Mac上安装Kaggle Docker 镜像的方法http://frankchen.xyz/2017/02/13/Install-Kaggle-Docker-on-Mac/[2]:爆速でKaggle環境を構築するhttps://qiita.com/bam6o0/items/354faa9394755a984661