This is a set of dotfiles based loosely on the set of fastai dotfiles. I have chosen to restructure them to be more zsh friendly because of Apples migration from bash to zsh and, because of licensing, including a very old version of bash that does not support all of the settings in the fastai bashrc. It is designed to support Mac and Linux OS environments, and to be fairly light weight. It is recommended that Windows users use Windows Subsystem for Linux (WSL) that will give you a linux environment where you can then use this set of dotfiles. In this setup, the bash shell environment is mainly for linux where bash tends to be the default shell. The zsh shell does try to install some plugins, but is still fairly light weight.
Like the fastai version, this set of dotfiles is based on using git to manage dotfiles using a Git bare repository. It includes simple configurations for vim, zsh, git, ctags.
Jeremy Howard as a set of youtube live coding videos where he shows how to setup a machine learning environment in different contexts. There is forum discussion on each of the videos that gives more detail on the setup. If you are setting up a Linux machine or VM, fastai fastsetup has files they use to quickly configure a machine or WSL in windows.
- Move vimrc from using Vundle as a plugin manager to minpac
- Creation of a seperate aliases file that can be shared between bash and zshrc
- Creation of a zshrc
- Homebrew for MacOS packages
- Homebrew for MacOS casks
- Some subdirectory refactoring using .local/bin
- Including a version of the fastai dotfiles
I encourage you to have a look through the scripts in this repository before you run them. It is good practice not to run any scripts unless you have an understanding of what they are doing.
A convenient method for accessing git is through ssh instead of using the HTTPS interface. This involves generating a ssh key pair and adding that key pair to your github account.
Generate ssh key pair for this computer replacing your email in the following command.
ssh-keygen -t ed25519 -C "[email protected]"Copy the newly generated public key and add to GitHub account.
pbcopy < ~/.ssh/id_ed25519.pubWhen cloning from a github repository, select "ssh clone* vs "HTTPS clone" from the dropdown. You can create a key pair for the cluster machines you will be doing development on. Jeremy Howard discusses setting up ssh keys and there is a forum topic.
It is highly recommended that you create your own copy of these dotfiles by forking the repository. This will allow you to create your own aliases and customize the dotfiles for different environments. After you have a fork of the repository, you can
If you are developing AI apps under MacOS, homebrew is a convenient way to provide command line applications that are either older versions, or not provided in MacOS. To build software, homebrew requires that xcode command line tools be installed. The free code camp has some extended instructions on how to do this. You can do this from the mac terminal
xcode-select --installand then install homebrew using the homepage instructions:
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"The "Live coding 1" session with Jeremy Howard does a walk through of installing using the mambaforge installer. Since this video, mamba has been integrated into the main Miniforge installer. This is the fastest way to get a functional python environment with some sensible defaults. The install instructions on the miniforge webpage are pretty good to get a working install. It is highly recommended to have a look at Beginner's Guide to Mambaforge Installation