intro

github actions provide an ci/cd platform to automate builds, tests and deployments. this post will include the core concepts of github actions and examples how to use it for machine learning or software projects.

terminology

we will give a quick introduction to the main terminology of github actions. for detailed and extensive explanations, we refer to github’s documentation.

in github actions, one can configure automated processes to build, test and deploy code. these processes are called workflows which contain one or more jobs. A single job is a set of actions where a single action is defined to perform complex but frequently repeated tasks. events are activities such as pull requests that trigger workflows . finally, runners are the servers that run your workflows .

there are many options to customize your workflows such as actions developed by the community, using variables, execute custom scripts and process using various expressions and literals.