Bodywork is distributed as a Python package that exposes a Command Line Interface (CLI) for interacting with your Kubernetes (k8s) cluster. The Bodywork CLI configures your cluster to run pre-built Bodywork containers, that clone your project from its Git repository and then deploy it.
Bodywork can be downloaded and installed from PyPI with the following shell command,
$ pip install bodywork
Or directly from the master branch of the bodywork-core repository on GitHub using,
$ pip install git+https://github.com/bodywork-ml/bodywork-core.git
Check that the installation has worked by running,
Which should display the following,
Manage machine learning model deployments on k8s. --> see bodywork -h for help
Required Python Version¶
Bodywork has been built and tested using Python 3.7. We recommend that Bodywork-compatible ML projects should also be developed and tested using Python 3.7, but in-practice your code is likely to work with other versions.
Required Kubernetes Version¶
Bodywork relies on the official Kubernetes Python client, whose latest version (12.0.0) has full compatibility with Kubernetes 1.16. We recommend that you also use Kubernetes 1.16, but in-practice Bodywork will work with other versions - more information can be found here. Bodywork is tested against Kubernetes 1.16 running on Amazon Elastic Kubernetes Service.