The platform for Kubermatic Kubernetes is an open-source, production-grade, cluster-management tool. It allows for automation and flexibility in the integration of the lifecycle management of a cluster as well as the ML/DL workflow.

Get Started

Before you deploy a model to your Kubernetes program, you need an installation of Docker and the creation of a container image of your model.

Containerization of the model includes the following steps:

Download and start Docker

Make a directory

Organize the dependencies and your code

Make a file for requirements. The .txt file should contain code to be run.

Make a Dockerfile for Docker to read, build and finally use to run the model

Build a Docker container and ensure that it is working

In Docker Hub, the upload model

Deploy the model to your Kubernetes cluster