/mlflow-tracking-server

This repository hosts the code to make it easier to deploy a customizable and flexible MLflow tracking server solution to your Kubernetes cluster.

Primary LanguageShellApache License 2.0Apache-2.0

MLflow Tracking Server on Kubernetes 🚀: A customizable and flexible solution ✨

This repository hosts the code to make it easier to deploy an MLflow tracking server to your Kubernetes cluster.

Project Structure 🗺️

  • scripts directory has the entrypoint that is executed when the image defined by the Dockerfile is run.
  • The Dockerfile specifies the recipe for building the image.

Contributions Welcome! 👋🥰

The script inside the scripts folder currently has options for adding the metadata store and the artifact store. It can be expanded to include the flags

Blog ✍️

A detailed idea behind this repository and the steps to execute it in a cloud environment can be obtained from this blog post that I wrote on my experience building a solution like this.

Read it here 🧑‍💻: Not just another MLflow on Kubernetes article