kubeflow/community-distribution

Kubeflow Community Distribution

What it solves

This project provides a standardized way to install and deploy the Kubeflow Platform across various popular Kubernetes clusters (such as Kind, Minikube, Rancher, EKS, AKS, and GKE). It simplifies the process of setting up a complete machine learning platform that includes essential components for training, serving, and orchestration, while supporting strict security and multi-tenancy requirements for both enterprises and academic users.

How it works

The repository contains a collection of Kubernetes manifests and Kustomize overlays. Users can deploy the entire platform or select individual components. The distribution is organized into three main categories:

  • Applications: Official Kubeflow components like Pipelines, KServe, Katib, and Notebooks.
  • Common: Shared services including Istio for traffic management, Dex for identity management, and Cert-manager for certificates.
  • Experimental: Third-party integrations and platform experiments (e.g., Ray).

Installation can be performed via a single-command bash one-liner or through a step-by-step manual process for better debugging and customization.

Who it’s for

  • ML Engineers and Data Scientists: Who need an end-to-end ML workflow environment.
  • Enterprise IT/Platform Teams: Who require a secure, multi-tenant ML platform on Kubernetes.
  • Academic Researchers: Who want to explore and contribute to the Kubeflow ecosystem.

Highlights

  • Broad Cluster Support: Compatible with Kind, Minikube, Rancher, EKS, AKS, and GKE.
  • Modular Installation: Ability to install the full platform or pick and choose specific components.
  • Integrated ML Tooling: Includes Kubeflow Pipelines, KServe for model serving, Katib for hyperparameter tuning, and Notebooks.
  • Enterprise-Ready: Built-in support for multi-tenancy, OIDC identity providers (via Dex), and network security (via Istio).

Related

  • Project
    kubeflow/pipelines

    A platform for building and deploying reusable, scalable end-to-end machine learning workflows on Kubernetes.

  • Project
    kubeflow/kubeflow

    Kubeflow is a Kubernetes-native foundation of tools for building AI platforms, providing a modular ecosystem to manage the entire AI lifecycle.

  • Project
    kubeflow/trainer

    A Kubernetes-native distributed AI platform for scalable LLM fine-tuning and training across various frameworks like PyTorch, JAX, and XGBoost.

  • Project
    Winson-030/dify-kubernetes

    A collection of Kubernetes deployment manifests that enable users to easily deploy and deploy Dify on a Kubernetes cluster.

  • Project
    kubesphere/kubesphere

    KubeSphere is a distributed operating system for cloud-native application management that simplifies Kubernetes management across multi-cloud, edge, and datacenter environments.