Installing Kubeflow
This guide describes how to install Kubeflow projects, or Kubeflow AI reference platform using package distributions or Kubeflow manifests.
Read the introduction guide to learn more about Kubeflow, Kubeflow projects, and Kubeflow AI reference platform.
Installation Methods
You can install Kubeflow using one of these methods:
Kubeflow Projects
Kubeflow projects in the Kubeflow ecosystem can be deployed as a standalone services, without the need to install the entire Kubeflow AI reference platform.You might integrate these services as part of your existing AI platform or use them independently.
These projects are a quick and easy method to get started with the Kubeflow. They provide flexibility to users who may not require the capabilities of a full Kubeflow AI reference platform.
The following table lists Kubeflow projects that may be deployed in a standalone mode. It also lists their associated GitHub repository and corresponding AI lifecycle stage.
| Kubeflow Project | AI Lifecycle Stage | Source Code |
|---|---|---|
| KServe | Model Serving | kserve/kserve |
| Kubeflow Katib | Model Optimization and AutoML | kubeflow/katib |
| Kubeflow Model Registry | Model Registry | kubeflow/model-registry |
| Kubeflow Pipelines | ML Workflows and Schedules | kubeflow/pipelines |
| Kubeflow Spark Operator | Data Preparation | kubeflow/spark-operator |
| Kubeflow Trainer | Model Training and LLMs Fine-Tuning | kubeflow/trainer |
Kubeflow AI Reference Platform
You can use one of the following methods to install the Kubeflow AI reference platform and get the full suite of Kubeflow projects bundled together with additional tools.
Packaged Distributions
Packaged distributions are maintained by various organizations and typically aim to provide a simplified installation and management experience for your Kubeflow Platform. Some can be deployed on multiple Kubernetes distributions, while others target a specific platform (e.g. EKS or GKE).
The following table lists distributions which are maintained by their respective maintainers:
| Maintainer Distribution Name | Kubeflow Version | Target Platform | Link |
|---|---|---|---|
| Aranui Solutions deployKF | [version matrix] | Multiple [list] | Website |
| Canonical Charmed Kubeflow | [release notes] | Multiple | Website |
| Google Cloud | [release notes] | Google Kubernetes Engine (GKE) | Website |
| IBM Cloud | [release notes] | IBM Cloud Kubernetes Service (IKS) | Website |
| Microsoft Azure | [release notes] | Azure Kubernetes Service (AKS) | Website |
| Nutanix | Nutanix Kubernetes Platform | Website | |
| QBO GPU Cloud | [release notes] | QBO Kubernetes Engine (QKE) | Website |
| Red Hat Open Data Hub | OpenShift | Website |
Kubeflow Manifests
The Kubeflow manifests are a collection of community maintained manifests to install Kubeflow AI reference platform in popular Kubernetes clusters such as Kind (locally), Minikube (locally), Rancher, EKS, AKS, GKE. They are aggregated by the Manifests Working Group and are intended to be used by users with Kubernetes knowledge and as the base of packaged distributions.
Kubeflow Manifests contain all Kubeflow projects, Kubeflow Central Dashboard, and other Kubeflow applications that comprise the Kubeflow AI reference platform. This installation is helpful when you want to try out the end-to-end Kubeflow AI reference platform capabilities.
If you want a stable / conservative experience we recommend to use the latest stable release:
You can also install the master branch of kubeflow/manifests by following the instructions here and provide us feedback.
Next steps
- Review our introduction to Kubeflow.
- Explore the architecture of Kubeflow.
- Learn more about the Kubeflow projects.
Feedback
Was this page helpful?
Thank you for your feedback!
We're sorry this page wasn't helpful. If you have a moment, please share your feedback so we can improve.