Skip to content

FootprintAI/kubeflow-workshop

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

59 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

kubeflow workshop

This workshop is for educational purpose, please contact [email protected] if you want to introduce kubeflow to your orgainzation.

Pipelines

we have listed few good-for-beginner pipelines to get started, please check Pipeline folder

Tutorials

We have listed some tutorials on how advanced use kubeflow and it's components, please check Tutorials folder

Talks & Slides

We put our slides from invited talks and workshops under here

Prerequsite

CPU: 4 Core+
Memory: 16G+
Gpu: Optional

CLI Installation

  • multikf: a simplified cli tool to install one or many kubeflow clusters in the same host.

Script Installation (w/o GPU)

To simplify the installation process, we use root to install everthing (although it is NOT recommended using in production environment, use it on your own risk)

To install kubeflow, you need to have a running kubernetes cluster, either an existing running one or you can just create one. In this example, we provides both minikube installation and kind installation for creating a cluster, you can just pick either one.

Since KinD use containerd not dockerd as its runtime, so we have to use emissary as Kubeflow/Pipeline's default executor.) See script below to replace its default runtime executor (tested in v1.4)

We then install kubeflow on the cluster we just created.

For Windows user

We have scrips to Windows user to install kubeflow, please noted that you have to use Windows 10 Pro or later version which supports Hyper-V.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages