• Post category:StudyBullet-3
  • Reading time:9 mins read


Learn Kubeflow by Example with Machine Learning – Deploy ML AI Pipelines on Google Cloud Platform – Kubernetes & AWS

What you will learn

How to build ml/ai pipelines with Kubeflow from scratch

Deploy Kubeflow on GCP and AWS with real-world examples, and best practices

Kubernetes & Kubeflow fundamentals

Run multiple ML pipelines with the Kubeflow UI

Description

In this course we will cover all the fundamentals first of Kubeflow with slides and presentations and then build and deploy ML/AI Pipelines with Kubeflow  together using the Google Cloud Platform (GCP) along with the GKE and active cloud shell. We will also learn the fundamentals of Kubernetes and Kubeflow along with GCP project management as we move forward together with the code lab.

Get hands on experience early with an exciting technology making ML deployments much easier thanks to the power of Kubeflow!

This is the course you’ve been looking for to get a clear and concise explanation of what is Kubeflow and the value it presents for creating efficiency with Machine Learning.

If you’d like to quickly and simply go through each step of code together and discuss the conventions and the commands for setting up cloud native and run multiple pipelines together – we’re even going to take a look at a recursive tutorial which runs  iterative prediction calculations with increasing margins of acceptable results, then this is perfect course is for you!

This course is modular and intended to be beginner friendly as well, so that if you are coming from a less technical or more business minded side or you are just keen on reviewing the fundamentals of kubernetes and, vms, containers and clusters and how they have significant value in relation to deploying and running machine learning pipelines then you will also find clear, simplified and contextualized examples as part of this course as well. Just remember, those sections are purely optional and if you already have fundamental knowledge please feel free to skip directly to the code lab and get started hands on with me.

What you will learn in this course:

  • Setting up the Google Cloud Platform development environment
  • Build and successfully deploy ML/AI Pipelines with Kubeflow
  • Learn the fundamentals of Kubernetes, GKE, Containers and Clusters in relation to Machine Learning
  • Work on a code lab with the GCP active cloud shell
  • Run ML Pipelines and examine events and logs – GPU, CPU and node management
  • Create buckets, OAuth, and credentials with Google Cloud Platform
  • Review the basics of Kubeflow for AWSEKS
  • Set up scheduling and billing on GCP for project administration and management
  • Check out deploying Jupiter notebook and for Kubeflow pipelines
  • And much more along the way!

Course Set up and Tools


Get Instant Notification of New Courses on our Telegram channel.


This course develops its Kuebflow project and source code with Active Cloud Shell on the Google Cloud Platform – it’s free to set up, but deploying and running the pipelines to completion yourself will require you to activate a billing account and it’s important that you monitor your costs in that case (this is optional and we explain the steps and procedure if you’re interested in spending a bit more to see kubeflow machine learning pipelines in action).

Is this the right course for you?

This course is straight to the point, time sensitive, and focuses on completing the project at hand (the reasons and explanations for the code and how it works) as the primary. Besides the initial sections which is meant for a 101 introduction into the basics of Kubeflow and Kubernetes for all levels, pretty much all of this course after that is just building out our Kubeflow Pipeline stopping to explain the techniques and dependancies connections along the way. If you are the type of person who gets the most out of learning ‘by doing’, then this course will be for you.

I’m looking forward to discovering the value and real ease of what it means to make our lives much more simple and efficient thanks to what kubeflow can offer!

And whenever you’re ready,  I’ll see you in the lessons!

C

English
language

Content

What Are Containers & Virtual Machines – Introduction
What Are Containers & Virtual Machines – 101 (Kubeflow)
How Do Containers Work
Isolation Differences Between Virtual Machines & Containers
Modular Adaptability & Customization Of Containers
Portability & Flexibility From VMs and Containers
What Is Kubernetes – Fundamentals
Quick Note – Kubernetes Section
Introduction To Kubernetes & Container Deployment
Tradition & Virtual Deployment Eras
The Container Deployment Era & Benefits
Kubernetes & Container Benefit Recap
Why Use Kubernetes
How is Kubernetes Useful
Kubernetes Review
Kubernetes & Clusters – Fundamentals
What is a Kubernetes Cluster – Containers & Hosts
Worker Nodes & The Master Node
Kubernetes Microservice Application Example Part I
Kubernetes Microservice Application Example Part II
What Is Kubeflow – Introduction
How Machine Learning Benefits From Kubernetes
Kubeflow Beginning with TFX
How Kubefow Makes It Easier For Developers
How Kubeflow Works – Basics
Kubelfow Project Set Up – Google Cloud Platform GCP
Important Note – Codelab & GCP Billing
Learn Kubeflow Lab Overview
Set Up A Google Cloud Platform Project for The Kubeflow Example
GCP GCloud Config Kubeflow Project Setup
Create A Bucket For Kubeflow Example Storage
Deploy A Kubeflow Pipeline – Kubernetes Engine Part I
Deploy A Kubeflow Pipeline – Kubernetes Engine Part II
Google Cloud Pipeline Billing And Budget Alerts
Set Up GKE Cluseter
Request GPU Quota Process
Request GPU Quota Process – Once Approved
Kubeflow Run A Pipeline From UI – Google Cloud Platform GCP
Kubeflow Run A Pipeline From UI – Google Cloud Platform GCP
Upload Yaml Pipeline Config Kubeflow File
Input Parameters For Kubeflow Pipeline Run
Kubeflow Pipeline Runs – Events & Logs
Pipeline Deployment Checkpoint – Deprecated Example
Iterative Recursive Example For Kubeflow – Pipeline Completion
Quick Look at Kubeflow Teardown Command
Optional – Deploying A Notebook WIth AI Platform GCP Kubeflow
Optional – Kubeflow on AWS