Amazon SageMaker Lifecycle Scripts and Custom Images

Name: Amazon SageMaker Lifecycle Scripts and Custom Images
Rating: 5.0 (13 reviews)

Customize the SageMaker Environment for Machine Learning and AI Projects on AWS with Lifecycle Scripts and Custom Images

Created byJoe Cline

Last updated 10/2025

English

What you'll learn

What Amazon SageMaker is
The Amazon SageMaker and SageMaker Studio interface
The Amazon SageMaker Studio JupyterLab environment
The training, inference, and deployment features in SageMaker
What the BASH shell is
Understand AWS IAM permissions and policies for SageMaker
What a SageMaker Lifecycle Script is and how it’s structured
Create and configure SageMaker Lifecycle configuration scripts
What SageMaker images are
What Docker is
The difference between containerization and virtual machines
What container repositories are
When to use a lifecycle script or a custom image
Best practices for creating and deploying lifecycle scripts and custom images
How to create a Docker container build environment with EC2 and ECR
Create and configure a Amazon SageMaker Studio Custom Image

Course content

5 sections • 26 lectures • 1h 33m total length

Introduction to Amazon SageMaker Customization1:24
Welcome to this mini-course on Amazon SageMaker.

Our mini-courses are designed to be under two hours so you can get to the information you need quickly.
This course is for those who need to customize a SageMaker environment and don't have 40, 20 or even 10 hours to spend on training.

There are some explainer videos, but only enough to understand what we are doing in the demonstrations and why.
I see plenty of other courses that teach how to create ML models in SageMaker, but not how to set it up and customize it.
I hope you find great value in this course. Please shoot me an email if you have suggestions for this course or any mini-course you would like to see us make.

An Overview of SageMaker and the SageMaker Studio Space4:15
In this video, we first learn what SageMaker is and then give a tour of the SageMaker dashboard and SageMaker Studio.
An Overview of JupyterLab in SageMaker Studio Part 13:34
In this video we first launch a SageMaker Studio Space and tour the JupyterLab environment
An Overview of JupyterLab in SageMaker Studio Part 22:26
This video continues our tour of JupyterLab.
An Overview of JupyterLab in SageMaker Studio Part 35:06
This video continues our tour of JupyterLab
An Overview of JupyterLab in SageMaker Studio Part 42:57
This video completes our tour of JupyterLab.
Module 1: Key Components of SageMaker Quiz, Lectures 1-6
SageMaker Key Componants: Training, Inference and Deployment Features4:24
In this video, we learn about the training, inference and deployment features in SageMaker.
Demo: Setting up a SageMaker Studio Domain and User Profile6:09
In this video, I demonstrate setting up a SageMaker Studio domain and user profile.
Demo: Setting Up a Legacy SageMaker Jupyter Notebook Instance4:03
In this video, I demonstrate setting up a legacy SageMaker notebook instance.
Understanding IAM Permissions and Policies for Amazon SageMaker5:42
In this video, we revisit IAM permissions and policies for SageMaker in more detail.
Module Summary0:39
In this video, we summarize what we learned in this module.
Module 1: Key Components of SageMaker Quiz, Lectures 1-11

An Overview of the Amazon SageMaker Lifecycle Script1:29
In this video, we discuss what a SageMaker lifecycle script is and why we might want to create one.
The Anatomy of the SageMaker Lifecycle Script4:08
In this video, I break down the typical structure of a lifecycle script.
Demo: How to Create and Deploy an SageMaker Lifecycle Script4:11
In this video, I give a demonstration on how to create and deploy a SageMaker Lifecycle script.
(See resources to download the lifecycle script used in this demo.)
Amazon SageMaker Lifecycle Script Best Practices1:14
In this video, we discuss best practices and sum up what we learned in this module.
Module 2: Amazon SageMaker Lifecycle Scripts Quiz, Lectures 12-15
Stretch Assignment #1: Create and configure an Amazon SageMaker Lifecycle Script

An Overview of Amazon SageMaker Images3:46
In this video, we learn what an image is in SageMaker.
An Introduction to Containerization With Docker6:10
In this video, we get an introduction to Docker and containerization in general, as well as, leaning the difference between an AWS EC2 instance (a virtual machine) and a container.
An Introduction to Container Repositories with DockerHub2:05
In this video, I introduce you to container repositories starting with DockerHub.
An Introduction to Container Repositories with Amazon Elastic Container Registry2:12
In this video, we pick from where the last video left off with an overview of Amazon Elastic Container Registry, or ECR. After that, I compare DockerHub to ECR.
Module 3: Custom SageMaker Images Quiz, Lectures 16-19
Demo: Create and Deploy a Custom SageMaker Image - Part 16:18
Part 1 of a demonstration to create and deploy a custom image. In this video, we start with launching an EC2 instance to build our Docker container.
Demo: Create and Deploy a Custom SageMaker Image - Part 23:58
Part 2: In this video, we learn how to connect via SSH to the EC2 instance we launched in the last video.
Demo: Create and Deploy a Custom SageMaker Image - Part 33:40
Demo Part 3: In this video, we learn how to create our Docker build environment on EC2 by installing docker, uploading a Dockerfile and creating our build directory.
(See resources to download the Dockerfile used in the demo.)
Demo: Create and Deploy a Custom SageMaker Image - Part 45:30
Demo Part 4: In this video, we learn how to create an IAM user and a user's access key. Then, we build the image and push it to ECR. (See resources for the JSON policy for your ECR repo permissions)
Demo: Configuring SageMaker to Use A Custom Image4:33
In this demonstration, I show you how to configuring SageMaker to use the new, custom image.
Creating Amazon SageMaker Custom Images Best Practices and Module Summary1:30
In this video, we examine some best practices for creating and deploying custom images.
Module 3: Custom Amazon SageMaker Images Ouiz for Lectures 16-25
Stretch Assignment #2: Build an Amazon SageMaker Custom Image with a Dockerfile

Requirements

Basic AWS familiarity is preferred but not required
Knowlege of BASH and Linux helpful but not required
Basic understanding of Docker helpful but not required
Knowledge of basic Python helpful but not required

Description

Disclaimer: This course contains the use of artificial intelligence.

Due to a disability, I have used AI to assist with the production and editing of the audio in this course. The preview/promo video uses an AI generator of my image – but it is my image. I also use a "text-to-speech" AI tool to create AI audio of my voice recorded from an earlier date.

Oh, and I used it to create the image used in Udemy’s search results and in a couple of slides.

The ideas, most of the slides, script, demos, and the downloadable materials are of my own production. I hope this does not dissuade you from this course.

Now, back to the course description.

~ Joe Cline

Do you need to customize an Amazon SageMaker environment but don't have 40, 20, or even 10 hours to spend on training? I see plenty of other courses that teach how to create machine language (ML) models and AI pipelines in AWS SageMaker, but NOT how to set it up and customize it.

Introducing, d8aland's new mini course on how to customize SageMaker

Our mini courses are designed to be under two hours so you can get to the information you need quickly. They are created for specific use-cases instead of just a lengthy, high level tech overview. In other words, "how do I. . .?"

There are some explainer videos, but only enough to understand what we are doing in the demonstrations and why. In this case, how to create an SageMaker Studio domain, launch JupyterLab and, configure lifecycle scripts and custom images.

Who Should Enroll?
Data scientists, ML engineers, and AWS professionals who want to optimize SageMaker ML for advanced AI projects or prepare for the AWS Machine Learning Specialty exam.

I hope you find great value in this mini-course from D8aland. Please shoot me a message if you have suggestions for this course or any other mini course you would like to see us make.

Now, let's go build something cool.

Joe Cline

A senior level data engineer with over 25 years experience in enterprise data management.

Who this course is for:

Data scientist and analyst who build models in SageMaker but want to learn SageMaker Studio customization
AWS Cloud Platform Enginners who wants to set up a custom SageMaker environment
Engineers who work on a cloud platform other than AWS and need to learn about the SageMaker service

Amazon SageMaker Lifecycle Scripts and Custom Images

What you'll learn

Explore related topics

Course content

Introduction1 lecture • 1min

Module 1: Key components in AWS SageMaker10 lectures • 39min

Module 2: Amazon SageMaker Lifecycle Scripts4 lectures • 11min

Module 3: Custom SageMaker Images10 lectures • 40min

Course Summary1 lecture • 2min

Requirements

Description

Who this course is for: