aws/amazon-sagemaker-examples

★ 10,922Jupyter NotebookAudience · dataComplexity · 3/5LicenseSetup · moderate

Mindmap

mindmap
  root((SageMaker Examples))
    What it covers
      Model training
      Model deployment
      Data labeling
      Geospatial ML
    How to use
      Run in SageMaker
      Jupyter notebooks
      Adapt to own data
    Prerequisites
      AWS account
      IAM permissions
      S3 bucket
    Repo structure
      Official examples
      Community companion repo
      New PRs for gaps only
    Audience
      Data scientists
      ML engineers
      AWS practitioners

mindmap root((SageMaker Examples)) What it covers Model training Model deployment Data labeling Geospatial ML How to use Run in SageMaker Jupyter notebooks Adapt to own data Prerequisites AWS account IAM permissions S3 bucket Repo structure Official examples Community companion repo New PRs for gaps only Audience Data scientists ML engineers AWS practitioners

Click or tap to explore — scroll the page freely

Things people build with this

USE CASE 1

Learn how to train a machine learning model on SageMaker by following a working, runnable notebook.

USE CASE 2

Deploy a trained model as an API endpoint on AWS using SageMaker Hosting services.

USE CASE 3

Set up automated data labeling for a computer vision or NLP dataset using SageMaker Ground Truth.

USE CASE 4

Run geospatial machine learning analysis using SageMaker's built-in geospatial processing capabilities.

Tech stack

PythonJupyter NotebookAWS SageMakerS3IAM

Getting it running

Difficulty · moderate Time to first run · 30min

Requires an AWS account with SageMaker and S3 permissions, running notebooks incurs AWS compute and storage charges.

Apache 2.0, use freely for any purpose including commercial, keep the license notice.

In plain English

This is the official AWS repository of example notebooks for Amazon SageMaker, Amazon's cloud-based machine learning platform. The notebooks use the Jupyter format, which presents code alongside explanations and output in a single interactive document, making them a practical way to follow along with specific tasks rather than just reading documentation. SageMaker handles the heavy infrastructure side of machine learning: spinning up computing resources for training models, managing where data is stored, and hosting trained models so they can respond to requests. These examples walk through how to use those capabilities for a wide range of tasks, from basic model training to more specialized workflows like geospatial analysis, automated data labeling, and deploying models at scale. The README notes the collection is split into two repositories. This one is the official set maintained directly by the SageMaker team and focuses on breadth across SageMaker features. A companion community repository exists for additional examples and reference solutions contributed by AWS engineers and architects outside the core team. New pull requests to this official repository are only accepted for features not yet covered anywhere in the existing notebooks. Getting started requires an AWS account, appropriate permissions set up through AWS Identity and Access Management, an S3 storage bucket for data, and a SageMaker Notebook Instance. Once set up, these notebooks are available directly within the SageMaker interface and can also be run with minimal changes outside of SageMaker by updating the permissions configuration and installing the required Python libraries. The full README is longer than what was shown.

Copy-paste prompts

Prompt 1

Which SageMaker example notebook should I start with to fine-tune a text classification model on my own dataset?

Prompt 2

Take the SageMaker XGBoost training example and adapt it to use my own CSV dataset stored in an S3 bucket.

Prompt 3

How do I set up the IAM permissions and S3 bucket needed to run the SageMaker example notebooks without errors?

Prompt 4

Find the SageMaker notebook that shows how to deploy a PyTorch model as a real-time endpoint and explain each step.

Open on GitHub → Explain another repo

← aws on gitmyhub — every repo by this author, as a profile.

Verify against the repo before relying on details.