Official repository of Jupyter notebooks for building, training, evaluating, deploying, and operationalizing ML models.
End-to-end code guides for model training, fine-tuning, evaluation, inference, and MLOps on AWS infrastructure.
A curated collection of runnable Jupyter notebooks maintained directly by the AWS SageMaker team. It provides reference code for machine learning practitioners building, fine-tuning, and operationalizing models using the SageMaker Python SDK and SageMaker-Core.
The repository organizes self-contained notebooks across the entire ML lifecycle, including distributed training, LLM fine-tuning (SFT, DPO, RLVR), model evaluation, and deployment options like real-time or serverless endpoints. It also demonstrates MLOps capabilities including SageMaker Pipelines, Feature Store, and MLflow tracking.
Suited for data scientists, ML engineers, and AWS architects who build production ML pipelines on AWS. It is not suitable for practitioners looking for cloud-agnostic tools or non-AWS ML infrastructure.
Demonstrates script and framework training, distributed compute, managed spot training, and local code submission via the remote decorator.
Includes recipes for supervised fine-tuning, direct preference optimization, reinforcement learning, and continued pre-training.
Shows how to configure real-time, serverless, and asynchronous endpoints alongside A/B testing and latency optimizations.
Features examples using an object-oriented Python interface with resource chaining, type hints, and automatic state management.
Covers automated pipelines, feature stores, experiment tracking, MLflow integration, and Clarify bias analysis.
The code in this repository is completely free under the Apache 2.0 license. However, running the code requires an AWS account, and you will be billed for all compute instances, S3 storage, training jobs, and hosted endpoints created by the notebooks.
880 open remote AI-training roles, updated November 16, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
These reviewed AI tools do the same job instantly — from our scored directory.
Yes, provided you install the required Python libraries, configure AWS credentials, and set an active IAM execution role.
This repository is maintained by the core SageMaker team to showcase official features, while the community repository hosts external solutions and broader reference architectures.
The repository code is free under the Apache-2.0 license, but executing the notebooks provisions paid AWS cloud services such as EC2 instances and S3 storage.
See Zekai first in Google