The home for great developer writing.

We surface the best developer writing from thousands of independent blogs, updated daily.

Introducing checkpointless and elastic training on Amazon SageMaker HyperPod

1 · Amazon Web Services · Dec. 3, 2025, 5:35 p.m.

Amazon SageMaker HyperPod artificial-intelligence AWS re:Invent Launch Amazon SageMaker AI Model Development Machine Learning Cloud Computing

Summary

Amazon SageMaker introduces new training features, including checkpointless training for instant failure recovery and elastic training for automatic scaling based on resource availability, aimed at enhancing AI model development efficiency.

Read full post on aws.amazon.com →

MORE POSTS LIKE THIS

New serverless customization in Amazon SageMaker AI accelerates model fine-tuning

Amazon Web Services · Dec 3, 2025

Amazon SageMaker AI artificial-intelligence

SQL with GenAI: Building an Apache Iceberg lakehouse on Red Hat OpenShift

Red Hat · Jun 22, 2026

SQL AI Functions

Google Colab CLI fuses local terminals with cloud GPUs

Developer Tech · Jun 8, 2026

AI Tools APIs

DigitalOcean Dedicated Inference: A Technical Deep Dive

DigitalOcean · Apr 25, 2026

engineering AI

Writing an LLM from scratch, part 32j -- Interventions: trying to train a better model in the cloud

gpjt · Apr 9, 2026

Machine Learning artificial-intelligence

Jupyter Notebooks on Heroku with Persistent Storage

Heroku · Apr 2, 2025

Jupyter Notebooks Heroku

Discover more posts →

AUTHOR