Accelerate FM training with Amazon SageMaker HyperPod training plans | AWS OnAir re:Invent 2024

Описание: Amazon SageMaker HyperPod offers a new way to train foundation models (FMs) within budget and timeline. In this episode, learn how SageMaker HyperPod finds the most cost-efficient training plan across multiple blocks of compute capacity, automatically provisions the infrastructure and runs the training jobs on compute resources without requiring any manual effort or intervention. When there is no training plan that meets the timeline, budget, or compute requirements, SageMaker HyperPod suggests alternate plans as the next best options. SageMaker HyperPod may suggests a training plan that includes executing training on multiple blocks of compute resources that may reside in different AWS Availability Zones. SageMaker HyperPod automatically takes care of moving the workload closer to the compute resources, while pausing and resuming training between gaps in compute availability, without requiring any intervention. This new capability can save weeks of effort for every model training, accelerating time to market for new FMs.

Get started with Amazon SageMaker Hyperpod: https://awsonair.net/sagemakerhyperpod

Follow AWS OnAir on
LinkedIn https://awsonair.net/LinkedInAWSOnAir
Twitch https://awsonair.net/subscribe

ABOUT AWS
Amazon Web Services (AWS) hosts events, both online and in-person, bringing the cloud computing community together to connect, collaborate, and learn from AWS experts. AWS is the world’s most comprehensive and broadly adopted cloud platform, offering over 200 fully featured services from data centers globally. Millions of customers—including the fastest-growing startups, largest enterprises, and leading government agencies—are using AWS to lower costs, become more agile, and innovate faster.

#AWS #AWSEvents #generativeai #foundationmodel #sagemaker #awsonair