Home > Categories > Machine learning and neural networks > Model Deployment > Amazon SageMaker Hosting

Amazon SageMaker Hosting

Rating:

4.7 / 5.0

Tags

Machine Learning, Model Deployment, Cloud Service, AWS, MLOps, Prediction, Real-time AI

Categories:
Machine learning and neural networks
Creator Amazon Web Services (AWS)
Date 2017-11-29
Platforms Cloud Platform, API, AWS Console
Status Active
Website aws.amazon.com
Price Model Pay-as-you-go
Sections:
ML PlatformsModel Deployment

Pricing Details

Pay-as-you-go based on the compute instance time used for hosting. Pricing depends on the instance type. A free tier is available.

Features

Real-time prediction, low latency, scalable infrastructure, model hosting, A/B testing, Multi-Model Endpoints, Asynchronous Inference, integrated with AWS.

Integrations

Deep integration with Amazon S3 (for models and data), Amazon CloudWatch (for monitoring), and other AWS services. API for integration into applications.

Preview

Amazon SageMaker Hosting is a fully managed service from Amazon Web Services (AWS) that allows easily deploying trained machine learning models for real-time, low-latency predictions. The service provides a scalable infrastructure that automatically adjusts to the load, ensuring high availability. SageMaker Hosting supports models trained in SageMaker Training or imported from other sources. It offers various deployment options, including Multi-Model Endpoints for hosting multiple models on a single instance and Asynchronous Inference for handling requests with large payloads. It is a critical component for integrating ML models into production applications.