Скачать или смотреть SageMaker Cost Optimization (Part 1): Serverless Endpoints

Welcome to Mount Pointe Channel!


This video, the first in a series on SageMaker cost optimization, explains how to use serverless endpoints to reduce costs. The video highlights key benefits such as paying per inference, automatic scaling, and eliminating idle costs. It provides a detailed comparison, showing how serverless endpoints can be significantly cheaper than real-time endpoints for unpredictable or low-traffic workloads. The video also includes a hands-on demo, guiding you through the steps of deploying a SageMaker serverless endpoint, from importing libraries and setting up the environment to packaging the model, uploading it to S3, and testing the final deployed endpoint. It's a great resource for anyone looking to make their machine learning deployments more cost-effective.


Github Link: https://github.com/tidika/sagemaker-c...

Chapters
0:00 Introduction to SageMaker Serverless Endpoints
0:45 Pay-per-inference and automatic scaling
2:35 Ideal use cases for serverless endpoints
3:31 When NOT to use serverless endpoints
4:40 Cost comparison analysis
6:55 Hands-on demo begins
11:14 Deploying the serverless endpoint
13:37 Testing the deployed model