Serverless Deep Learning
summary
Demonstrates running inference as a serverless API by bundling a PyTorch MNIST model into AWS Lambda container images, building and testing locally with the AWS SAM CLI, pushing the image to ECR, and deploying an API Gateway + Lambda stack via SAM deploy; the deployed endpoint returns the correct prediction for the sample “3.” Shows exact commands used (sam init/build/invoke, aws ecr create-repository, sam deploy guided), notes Lambda container limit of 10 GB, and recommends serverless for prototyping and parallel batch jobs rather than high-performance serving.
AI-written summary · about
chapters (7)
- 0:00Serverless ML introduction
- 1:24AWS Lambda for ML
- 3:55Deploy ML with SAM CLI
- 7:49Local testing & builds
- 9:55Cloud deployment with SAM
- 10:59Testing serverless ML API
- 11:42Serverless ML considerations
related talks
41mBuilding Serverless Machine Learning Models in the Cloud
34mBuilding scalable end-to-end deep learning pipelines in the cloud
15mServerless Tensorflow 1.0.0 on AWS. Wait what?
01h 27mDeploying a Machine Learning Model to the Cloud Using AWS Lambda
39mBuilding model testing infrastructure and scaling AI predictions
18mServerless AI's Rise: Revolutionizing ML Deployment
04mServerless applications, Microservices, AWS
02h 37mServerless Deep Dive