Pragmatic Machine Learning in the Cloud
summary
Shows concrete patterns and working code to run ML on AWS SageMaker: built-in algorithms (about 17 algorithms), dataset processing, Debugger snapshots to S3, AutoTuning that runs many trials and often finds a best model in 15–20 minutes, and SHAP explainability jobs that return results in 3–7 minutes. Demonstrates deploying multi-model endpoints and Lambdas, claims a working POC in 4–8 hours, and emphasizes transient instances and managed spot to cut cost while keeping reproducible pipelines.
AI-written summary · about
chapters (15)
- 0:00Introduction to machine learning in the cloud
- 1:16What is ML and frameworks
- 4:29Practical ML tools
- 7:18SageMaker simplifies ML
- 10:34Cost optimization with SageMaker
- 13:02SageMaker best practices
- 15:34Migrate ML code to SageMaker
- 18:44Key SageMaker features
- 23:46Debugging with SageMaker Debugger
- 27:02Auto model tuning in SageMaker
- 30:14ML explainability with Clarify
- 34:25Deploying models with SageMaker
- 37:32Automating ML workflows
- 42:10SageMaker evolving features
- 43:11Conclusion
related talks
32mBuild, Train, and Deploy Machine Learning Models on AWS
33mBuild, train, and deploy ML models using Amazon SageMaker
37mStreamlining Machine Learning Operations
41mBuilding Serverless Machine Learning Models in the Cloud
34mLat Designing and Building Complex Machine Learning....
40mCloud-based Machine Learning Offerings - Beginner's Perspective
45mDeploying ML Models to Production with Amazon SageMaker
19mUsing Amazon SageMaker to develop and deploy ML models