AWS · AIP-C01
The lifecycle of customized models
AIP-C01 · Lesson 3 of 23 · 17:40
YouTube loads only after you press play. This chapter opens at its timestamp in the full course. Watch on YouTube ↗
In this lesson
Prepare for the AIP-C01 exam (AWS Certified Generative AI Developer - Professional). In this lesson: hosting fine-tuned models on SageMaker AI endpoints with inference components, LoRA adapters, SageMaker Model Registry and deployment pipelines, blue/green, canary, and linear traffic shifting with automatic rollback, shadow tests, and the Bedrock model lifecycle, from Active to Legacy to End-of-Life. Includes three original practice questions.
Preview the topics
- AnyCompany needs twelve variants of the same base model, one per regional brand voice, and twelve full copies would need twelve times the GPU memory. What should it use?
- AnyCompany wants to send ten percent of live traffic to a new model version first, and switch back automatically if errors rise. What should it configure?
- A Bedrock model that AnyCompany uses has just entered its legacy period. What should the team do?