AWS · AIP-C01
Deploying foundation models
AIP-C01 · Lesson 10 of 23 · 1:15:20
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: deployment options from Bedrock on demand and Provisioned Throughput to SageMaker AI endpoints, the memory challenges of large models, tensor parallelism, quantization, and LMI containers, loading strategies for large models, and cost-aware deployment with small models, cascades, and intelligent prompt routing. Includes three original practice questions.
Preview the topics
- AnyCompany’s open-weight model is too large for the memory of one GPU. Which technique lets it run on one instance with several GPUs?
- The endpoint for AnyCompany’s 150 GB model keeps failing during startup, because the model download takes longer than allowed. What should the team change?
- Most of AnyCompany’s support questions are routine, but a few need deep reasoning, and sending everything to the largest model is too expensive. What should it do?