Model Architectures · Practitioner
Customising Models on AWS: Fine-Tuning, Distillation and Custom Model Import
The four ways to change a model's behaviour on AWS, what each actually costs, and the honest test for whether you need any of them.
- 1 hours lesson
- advanced level
- AWS AI Track
- Narrated presentation
What you’ll learn
- Prompting and RAG before fine-tuning: the cost ladder
- Supervised fine-tuning with labelled pairs
- Reinforcement fine-tuning with Lambda-defined reward functions
- Distillation: teacher to student for unit-cost reduction
- Custom Model Import and SageMaker JumpStart for full control
- Training billed per token times epochs, plus monthly per-model storage
Inside this lesson
A narrated, slide-by-slide presentation with 1 prediction exercise, 1 knowledge check, 3 videos. You answer as you go, and the platform checks your work.