AWS Certified AI Practitioner
Fine-Tuning Methods
Fine-tuning is an umbrella term. This lesson separates supervised fine-tuning, continued pre-training, and the other Bedrock customization methods, and shows which problem each one solves.
Intermediate 17 minutes 4 Learning Objectives
- Distinguish supervised fine-tuning from continued pre-training by their data and their purpose
- Describe the other Bedrock customization methods: reinforcement fine-tuning and distillation
- Explain what creating a custom model on AWS produces and how it is billed
- Match a customization goal to the right fine-tuning method