Prompting, RAG, or fine-tuning
On This Lesson 4 topics
This lesson is planned. The outline below lists the topics it will cover. The full lesson has not been published yet.
In this lesson
pick the lightest technique that meets the quality bar.
Planned topics
- Prompting: instructions and examples; cheapest and fastest to change.
- Retrieval-augmented generation (RAG): fetch relevant documents and give them to the model; good for changing or private knowledge.
- Fine-tuning: adjusts model behavior or style from examples; needs data, evaluation, and maintenance.
- Escalate only when evaluation shows the lighter option is not enough.
Draft outline details
Demonstration
a flow: start with prompting and an evaluation set; if knowledge is missing add RAG; if style or format is inconsistent after that, consider fine-tuning, with the cost of each step.
Check your understanding
What is the usual order of escalation?
Show answer
Prompting, then retrieval, then fine-tuning, based on evaluation results.
Practice
Apply this lesson with the module's lab, exercises and quiz.
Marking a lesson complete is optional and you can undo it any time. Your progress stays in this browser.