Part II. Crafting Agentic Systems
IN THIS LESSON
This is the fourth lecture in the Language Models and Intelligent Agentic Systems course, run by Meridian Cambridge in collaboration with the Cambridge Centre for Data Driven Discovery (C2D3).
This lecture introduces the post-training pipeline for LLMs, which is used to turn them from text generators into economically useful assistants. We discuss the post-training pipeline, and then cover supervised fine-tuning in depth. Finally, we discuss fine-tuning attacks that can be used to remove safety guardrails from models.