Episode-08 | From a Base Model to an AI Assistant
Understand how a base AI model is transformed into a useful AI assistant through training, prompting, and additional capabilities.
More Resources & Guides
What is the difference between a base model and an AI assistant?What are the main stages involved in building an AI assistant?What happens during the pre-training of a large language model?Why must raw internet data be cleaned before training an AI model?How does the quality of training data affect an LLM’s performance?What is supervised fine-tuning?How does instruction tuning teach a model to follow user requests?What is a reward model, and how is it trained using human preferences?What is reinforcement learning with human feedback (RLHF)?What limitations can human bias introduce into RLHF?