Chat GPT Fine Tuning
John discusses the challenges of fine-tuning Chat GPT models, emphasizing the need for iterative supervised fine-tuning to improve model outputs. He highlights the importance of human intervention to refine the model's understanding of limitations and factuality.In this clip
From this podcast
Dwarkesh Podcast
John Schulman (OpenAI Cofounder) - Reasoning, RLHF, & Plan for 2027 AGI
Related Questions