This is the first day of Kaggle's five-day Generative AI Intensive Course, focusing on foundational models and prompt engineering. The session includes a Q&A segment with guest speakers and covers topics ranging from the Gemini app's use of RLHF to improve responses, whether large language models can make new discoveries, and using larger models to train smaller ones through distillation. The speakers also discuss approaches to evaluating large models, chain of thought prompting, and using LLMs for code generation. The session also includes a walkthrough of code examples and concludes with a pop quiz to test the audience's comprehension of the material.
Sign in to continue reading, translating and more.
Continue