Cracking the Code to Customizing Large Language Models

Talk details
Level: General

Large Language Models such as ChatGPT have taken the world by intrigue and surprise at how they've been able to analyze and generate text. These general models can be used for a wide variety of tasks. While they're generally interesting for a demo, they have fallen short of many tasks that are actually useful to an end user. Right now, countless hours, days, years of people-hours are spent on labeling data just to improve these models. This talk will touch on that pain now (with a story!) and then discuss some of the programmatic ways of customizing these models to work for an end user. This is an important step to democratizing who can build and customize these models.

Speaker
Reinforce 2023 - Sharon Zhou
Sharon Zhou
Co-Founder at Stealth Startup

Sharon Zhou most recently was a computer science (AI) faculty member at Stanford, where she received her PhD in generative AI and taught generative AI classes to over 80K students on Coursera. Prior to this, she was a PM at Google and various startups, and was the first to major in Classics and Computer Science in Harvard University's history. She is now working on democratizing access a...