NCSA staff who would like to submit an item for the calendar can email newsdesk@ncsa.illinois.edu.
Registration Required: https://forms.illinois.edu/sec/1821324705
Join the AI User Group on Friday, November 8 from 12:00 to 1:00 PM. All Research Park events are free; however, registration is required to ensure that we have adequate seating and food for all participants.
The aim of the user group is to build a community of interest around data + AI to foster learning, collaboration, and networking. The group will serve both students and professionals to bridge the gap between the analytical sciences and practical applications of industry.
Nathan Price will be presenting on serving large language models (LLMs) at scale. In his talk, he’ll explore the challenges and solutions involved in serving large language models (LLMs) at scale. He'll also cover key trade-offs in selecting the right software platforms, considerations for autoscaling and compiling models, and techniques such as in-flight batching that help optimize performance. The session will highlight both technical decisions and practical experiences from building a robust AI infrastructure for applications.
Speaker Bio:Nathan Price has over 11 years of experience in technology and AI development. He spent 8 years at Sandia National Laboratories, working on DARPA projects, neutron generator development, and machine learning applications for non-proliferation detection. For the past 3 years, Nathan has been at Abridge, an AI healthcare startup, where he focuses on hosting and serving large language models (LLMs) to automate doctor note generation.
Presentation Abstract:In this talk, we’ll explore the challenges and solutions involved in serving large language models (LLMs) at scale. I’ll cover key trade-offs in selecting the right software platforms, considerations for autoscaling and compiling models, and techniques such as in-flight batching that help optimize performance. The session will highlight both technical decisions and practical experiences from building a robust AI infrastructure for applications.
Check out the Meetup Page with more information: https://www.meetup.com/cu-data-ai/ Interested in joining other peer-sharing groups about tech and entrepreneurialism? Be sure to check out our other monthly meetup groups here.