| Description: |
Unlocking the Power of Large Language Models for Research Innovation
Date and time
Thursday, May 2, 2024; 12:00 - 1:00 PM EDT
Presenter
Juan C. Verduzco, Ph.D., Network for Computational Nanotechnology (NCN) at Purdue University
Register
https://purdue.webex.com/weblink/register/r953e51dc946b89510c918d72ebe0c787
In this hands-on workshop, Dr. Juan C. Verduzco will showcase the transformative potential of large language models (LLMs) and their applications to research tasks. First, this presentation explores the essential building blocks of LLMs, such as embeddings, attention mechanisms, and transformer blocks for attendees to understand the mechanics behind these powerful tools. Following this foundation, the focus will shift to practical strategies for customizing LLMs, including prompt engineering, retrieval-augmented generation, and fine-tuning, enabling researchers to tailor these technologies to their unique requirements. Finally, Dr. Verduzco will bring these concepts to life with demonstrations that use these tools to summarize academic papers and debug plotting code, illustrating the practical applications and versatility of LLMs to enhance research productivity and problem-solving.
Please note, this webinar is Part 1 of 2. You can register for Part 2 here.
Presenter bio
Dr. Juan Carlos Verduzco is a Senior Data Science Engineer at the Network for Computational Nanotechnology at Purdue University. His contributions to the nanoHUB platform for online simulations and data involve the use of artificial intelligence to improve the user experience in education and research. Dr. Verduzco's research interests lie in accelerating the discovery and optimization of new materials through the combination of experiments, simulations, artificial intelligence and cyberinfrastructure. Dr. Verduzco earned his Ph.D. in Materials Engineering from Purdue University and a B.S. in Mechatronics Engineering from the Monterrey Institute of Technology and Higher Education. |