GenAI4Lib: A Practical Introduction to Large Language Models
The developers from the Digital Products & Data Curation workgroup at Northwestern University Libraries are proposing an immersive, full-day workshop designed to introduce and deepen your understanding of Large Language Models (LLMs) in the context of library services and operations. In “GenAI4Lib,” participants will learn from practitioners with firsthand experience creating, configuring, and deploying LLM-centric applications on the web.
- Real World Insights: Gain “behind-the-scenes” access to Northwestern University Library’s Digital Collections chat prototype to explore LLM content generation using state-of-the-art commercial models
- Comprehensive Overview: Understand the current capabilities of generative AI models, both open-source and commercial, and develop an intuition for the numerous factors that can influence LLM output
- Direct Interaction: Learn how to use LLMs to generate content locally with open-source models
- Innovative Techniques: Discover how Retrieval Augmented Generation (RAG) can ground LLMs in institutional data, addressing issues like “hallucination”
Learning Objectives: At the workshop's end, participants will be able to answer questions about LLMs, multimodal models, vector databases, embeddings, tokens, context windows, LLM "hallucinations," and how to use this technology, even without a budget.