Time: Saturday, August 29, 2026, 9:00 AM - 5:00 PM. Lunch included.

Location: In person at NC State University, online via Zoom.

Co-hosted by: Research Triangle AI Society and AI Hub for Science of NC State University.

Registration Fee
Regular: $300 / $270 early bird before August 15
RTAI Pro Member: $280 / $250 early bird before August 15
College and K-12 students: $150. Email info@research-triangle.ai to request a coupon code with student ID.

Instructor: Dr. Paul Liu, Director of AI Hub for Science at NC State University

Dr. Liu is a professor at NC State University and Director of the AI Hub for Science. His AI expertise spans LLM fine-tuning, AI agent and RAG system development, and large dataset processing and modeling. Dr. Liu is the author of How to Build and Fine-Tune a Small Language Model and Generative AI for Science, making him a leading voice in applying AI to scientific domains.

By the end of this full-day hands-on workshop, participants will be able to:

• Describe how small and large language models are structured, including tokenization, embeddings, attention, and transformer blocks, and explain where SLMs fit relative to LLMs in scale, cost, and use case.
• Prepare and format their own datasets for each stage of training, from raw pre-training text to instruction pairs and preference data.
• Pre-train a small language model from scratch and interpret what it learns at this stage, along with its limits.
• Apply supervised fine-tuning (SFT) to turn a base model into an instruction-following assistant with a consistent voice and identity.
• Use Direct Preference Optimization (DPO) to align a model's responses with preferred behaviors.
• Fine-tune efficiently with Low-Rank Adaptation (LoRA), and judge when parameter-efficient methods are preferable to full fine-tuning.
• Deploy a small language model locally and on a server, including as an OpenAI-compatible API endpoint.
• Access and integrate their model from applications, including chat and function/tool calling.