Ian Ingram
An AI Workshop Hosted by
Southern Oregon University
Join us for the AI Retrieval Augmented Generation Workshop hosted by Southern Oregon University in collaboration with
the Regen AI Summit (Summit described below Workshop description).
Apply here
The AI RAG workshop is open to all levels of participants (faculty, researchers, students, HPC/IT admins, and industry).
This workshop is free but registration is required. The event will be held in-person on Southern Oregon University
campus. Some funding is available to support travel. Participants in IT, research, education, and CS undergraduates are
especially encouraged to apply.
The deadline for workshop applications is Midnight, July 17th AoE (Anywhere on Earth).
Notifications will be sent the following week.
Details
Date:
September 13th, 2024 9am - 12pm
Location:
Hannon Library
Southern Oregon University
1290 Ashland St, Ashland OR 97520
Coffee and lunch included.
Travel Awards available (see application form).
Workshop Description:
As artificial intelligence technology continues to evolve at an unprecedented pace, the economic and productivity
benefits of integrating AI tools into our workflows are becoming increasingly apparent. However, this integration is not
without its challenges, particularly in terms of data privacy and accessing specialized data not included in large
language models' (LLMs) pre-training datasets. To address these critical issues, our workshop will introduce
participants to a state-of-the-art Modular Retrieval-Augmented Generation (RAG) workflow. This innovative approach is
designed specifically to use capable offline LLMs, enabling them to utilize proprietary and protected data without the
risk of privacy breaches when implemented using appropriate industry data practices.
This workshop will cover:
AI workflow concepts:
Introduction to LlamaIndex: A leading open source framework used to develop LLM applicatons.
Basics of Retrieval-Augmented Generation: Understanding the components and operation of RAG. Building LLM compatible
vector database for proprietary documents and data.
Implementing RAG with Offline LLMs: Detailed instruction on setting up LLMs to securely access and process external data
sources while operating entirely offline. This workshop will use the latest generation of open source LLMs but the
framework is modular and supports the use of other foundation models such as GPT-4.
Speakers:
Ian Ingram
More speakers to be announced.
Agenda:
9:00-9:15 Set-up/introductions
9:15-9:45 Getting started with ACCESS tools
9:45-10:30 LLM Ops and RAG Concepts
10:30 - 11:00 Spinning up a RAG workflow
11:00 - 11:15 Break
11:15 - 12:00 Testing and optimization techniques/deployment
Objectives:
This workshop will leave participants with knowledge of an LLM workflow that uses RAG to allow users to query files and
documents that AI models wouldn't otherwise have access to. This workflow is presented as a template to assist users in
incorporating AI for increased capability and productivity.
The RAG Workshop 2024 is held in collaboration with Ashland's first major AI event:
For more information, Please contact Ian Ingram at ingrami@sou.edu
Southern Oregon University and ACCESS-CI, a program supported by the National Science Foundation