BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//cfp.scipy.org//2024//talk//W3ZJWG
BEGIN:VTIMEZONE
TZID:US/Pacific
BEGIN:DAYLIGHT
DTSTART:20230710T000000
TZNAME:PDT
TZOFFSETFROM:-0700
TZOFFSETTO:-0700
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20231105T020000
RDATE:20241103T020000
TZNAME:PST
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20240310T030000
RDATE:20250309T030000
TZNAME:PDT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
SUMMARY:From RAGs to riches: Build an AI document inquiry web-app - Pavith
 ra Eswaramoorthy\, Dharhas Pothina\, Andrew Huang
DTSTART;TZID=US/Pacific:20240709T080000
DTEND;TZID=US/Pacific:20240709T120000
DTSTAMP:20260814T033352Z
UID:pretalx-2024-W3ZJWG@cfp.scipy.org
DESCRIPTION:As we descend from the peak of the hype cycle around Large Lan
 guage Models (LLMs)\, chat-based document inquiry systems have emerged as 
 a high-value practical use case. Retrieval-Augmented Generation (RAG) is a
  technique to share relevant context and external information (retrieved f
 rom vector storage) to LLMs\, thus making them more powerful and accurate.
 \n\nIn this hands-on tutorial\, we’ll dive into RAG by creating a person
 al chat app that accurately answers questions about your selected document
 s. We’ll use a new [OSS project called Ragna](https://ragna.chat/en/late
 st/) that provides a friendly Python and REST API\, designed for this part
 icular case. We’ll test the effectiveness of different LLMs and vector d
 atabases\, including an offline LLM (i.e.\, local LLM) running on GPUs on 
 the cloud-machines provided to you. We'll then develop a web application t
 hat leverages the REST API\, built with [Panel](https://panel.holoviz.org)
 –a powerful OSS Python application development framework.
LOCATION:Ballroom B/C
URL:https://cfp.scipy.org/2024/talk/W3ZJWG/
END:VEVENT
END:VCALENDAR