Skip to Main Content
Talk Intermediate

Crafting Specialized AI with Hugging Face and LangChain

Proposal status is Rejected
Session Description

What if you could turn a generic AI large language model into a domain expert by combining retrieval, adaptation, and fine-tuning, all with Hugging Face' Opensource models?


In this talk we will explore multiple techniques like Weight-Modification tuning (PEFT), RAG (Retrieval-Augmented Generation) and prompt engineering to specialize AI models.


Agenda:

  1. Elucidate the need of specializing the LLM
  2. Concepts and code overview of techniques like PEFT(LoRa) using hugging face libraries and langchain for RAG.
  3. Insights on picking the right technique for the specific usecase.
  4. Demo: Contextual Chatbot for a Knowledge Base

Session Categories

FOSS

Speakers

Bharat Rajani
Bharat Rajani Senior Software Engineer | Red Hat

My name is Bharat Rajani and I am a senior software engineer at Red Hat. Over the past 5 years I've dabbled with multiple technologies to solve some of the most interesting problems in distributed computing, data-warehousing and subscriptions space. My passion for open source development and solving problems drives me daily. When not coding, I can be found playing table tennis or strumming guitar.