Description:

LlamaIndex is a toolkit designed to enhance the utility of Large Language Models (LLMs). It provides powerful capabilities to bridge the gap between your custom data and LLMs, thereby enabling the construction of sophisticated, data-driven applications. With LlamaIndex, harnessing the potential of retrieval and generation in language models becomes seamless and efficient.

In this hands-on workshop, we will dive into the intricacies of the Retrieval Augmented Generation (RAG) paradigm, demonstrating how it can be leveraged to build potent systems such as Q&A systems, chatbots, and data agents. A core component of our session will be the exploration of LlamaIndex as an essential bridge between LLMs and your custom data, equipping participants with the knowledge to craft tailored applications efficiently. Here is the outline of the workshop.

Introduction

• Brief on Large Language Models and LlamaIndex.
• Overview of the Retrieval Augmented Generation (RAG) paradigm

Understanding RAG Paradigm

• Detailed explanation of RAG and its applications
• How RAG enhances Large Language Models with custom data

Mastering Indexing and Querying Techniques

• Preparing a robust knowledge base using LlamaIndex's data connectors and indexing capabilities
• Understanding the retrieval process for the most relevant context given a user query
• Synthesizing responses using LLMs and LlamaIndex for effective querying
• Practical demonstration and hands-on practice of these techniques

Working with Advanced Building Blocks

• Practical experience with retrievers, node postprocessors, and response synthesizers
• Understanding the use of these tools in different applications

Router Engines

• Introduction to Router Engines as decision-making systems in LlamaIndex
• Choosing the right query engine/index based on user’s query

Evaluation

• Response Synthesis Evaluation.
• Response + Context Evaluation.
• Question Generation and Evaluation.

Exploring Data Agents

• Understanding Data Agents and their role in interacting with data
• Dynamic interaction with external tools using Data Agents

Prerequisites:

1. Python and Experience with Large Language Models (LLMs) Participants should have a fundamental understanding of Python programming and prior experience with Large Language Models such as ChatGPT, which will provide a basic comprehension of LLM workings.

2. Access to Google Colab Our sessions will be conducted using Google Colab, so please ensure you have access to this platform.

3. OpenAI API Key We'll be utilizing GPT-based models by default for building applications with LlamaIndex, so having an OpenAI API key will be essential.

Video URL:

https://www.youtube.com/watch?v=7H7sNDg6j34

Content URLs:

Here is a version of the slides for the workshop. Notebooks for each section will be updated after the proposal acceptance.

Speaker Info:

I am an Open Source Contributor at LlamaIndex and Senior Data Scientist at Glance (Inmobi).

1. Contributed different data loaders and evaluation modules to LlamaIndex.
2. Built Recommender Systems, NLP, and GenAI applications at Glance. I am part of Glance TV team which is a product built entirely using LLM's and vector DB's. Demo
3. I have published papers at ACL and COLING workshops.
4. I have given talks about LlamaIndex at the Hasgeek GenAI meetup, Fifth Elephant Conference, Analytics Vidhya Data Hack Summit, and different VC firms (Accel, Together, Speciale).
5. Published blogs and videos on LlamaIndex.
6. Recognised as Top 5 GenAI Experts in India by Analytics Vidhya at Data Hack Summit.

Speaker Links:

Linkedin

Twitter

Github

Medium