Duration

Generative AI : Build LLM Application with LangChain

2 days

Overview

This course provides an overview into the world of generative artificial intelligence (AI) and the use of large language models. As a participant, you'll explore the key concepts, techniques, and best practices involved in working with large language models.

LangChain is an open-source framework that allows developers working with AI to combine large language models (LLMs) like GPT-4 with external sources of computation and data. It makes it easy to build and deploy AI applications that are both scalable and performant. It also facilitates entry into the AI field for individuals from diverse backgrounds and enables the deployment of AI as a service. In this course, we'll go over LangChain components, LLM wrappers, Chains, and Agents. We'll dive deep into embeddings and vector databases such as Pinecone.

DAY 1

9:00am - 10:30am

Introduction

Getting Started - AI vs ML vs Generative AI
What is Generative AI
Key Step In Generative AI For Text - Predicting Next Word
Exploring Predictive Machine Learning vs Generative AI

10:30am - 10:45am Morning Break

10:45am - 12:30pm

Understanding LangChain with its Modules

Introduction to LangChain
Setting Up the Environment: LangChain, Pinecone, and Python-dotenv
LLM Models (Wrappers): GPT-3
ChatModels: GPT-3.50

12:30pm - 1:30pm Lunch Break

1:30pm - 3:30pm Prompt, Chain & Agents

Prompt Templates
Simple Chains
Sequential Chains
Introduction to LangChain Agents
LangChain Agents in Action

3:30pm - 3:45pm Afternoon Break

3:45pm - 5:00pm

Vector DB (Pinecone)

Short Recap of Embeddings
Introduction to Vector Databases
Diving into Pinecone, Part 2
Splitting and Embedding Text Using LangChain
Inserting the Embeddings into a Pinecone Index
Asking Questions (Similarity Search)

DAY 2

9:00am - 10:30am

Code-Along #1: Using LangChain Modules Built a Streamlit Application

10:30am - 10:45am Morning Break

10:45am - 12:30pm

Hands-on #1 : Question-Answering Application on Your Custom (or Private) Documents

Project Introduction
Loading Your Custom (Private) PDF Documents
Loading Different Document Formats
Public and Private Service Loaders
Chunking Strategies and Splitting the Documents
Embedding and Uploading to a Vector Database (Pinecone)
Asking and Getting Answers
Adding Memory (Chat History)

12:30pm - 1:30pm Lunch Break

1:30pm - 3:30pm

Hands-on #2: Summarizing With LangChain and OpenAI

Project Introduction
Summarizing Using a Basic Prompt
Summarizing using Prompt Templates
Summarizing Using StuffDocumentsChain