Chapter 1: Overview of the LLM

__1.1 LLM Concept
____1.1.1 Language Model
____1.1.2 Giant Language Model
__1.2 LLM Characteristics and Types
____1.2.1 Characteristics of LLM
____1.2.2 Types of LLM
____1.2.3 LLM and GAI, SLM
__1.3 LLM Creation Process
__1.4 Additional Considerations After Creating an LLM

Chapter 2: Using LLM

__2.1 How to use LLM
____2.1.1 Fine Tuning
____2.1.2 RAG
____2.1.3 Few-shot running
__2.2 Precautions when using LLM
__2.3 Limitations of LLM

Chapter 3: RAG Overview

__3.1 RAG concept
__3.2 RAG implementation process
____3.2.1 Information Retrieval
____3.2.2 Advanced Information Retrieval
____3.2.3 Text Generation
__3.3 What is needed to implement RAG
____3.3.1 Data
____3.3.2 Vector Database
____3.3.3 Framework (Langchain)

Chapter 4: Getting Familiar with the Langchain

__4.1 Overview of the Rank Chain
__4.2 Configuring the environment for using Langchain
____4.2.1 Configuring the Anaconda Environment
____4.2.2 Installing required libraries
____4.2.3 Key Issuance
__4.3 Langchain main modules
____4.3.1 Model I/O
____4.3.2 Data Connection
____4.3.3 Chain
____4.3.4 Memory
____4.3.5 Agents/Tools

Chapter 5: Implementing RAG with Langchain

__5.1 Creating a Simple Chatbot
__5.2 Creating a RAG-based chatbot
__5.3 Creating a PDF Summary Website
__5.4 Creating a Standalone Question Chatbot
__5.5 Creating a Conversational Chatbot
__5.6 Creating a Translation Service
__5.7 Creating a mail composer
__5.8 Analyzing CSV Files

Chapter 6: Understanding Services Using LLM

__6.1 Call Center
__6.2 Product Recommendation
__6.3 Insurance Underwriting
__6.4 Code Generation and Review
__6.5 Sentence Generation, M365 Copilot

Appendix How to Use Colab
__A.1 How to use Colab
__A.2 Uploading files to Colab

Search

Anyone can learn and create an LLM easily and quickly!
By creating eight available LLM services and practicing various scenarios,
Let's make it with our own hands using our data!

What this book covers: LLM concepts, principles, and examples, and how to use RankChain.

Anyone can learn LLM and create LLM services.
Rather than simply following along, if you understand the concepts and principles, learn the rank chain, and apply it in various ways, you will be able to experience LLM much more efficiently.
- [Concept] Understanding the concept of what LLM, fine tuning, and RAG are and what they can do.
- [Operating Principle] Understanding the operating principle through the LLM generation process and RAG implementation process.
- [How to use Langchain] Setting up an environment for using Langchain & practicing major modules
- [Implementation Method, Practice] Implementing 8 LLM services at a practical usable level with Langchain.
- [Case] ​​Case of a service that actually utilizes LLM

This book's various practical scenarios: Creating eight usable LLM services.

Let's create an LLM service together using Langchain and LLM.
We've structured the scenarios in a variety of practical use cases, and kept the code short and focused on the core content.
- Simple chatbot using LLM
- RAG-based chatbot using Langchain and ChatGPT
- A website that summarizes PDFs
- A chatbot that asks standalone questions to PDF files.
- Conversational chatbot
- Translation service
- Mail composer
- Analyzing CSV files using LLM

Author's Preface

Anyone can quickly and easily create an LLM service by simply using the created LLM rather than creating the LLM directly.
In particular, with the emergence of the LLM development platform called Langchain, service development using LLM has become considerably easier.
However, if you don't have any knowledge about artificial intelligence or don't know how to develop it, you may think that this story has nothing to do with you.
So this book is for people like you:


- People who have little knowledge of artificial intelligence but want to create a service using LLM
- Someone who doesn't know much about development but still wants to create a service using LLM

As this book is written for beginners, the code used in the exercises is short and focused on core content, and the scenarios are also structured in a variety of ways.
Of course, it may be difficult at first, but I recommend that you create a service using the provided code as is and then modify it little by little to implement your own service.