RAG-driven Generative AI, First Edition

This is the code repository for RAG Driven GenAI, First Edition, published by Packt.

Last updated: December 5, 2024.

See the CHANGELOG.md for details.

Build custom retrieval augmented generation pipelines with LlamaIndex, Deep Lake, and Pinecone

Denis Rothman

                       

About the book

RAG-Driven Generative AI provides a roadmap for building effective LLM, computer vision, and generative AI systems that balance performance and costs. This book offers a detailed exploration of RAG and how to design, manage, and control multimodal AI pipelines. By connecting outputs to traceable source documents, RAG improves output accuracy and contextual relevance, offering a dynamic approach to managing large volumes of information. This AI book also shows you how to build a RAG framework, providing practical knowledge on vector stores, chunking, indexing, and ranking. You'll discover techniques to optimize your project's performance and better understand your data, including using adaptive RAG and human feedback to refine retrieval accuracy, balancing RAG with fine-tuning, implementing dynamic RAG to enhance real-time decision-making, and visualizing complex data with knowledge graphs. You'll be exposed to a hands-on blend of frameworks like LlamaIndex and Deep Lake, vector databases such as Pinecone and Chroma, and models from Hugging Face and OpenAI. By the end of this book, you will have acquired the skills to implement intelligent solutions, keeping you competitive in fields ranging from production to customer service across any project.

Key Learnings

• Scale RAG pipelines to handle large datasets efficiently
• Employ techniques that minimize hallucinations and ensure accurate responses
• Implement indexing techniques to improve AI accuracy with traceable and transparent outputs
• Customize and scale RAG-driven generative AI systems across domains
• Find out how to use Deep Lake and Pinecone for efficient and fast data retrieval
• Control and build robust generative AI systems grounded in real-world data
• Combine text and image data for richer, more informative AI responses

Chapters

This repo is continually updated and upgraded.
📝 For details on updates and improvements, see the Changelog.
🐬 New bonus notebooks to explore, see Changelog.
🚩 If you see anything that doesn't run as expected, raise an issue, and we'll work on it!

Platforms

You can run the notebooks directly from the table below:

Chapter	Colab	Kaggle
Part I The RAG Framework
Chapter 1: Why Retrieval Augmented Generation(RAG)?
RAG_Overview.ipynb
RAG overview with Elon Musk's xAI grok-beta LLM model
🐬RAG_Overview_Grok.ipynb with Elon Musk's xAI grok-beta
Chapter 2, RAG Embeddings and Vector Stores with Deep Lake and OpenAI
1_Data_collection_preparation.ipynb
2_Embeddings_vector_store.ipynb
3_Augmented_Generation.ipynb
RAG with OpenAI Reasoning models: the o1-preview API
🐬3_Augmented_Generation_o1_preview.ipynb
Chapter 3, Building Index-based RAG with LlamaIndex, Deep Lake, and OpenAI
Deep_Lake_LlamaIndex_OpenAI_RAG.ipynb
Chapter 4, Multimodal Modular RAG for Drone Technology
Multimodal_Modular_RAG_Drones.ipynb
Chapter 5: Boosting RAG Performance with Expert Human Feedback
Adaptive_RAG.ipynb
Chapter 6, Scaling RAG Bank Customer Data with Pinecone
Pipeline_1_Collecting_and_preparing_the_dataset.ipynb
Pipeline_2_Scaling_a_Pinecone_Index.ipynb
Pipeline_3_RAG_Generative_AI.ipynb
Chapter 7, Building Scalable Knowledge Graph-based RAG with Wikipedia and LlamaIndex
Tree_2_Graph.ipynb
Wikipedia_API.ipynb
Knowledge_Graph__Deep_Lake_LlamaIndex_OpenAI_RAG.ipynb
Chapter 8, Dynamic RAG with Chroma and Hugging Face Llama
Dynamic_RAG_with_Chroma_and_Hugging_Face.ipynb
Chapter 9, Empowering AI Models: Fine-Tuning RAG Data and Human Feedback
Fine_tuning_OpenAI_GPT-4o-mini.ipynb
Chapter 10, RAG for Video Stock Production with Pinecone and OpenAI
Video_dataset_visualization.ipynb
Pipeline_1_Generator_and_Commentator.ipynb
Pipeline_2_The_Vector_Store_Administrator.ipynb
Pipeline_3_The_Video_Expert.ipynb

Requirements for this book

You should have basic Natural Processing Language (NLP) knowledge and some experience with Python. Additionally, most of the programs in this book are provided as Jupyter notebooks. To run them, all you need is a free Google Gmail account, allowing you to execute the notebooks on Google Colaboratory’s free virtual machine (VM). You will also need to generate API tokens for OpenAI, Activeloop, and Pinecone. You might require to download modules while running the notebooks or you can simply run the requirements_01.txt file in the env you create. Some of the modules are as follows:

Modules	Version
deeplake	3.9.18 (with Pillow)
openai	1.40.3
transformers	4.41.2
numpy	>=1.24.1
deepspeed	0.10.1

Note: This GitHub repository will be continually maintained and updated as the platforms evolve. As such, the versions will evolve in time in this repo so that you will always have access to state-of-art programs!

Get to know Author

Denis Rothman graduated from Sorbonne University and Paris-Cité University, designing one of the first patented encoding and embedding systems and teaching at Paris-I Panthéon Sorbonne.He authored one of the first patented word encoding and AI bots/robots. He began his career delivering a Natural Language Processing (NLP) chatbot for Moët et Chandon(LVMH) and an AI tactical defense optimizer for Airbus (formerly Aerospatiale). Denis then authored an AI optimizer for IBM and luxury brands, leading to an Advanced Planning and Scheduling (APS) solution used worldwide. LinkedIn

Other Related Books

• Transformers for Natural Language Processing and Computer Vision, Third Edition, First Edition
• Generative AI Application Integration Patterns, First Edition