ragbook notebooks

Report Abuse

Basic Information

This repository collects interactive Jupyter notebooks that serve as the companion material for the Towards AI RAG book. The notebooks are organized by chapter and cover foundational and advanced topics such as transformer architectures, prompt engineering, retrieval augmented generation, LangChain and LlamaIndex usage, agent construction, multimodal examples, fine tuning, reward modeling and inference benchmarking. Each listed notebook can be opened in Google Colab via the provided links. The collection is intended to let readers follow step by step examples from the book, reproduce experiments, and explore practical implementations of RAG pipelines, prompt templates, chains, output parsers and knowledge graph creation. It functions as hands-on, tutorial style material mapped to the book chapters.

Links

App Details

Features
Organized notebooks by chapter with direct Google Colab links for interactive execution. Coverage includes transformer architectures, prompt engineering tips and tricks, LangChain tutorials, LlamaIndex introductions and RAG agent examples. Includes practical projects such as news article summarizers, YouTube summarizers using Whisper and LangChain, knowledge graph creation from text, text splitters, chains and self critique chains. Advanced content covers RAG metrics and evaluation, LangSmith introduction, building agents with OpenAI Assistants and AutoGPT, multimodal finance examples with DeepMemory, and benchmarking inference. Several notebooks focus on fine tuning workflows including LIMA, QLoRA, Cohere examples and RLHF and reward model fine tuning.
Use Cases
The repository provides a practical learning path for practitioners, students and researchers who want hands on exposure to RAG and agent workflows. Readers can use runnable notebooks to learn core concepts such as prompt templates, example selectors, output parsing, vector stores and retrieval pipelines. It supplies end to end examples for building summarizers, agents that generate analysis reports, and multimodal experiments that combine audio and text. The fine tuning notebooks offer templates for model adaptation and reward model work while the evaluation notebooks show how to measure RAG performance. Overall the materials reduce setup friction by providing ready examples linked to Colab and map directly to book chapters for guided study.

Please fill the required fields*