Introduction to RAGxplorer π¦π¦Ί
RAGxplorer is a sophisticated yet user-friendly tool designed to build visual representations of Retrieval Augmented Generation (RAG) models. This tool aims to simplify the visualization process, making it accessible for both beginners and experienced users in the field of AI, particularly those working with information retrieval systems.
Quick Start β‘
Getting started with RAGxplorer is straightforward. The installation process is streamlined, allowing users to set up the tool with a single command:
pip install ragxplorer
Once installed, users can quickly dive into its capabilities. Here's a simple example of how to use RAGxplorer in a Python environment:
from ragxplorer import RAGxplorer
client = RAGxplorer(embedding_model="thenlper/gte-large")
client.load_pdf("presentation.pdf", verbose=True)
client.visualize_query("What are the top revenue drivers for Microsoft?")
In this example, RAGxplorer is used to load a PDF document and analyze it with a query regarding Microsoft's top revenue drivers. The tool then visualizes the results, providing insights at a glance.
For those eager to learn more, a quickstart Jupyter notebook tutorial is available here, along with an interactive Colab notebook linked here.
Streamlit Demo π
RAGxplorer also offers a demo through Streamlit, an application that provides an interactive interface for exploring RAG models. This demo is accessible online at this link, making it easy for users to experience the tool's capabilities without requiring local installation. A visual sample from the demo is also provided to illustrate the application's interface and utilities.
Contributing π
The RAGxplorer project welcomes contributions from the community. Interested individuals can contribute by following the contributing guidelines which are currently a work in progress. Contributions can range from code enhancements to documentation improvements.
License π
RAGxplorer is open-source software, licensed under the MIT License. This ensures the tool is freely available to modify and distribute, with the full license details accessible here.
Acknowledgments π
The development of RAGxplorer has been inspired by the collaborative efforts and educational materials from various resources. Notably, DeepLearning.AI and Chroma have contributed with their insights and codelabs from the Advanced Retrieval for AI course, and the supportive Streamlit community has played a crucial role in providing resources and encouragement during development.
RAGxplorer provides a powerful yet accessible means to engage with RAG visualizations, fostering deeper understanding and insights into retrieval-augmented models through graphical analysis.