Skip to content
/ RAG Public
forked from amoako420/RAG

Retrieval Augmented Generation (RAG) App with Langchain & streamlit

Notifications You must be signed in to change notification settings

Amoako419/RAG

 
 

Repository files navigation

RAG Application on Paul Graham Essay

Overview

This application is designed to analyze and generate insights from essays written by Paul Graham using Retrieval-Augmented Generation (RAG) techniques. By leveraging advanced text retrieval and generation models, the application provides a comprehensive analysis of the essays, offering new perspectives and deeper understanding.

Features

  • Text Retrieval: Efficiently retrieve relevant sections from Paul Graham's essays.
  • Text Generation: Generate new text based on the retrieved sections.
  • Analysis Tools: Tools to analyze the content and structure of the essays.

Streamlit Link

Access the Streamlit application here.

Installation

To install the necessary dependencies, run:

pip install -r requirements.txt

Usage

  1. Data Preparation: Ensure that the essay is available in the paul_graham_essay.txt file.
  2. Streamlit Interface: Launch the Streamlit interface for interactive analysis.
streamlit run app.py

Directory Structure

RAG/
├── paul_graham_essay.txt      # File containing Paul Graham's essay
├── app.py                     # Streamlit application file
├── README.md                  # This README file
└── requirements.txt           # List of dependencies

Technology Stack

  • LangChain: For building the retrieval and generation pipelines.
  • Gemini 2.0 Flash: For efficient and scalable text processing.
  • Streamlit: For creating an interactive web interface.

Contributing

Contributions are welcome! Please submit a pull request or open an issue to discuss any changes.

License

This project is licensed under the MIT License.

About

Retrieval Augmented Generation (RAG) App with Langchain & streamlit

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 53.4%
  • Python 46.6%