Question Answering RAG System

This project is a REST API for a Question Answering system using Retrieval-Augmented Generation (RAG) with OpenAI's language models. The system processes PDF documents, splits them into chunks, stores them in a vector store and allows users to ask questions based on the uploaded documents.

A RAG(Retrieval-Augmented Generation) model is a transformer-based model that combines the benefits of retrieval-based and generation-based models. It uses a retriever to find relevant passages from a large corpus of documents and then generates an answer based on the retrieved passages.

Table of Contents

• Features
• Requirements
• Using Docker
• Installation
• Usage
• API Endpoints
  • Root Endpoint
  • Upload PDF
  • Ask Question
• Postman Collection
• Project Structure
• License

Features

• Upload PDF files and process them into document chunks.
• Store document chunks in a vector store for efficient retrieval.
• Ask questions based on the uploaded documents using a RAG approach.
• Consistent JSON response format for all endpoints.

Requirements

• Python
• FastAPI
• Uvicorn
• PyPDF2
• Langchain libraries

Using Docker

You can also run the project using Docker. Refer to the Docker documentation for more information.

Installation

1. Clone the repository:

   git clone https://github.com/sthsuyash/chat-pdf.git
   cd chat-pdf

2. Create a virtual environment:

   • For Windows:

   python -m venv venv
   .\venv\Scripts\Activate

   • For MacOS/Linux:

   python3 -m venv venv
   source venv/bin/activate

3. Install the required dependencies:

   pip install -r requirements.txt

4. Rename the .env.example file to .env and update the environment variables.

Usage

1. Start the FastAPI server:

   uvicorn main:app --reload

2. The API will be available at http://127.0.0.1:8000.

API Endpoints

Root Endpoint

• GET /

  • Description: Root endpoint to test the API.

  • Response:

    {
      "status": "success",
      "data": {
        "message": "Welcome to the Question Answering RAG system!"
      }
    }

Upload PDF

• POST /upload_pdf

  • Description: Upload a PDF file and process it into chunks for the vector store.

  • Request:

    • File: A PDF file.

  • Response:

    • Success:

      {
        "status": "success",
        "data": {
          "message": "File uploaded successfully."
        }
      }

    • Error:

      {
        "status": "error",
        "data": {
          "message": "Error processing file: <error_message>"
        }
      }

Ask Question

• POST /ask_question

  • Description: Ask a question based on the uploaded documents.

  • Request:

    • JSON body:

      {
        "question": "Your question here"
      }

  • Response:

    • Success:

      {
        "status": "success",
        "data": {
          "answer": "The answer to your question."
        }
      }

    • Error:

      {
        "status": "error",
        "data": {
          "message": "Error answering question: <error_message>"
        }
      }

Postman Collection

You can import the Postman collection to test the API endpoints. The collection is available here.

Project Structure

.
├── pdf/                # Directory to store uploaded PDF files
├── main.py             # Main FastAPI application
├── routes/             # Directory containing API route definitions
│   ├── __init__.py
│   ├── ai.py           # Route for AI operations
│   ├── base.py         # Base route for the API
│   ├── chat.py         # Route for chat operations using pdf
├── services/           # Directory containing service classes
│   ├── __init__.py
│   ├── ai_service.py   # Service class for AI operations
│   ├── pdf_service.py  # Service class for PDF operations
├── utils.py            # Utility functions for the application
├── requirements.txt    # Project dependencies
├── .env.example        # Example environment variables
├── .env                # Environment variables
├── Dockerfile          # Dockerfile for building the project
├── docker-compose.yml  # Docker Compose configuration
├── .gitignore          # Files and directories to be ignored by Git
├── .dockerignore       # Files and directories to be ignored by Docker
├── README.Docker.md    # Documentation for running the project in Docker
└── README.md           # Project documentation

License

This project is licensed under the MIT License. See the LICENSE file for details.