📄Abstract

This project presents an AI-powered text summarization tool, LLaMA Text Summarizer, which leverages the LLaMA 2 large language model via the Ollama framework. The system provides a simple and efficient way to summarize long-form text inputs entirely offline. It combines a responsive Streamlit-based frontend with a FastAPI backend to handle user inputs and model communication, offering a smooth user experience for real-time summarization tasks.

🔬 Methodology

System Architecture:

The application is structured into two main components:

• Frontend: Built with Streamlit, this interface allows users to input long texts and receive summarized outputs with a single click.
• Backend: Developed with FastAPI, the backend receives text data, sends it to the LLaMA 2 model through Ollama, and returns the summarized version.

Technology Stack:

• Frontend: Streamlit (Python)

• Backend: FastAPI

• Model Execution: Ollama (for running LLaMA 2 locally)

• Language Model: LLaMA 2

• HTTP Communication: requests Python library

Setup Workflow:

• Clone the repository and set up a Python virtual environment.

• Install dependencies via pip or requirements.txt.

• Install Ollama from https://ollama.com and pull the LLaMA 2 model locally.

• Launch the FastAPI server and Streamlit UI separately.

Summarization Flow:

• User inputs text via the Streamlit UI.
• The text is sent as a POST request to the FastAPI backend.
• Backend communicates with the locally running LLaMA 2 model via Ollama.
• Model generates a summary which is returned to and displayed on the frontend.

🛠️ Setup & Execution

1. Set Up a Virtual Environment

python -m venv venv
# macOS/Linux:
source venv/bin/activate
# Windows (PowerShell):
.\venv\Scripts\Activate.ps1
# Windows (CMD):
.\venv\Scripts\activate.bat

2. Install Dependencies

 pip install fastapi uvicorn streamlit requests python-multipart

3. Install Ollama and Pull the LLaMA 2 Model

ollama pull llama2

5. Run the Application
   Start the FastAPI Backend

uvicorn backend.main:app --reload

Start the Streamlit Frontend (in a new terminal)

streamlit run frontend/app.py

📊 Results

The LLaMA Text Summarizer has demonstrated strong performance in summarizing lengthy textual content in real-time, all while running entirely offline. Key advantages include:

• Privacy-preserving inference (no external API calls)

• Low-latency summarization using local LLaMA 2 models

• Cross-platform compatibility through Python-based setup

• User-friendly experience with a clean UI/UX
  This tool is suitable for individuals or organizations needing quick, offline summarization of documents, articles, and long-form content.

🔗 Resources

• 💻 GitHub Repository: https://github.com/dhiaselmi1/text-summarizer-llama
• 🦙 Ollama: https://ollama.com