Human-AI Chatbot

This project is a full-stack chatbot application powered by FastAPI for the backend and React + TypeScript for the frontend. The chatbot uses a fine-tuned Qwen2.5-3B-Instruct model to generate responses.

📂 Project Structure

• /frontend # React + TypeScript (User Interface)
• /backend # FastAPI + SQLite (API and ML Inference)

---

🚀 Getting Started

1️⃣ Prerequisites

Ensure you have the following installed:

• Node.js (Recommended: >=18.x)
• Yarn (npm install -g yarn)
• Python 3.12
• Virtual Environment (venv)
• SQLite (comes bundled with Python)
• CUDA (for GPU acceleration, optional)
• Unsloth (Fine Tuning)

---

Download the saved GGUF Weights from GDRIVE - https://drive.google.com/file/d/12JloQZULnGc0NTeLDryThee7SmX3jmLJ/view?usp=sharing

Video Link to see the project in action - https://drive.google.com/file/d/1MjoGwq-ZAmMc6bu7eIyMeyYQtvN5RiTO/view?usp=sharing

For Environment details please go through the 'environment.yml' file. Main Libraries used are -

Unsloth Pytorch

---

🔧 Backend Setup (FastAPI + SQLite)

1. Navigate to the backend folder:

   cd backend

2. Create a virtual environment:

   python3 -m venv venv
   source venv/bin/activate  # MacOS/Linux
   venv\Scripts�ctivate     # Windows

3. Install dependencies:

   pip install -r requirements.txt

4. Start the FastAPI server:

   uvicorn app.main:app

5. Open Swagger Docs:

   • API Docs: http://127.0.0.1:8000/docs

---

🌐 Frontend Setup (React + TypeScript)

1. Navigate to the frontend folder:

   cd ../frontend

2. Install dependencies:

   yarn install

3. Start the development server:

   yarn start

4. Open the browser and go to:

   • http://localhost:3000

---

📡 API Endpoints

User Authentication

Method	Endpoint	Description
POST	/users/signup	Register a new user
POST	/users/login	Authenticate and receive a token
GET	/users/me	Fetch logged-in user details

Chat API

Method	Endpoint	Description
POST	/chats/	Ask a question, receive AI-generated response
GET	/chats/history	Fetch last 15 questions & answers

---

🛠 Deployment

Backend (FastAPI)

• Use Gunicorn or Uvicorn for production:

  uvicorn app.main:app --host 0.0.0.0 --port 8000

• Deploy via Docker, AWS Lambda, or GCP Cloud Run.

Frontend (React)

• Build for production:

  yarn build

• Deploy to Vercel, Netlify, or GitHub Pages.

---

🏗 Future Improvements

• OAuth authentication (Google, GitHub)
• Persistent chat history with PostgreSQL
• Streaming AI responses
• Frontend UI enhancements