💡 Very-Very-Light-Chatbot

A super lightweight, local chatbot project designed for desktop/laptop use — ideal for minimal setups and quick testing.

🛠️ This is my first chatbot project, but it has been refined and optimized to be shared here publicly for others to explore and learn from.

---

🚀 Features

• 💬 Local LLM chatbot powered by Ollama
• 🧠 Custom memory via memory.json
• 📝 Chat and journal history saved in chat_history.json
• ⚡ Super lightweight and clean architecture (can be accesed by others by using local wifi connection)
• 📦 Uses models directly pulled from Ollama (no LM Studio required)
• 🖥️ Optimized for laptops and desktops (offline-capable)
• 🔧 Built for simplicity and extensibility

---

🎞️ Demo Preview

Here’s a look at the chatbot running locally:

---

🖥️ System Requirements

💡 This chatbot was developed and tested on the following specs:

• 🔹 Intel Core i5-13420H
• 🔹 RTX 4050 Laptop GPU
• 🔹 16GB RAM
• 🔹 Windows 11

Minimum recommended:

• ✅ Quad-core CPU
• ✅ 8–16GB RAM
• ✅ Optional: Discrete GPU (for faster performance)
• ✅ Smaller quantized models (like Phi-2) work well on modest hardware

---

📁 Folder Structure

├── app.py                  # Main Flask app
├── brain.py                # Core chatbot flow and logic
├── functions/              # Modular helper functions
│   ├── __init__.py
│   ├── history_func.py     # Manages conversation history
│   ├── journal_func.py     # Handles journaling features
│   ├── memory_func.py      # Memory system logic
│   ├── model_runner.py     # Interacts with LLM via Ollama
│   └── prompt.py           # Prompt creation/injection
├── templates/
│   ├── index.html          # Main chat UI
│   └── history.html        # Journal/history viewer
├── static/
│   ├── style.css           # Basic CSS styling
│   └── script.js           # Frontend behavior
├── memory.json             # Persistent chatbot memory
├── chat_history.json       # Stored conversation/journal logs
├── demo.gif                # Optional: Local UI preview

---

🧠 Ollama-Based Model Setup

This project skips the LM Studio step and directly uses Ollama to pull and run the model:

🔹 Step 1: Install Ollama

Download and install from https://ollama.com

🔹 Step 2: Pull Model (e.g. Phi-2)

ollama pull phi:2

🧠 This project uses the Phi-2 model for its balance of speed and performance. Feel free to swap in other models (like mistral, llama3, or others).

🔹 Step 3: Run the Model

ollama run phi:2

Your local model is now running and ready for chat.

---

🔧 How to Run the App

1. Clone the Project

git clone https://github.com/yourusername/very-very-light-chatbot.git
cd very-very-light-chatbot

2. Create Virtual Environment (Optional)

python -m venv venv
venv\Scripts\activate   # On Windows
# OR
source venv/bin/activate  # On Mac/Linux

3. Install Flask

pip install flask

4. Start Ollama + Run Flask App

Make sure Ollama is running your model (phi:2 or another):

ollama run phi:2
python app.py

5. Open in Browser

Go to:

http://localhost:11434

---

✅ Check Your Versions

python --version
flask --version
ollama --version

✉️ Contact

Built with ❤️ by Pranziss/yubedaoneineed

This is my first public chatbot project — feel free to fork, star ⭐, or reach out with feedback or ideas!