Chat Bot

local AI chatbot powered by open models. It combines OpenHermes for text generation and LLaVA for image recognition, all running locally through llama.cpp — no external API calls required.

✨ Features

• User management
• Conversation history
• Offline conversation
• Image recognition (temporary and stored images)
• Multiple personalities

🛠 Tech Stack

• Conversational AI using OpenHermes
• Image understanding with LLaVA
• Local inference via llama.cpp
• Python backend for user, conversation and messages curl, model management and prompt orchestration
• React + Vite frontend
• Modular structure (backend and frontend separated for easy customization)
• sqlite3 for user, conversation and messages storage

📁 Project Structure

chat-bot/
├── client/ # React + Vite frontend
├── server/ # Python backend
│ ├── controller/ # HTTP handlers / controllers (API routes -> services)
│ ├── service/ # Model management & inference logic
│ ├── uploads/
│ ├── db/
│ ├── utils/
│ ├── prompts/
│ ├── models/
│ └── app.py

🚀 Getting Started

1. Clone the repository

git clone https://github.com/damian5/chat-bot.git

cd chat-bot

2. Run the backend

See server/README.md for full setup.

3. Run the frontend

See client/README.md for full setup.

HTTPS setup:

• Some browsers require a secure environment to make usage of the camera, for this reason, if you want to use that feature, you will have to run the app using HTTPS

1. Install mkcert

brew install mkcert

brew install nss # if you use Firefox

2. Install the certificates in your computer

mkcert --install

3. Create the certificates

You can choose the URL you want to use, I recommend to use your local hostname for simplicity You can configure this in System preferences > Sharing > local hostname

mkcert <ip_address | local_hostname> 127.0.0.1 ::1

Note If you want to access the app trough another device like your phone, you will need to send the root cert to the device and install it.

mkcert -CAROOT

4. Place your certificates inside /server and /client. You can use the same for both.

├── client/
│ ├── cert.pem
│ ├── cert-key.pem
├── server/
│ ├── cert.pem
│ ├── cert-key.pem

5. Now you can reference them in the vite.js file and app.py

🖼️ Screenshots

Here’s a quick look at the chatbot interface:

📄 License

MIT

Built with ❤️ by Damian