👁️ Project IRIS

Project IRIS is a robust, containerized ecosystem for running Large Language Models (LLMs) locally and privately.

It orchestrates a high-performance AI stack using Docker Compose, bridging the Ollama inference engine with modern, user-friendly web interfaces, fully optimized for Nvidia GPU acceleration.

🏗️ The Stack

Project IRIS connects three powerful components in an isolated network (iris-network):

1. 🧠 Ollama (Backend): The engine that runs the models (Llama 3, Mistral, Gemma, etc.). Configured for full Nvidia GPU passthrough.
2. 💬 Lobe Chat (UI 1): A modern, high-performance chatbot interface with plugin support and a clean design. accessible at port 3210.
3. 🌐 Open WebUI (UI 2): A feature-rich interface (formerly Ollama WebUI) offering advanced chat history, RAG (Retrieval Augmented Generation), and model management. Accessible at port 3211.

🚀 Prerequisites

• Docker & Docker Compose installed.
• Nvidia GPU (Optional but recommended).
• Nvidia Container Toolkit installed (Required for the GPU configuration in docker-compose.yml).

🛠️ Installation & Usage

1. Clone the repository

git clone https://github.com/Cluyverth/Project-IRIS.git
cd Project-IRIS

2. Start the Ecosystem

Run the stack in detached mode:

docker compose up -d

3. Access the Interfaces

Service	URL	Description
Lobe Chat	http://localhost:3210	Modern, aesthetic chat interface.
Open WebUI	http://localhost:3211	Advanced UI with RAG and management features.
Ollama API	http://localhost:11434	Direct API access.

🧠 Managing Models

Since the Ollama container starts empty, you need to pull a model first. You can do this via the Open WebUI interface or the command line:

# Example: Pulling Llama 3 inside the container
docker compose exec ollama ollama run llama3

📂 Project Structure

.
├── docker-compose.yml   # Orchestration logic
├── .gitignore           # Ignores Python caches & personal docs
├── .dockerignore        # Optimizes build context
└── PersonalDocsFolder/  # (Ignored) Place for local RAG documents/scripts

🐍 Python Development

This project is set up for Python development/scripting (e.g., for custom RAG pipelines or automation). The .gitignore is configured to keep your environment clean by ignoring __pycache__, .venv, and .env files.