DebatingAI

A Python-based AI debate system that facilitates structured discussions between two AI models (Zephyr and Mistral) using Ollama. The system provides a real-time, interactive debate experience with distinct AI personalities and perspectives.

Features

• 🤖 Dual AI Debate: Engage two different AI models (Zephyr and Mistral) in a structured debate
• 🎨 Colorful Interface: Clear visual separation between speakers with color-coded responses
• 💬 Real-time Streaming: Watch the debate unfold in real-time with streaming responses
• 🔄 Robust Error Handling: Automatic retries and graceful error recovery
• 🎯 Structured Conversation: Maintains conversation history and context
• 🎭 Distinct Personalities: Each AI has a unique debating style and perspective

Technical Details

Architecture

• Object-oriented design with DebateManager class for state management
• Type-hinted code for better IDE support and maintainability
• Modular structure for easy extension and modification

AI Models

• Zephyr (7B): Empathetic and thoughtful debate partner
• Mistral (7B): Dynamic and insightful counter-perspective
• Both models run locally via Ollama

Error Handling

• Automatic retry mechanism for failed API calls
• Graceful degradation when responses fail
• Clear error messages and status updates

Requirements

• Python 3.8 or higher
• Ollama installed and running locally
• Required Python packages (see requirements.txt)

Setup

1. Install Ollama:

   # macOS/Linux
   curl https://ollama.ai/install.sh | sh

2. Pull Required Models:

   ollama pull zephyr:7b
   ollama pull mistral:7b

3. Install Python Dependencies:

   pip install -r requirements.txt

4. Start Ollama:

   ollama serve

Usage

1. Run the script:

   python DebatingAI.py

2. Enter a debate topic when prompted

3. Watch the AI models debate in real-time

4. Choose to continue or end the debate after each round

Code Structure

DebatingAI/
├── DebatingAI.py      # Main application file
├── requirements.txt   # Python dependencies
└── README.md         # This documentation

Key Components

• DebateManager: Core class managing debate state and AI interactions
• call_ai(): Generic method for AI model communication
• add_message(): Message history management
• rainbow_text(): UI enhancement for user prompts