🐺 A tiny model powering an autonomous AI trading floor.

Developer @ashdebugs built Wall Street of AI Agents with MiniCPM5-1B-GGUF as the cognitive engine.

Using the same 1B model, four AI traders with different personalities make independent decisions, react to market changes, interact with each other, and compete in a continuously running trading simulation.

The project explores how lightweight models handle: 🧠 Instruction following 📈 Market adaptation 🤝 Multi-agent collaboration

Built with local inference and llama.cpp, the project shows how lightweight models can power complex autonomous agent systems without cloud APIs.

🔗 Try the demo: https://huggingface.co/spaces/build-small-hackathon/Wall-Street-of-AI-Agents

🧩 MiniCPM5-1B-GGUF: https://huggingface.co/openbmb/MiniCPM5-1B-GGUF