Show HN: Lumabri – What if LLMs worked like Napster?

Lumabri is a peer-to-peer system that allows running large language models across a distributed network of machines, inspired by Napster's decentralized approach. The system enables one machine to share a model while others access it on-demand, downloading only the data needed for inference and caching it locally for faster subsequent queries. It works with CPU and SSD as the primary architecture, making GPUs optional for speed rather than necessary, allowing networks without dedicated GPU resources to function effectively.