So far, running LLMs has required a large amount of computing resources, mainly GPUs. Running locally, a simple prompt with a typical LLM takes on an average Mac ...