I’ve been working on this Android project for running local AI, I've posted about this before and the latest version adds a pretty big batch of changes and additions. Main additions in this update: Benchmarking for local LLMs Users can benchmark their device and compare different thread counts to fi