Introducing Hexllama Hey, I’ve always found llama-server to be more than enough for testing out local models, mostly because it guarantees you always have the absolute latest llama.cpp features and architecture support. But keeping track of different CLI commands, context sizes, and batch settings f
