llama.cpp server

./build/bin/llama-server -m model.gguf --port 8080