Run llama.cpp server

./llama-server -m ./models/<model>.gguf -c 8192