module llama [system] {
    header "llama.h"
    header "ggml.h"
    header "ggml-alloc.h"
    header "ggml-backend.h"

    link "llama"

    export *
}