Conversion of z-lab DFlash models to the GGUF format using llama.cpp.