llama.cpp

export-lora : fix issue with quantized base models

#8687

Merged