llama.cpp

099afc62 - llama : fix quantization when tensors are missing (#5423)