llama.cpp

clip: Fix llama-llava-clip-quantize-cli quantization error under CUDA backend

#12566

Merged