llama.cpp

ef035803 - ggml: Add Apple support for GGML_CPU_ALL_VARIANTS (#14258)