llama.cpp

fix typo in ggml.c

#7603

Merged