llama.cpp

llama : fix incorrect model type for Gemma 270M

#15764

Merged