llama.cpp

b0ec5218 - metal : support MTLGPUFamily < Apple7, formatting, style (#3524)

metal : support MTLGPUFamily < Apple7, formatting, style (#3524) * metal : improve decoding speed for batches of 2-16 * metal : rename kernels mul_mat_ to mul_mv_ * metal : indentations * minor * metal : print more GPU info + disable mul_mm for MTLGPUFamiliy < Apple7