ggml-org/llama.cpp b9558

on GitHub

vulkan: Use cm2 decode_vector for mul_mat_id B matrix loads (#23991)

This allows vec4 loads of the B elements. Also increase BK to 64 when this is
enabled. Neither of these alone is consistently faster, but together these give
a nice speedup.

In ggml-vulkan.cpp, we need to make sure the B matrix alignment and stride are
multiples of 4.