Pin llama.cpp 358cafc: ZAYA1-8B decodes 1.9x faster on HRX; concurrent ZAYA on Vulkan 2.5x - #168

Merged

bong-water-water-bong merged 1 commit into

mainfrom

pin/hrx-zaya-kernels

Sep 26, 2026

Commits on Sep 26, 2026

Pin llama.cpp 358cafc: HRX kernels for ZAYA's router and grouped conv, ZAYA1-8B decode 25.5 -> 47.9 tok/s on HRX (llama.cpp #22, #24)

bong-water-water-bong
and
claude
committed
433f7a3