⬆️ Update ggerganov/llama.cpp #2919

Job	Run time
core-image-build (cublas, 12, 5, linux/amd64, false, -cublas-cuda12-ffmpeg-core, true, core, ubun... / reusable_image-build	1h 14m 48s
core-image-build (linux/amd64, false, -ffmpeg-core, true, core, ubuntu-latest, ubuntu:22.04, --jo... / reusable_image-build	21m 17s
core-image-build (sycl_f16, linux/amd64, false, quay.io/go-skynet/intel-oneapi-base:latest, ubunt... / reusable_image-build	31m 58s
core-image-build (vulkan, linux/amd64, false, -vulkan-ffmpeg-core, true, core, ubuntu-latest, ubu... / reusable_image-build	21m 30s
extras-image-build (cublas, 12, 5, linux/amd64, false, -cublas-cuda12-ffmpeg, true, extras, arc-r... / reusable_image-build	1h 26m 2s
extras-image-build (linux/amd64, false, -ffmpeg, true, extras, arc-runner-set, ubuntu:22.04, --jo... / reusable_image-build	50m 59s
extras-image-build (hipblas, linux/amd64, false, -hipblas, false, extras, rocm/dev-ubuntu-22.04:6... / reusable_image-build	3h 35m 7s
extras-image-build (sycl_f16, linux/amd64, false, quay.io/go-skynet/intel-oneapi-base:latest, ubu... / reusable_image-build	1h 4m 47s
	9h 26m 28s

Provide feedback