b4659

ggml : optimize and build warning fix for LoongArch (#11709)

* ggml : optimize convert f32<->f16 for loongarch_asx

* ggml : optimize loongarch_asx extend i16,i8,u8 to i32,i16

* ggml : Fix warnings when run cpu CI locally on LoongArch

b4658

llama : fix old glm4 models (#11670)

b4657

sync : ggml

b4651

build : fix llama.pc (#11658)

Signed-off-by: Adrien Gallouët <[email protected]>

b4649

vulkan: optimize coopmat2 iq2/iq3 callbacks (#11521)

* vulkan: optimize coopmat2 iq2/iq3 callbacks

* build: trigger CI on GLSL compute shader changes

b4648

vulkan: initial support for IQ4_XS quantization (#11501)

b4647

vulkan: use smaller combined allocations to avoid fragmentation (#11551)

b4646

metal : avoid breaking build when metal API predates TARGET_OS_VISION…

b4644

metal : adjust support conditions for norm operators (#11671)

cont #11659

ggml-ci

b4643

CUDA: support for mat. mul. with ne03 != ne13 (#11656)