git.djapps.eu Git - pkg/ggml/sources/ggml/commit

]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit

overview / pkg / ggml / sources / ggml / commit

author	Jeff Bolz <redacted>
	Fri, 26 Dec 2025 22:12:58 +0000 (16:12 -0600)
committer	Georgi Gerganov <redacted>
	Wed, 31 Dec 2025 10:39:43 +0000 (12:39 +0200)
commit	82db5d886a354f05b2745625ee1e981f501133a4
tree	e177dad4fe1e9a7f103fb14f7c423489e61839d9	tree
parent	792ec160b359326c8f80cf2e50ae9b9b0e68fbc6	commit \| diff

vulkan: preprocess mul_mat_id experts and discard workgroups more quickly (llama/18352)

Run a preprocess to count how many times each expert is used, and use this to
quickly discard workgroups that aren't needed.

Packaging of ggml-org/ggml

RSS Atom

src/ggml-vulkan/ggml-vulkan.cpp		diff \| blob \| history
src/ggml-vulkan/vulkan-shaders/count_experts.comp	[new file with mode: 0644]	blob
src/ggml-vulkan/vulkan-shaders/mul_mm.comp		diff \| blob \| history
src/ggml-vulkan/vulkan-shaders/mul_mm_cm2.comp		diff \| blob \| history
src/ggml-vulkan/vulkan-shaders/mul_mm_id_funcs.glsl		diff \| blob \| history
src/ggml-vulkan/vulkan-shaders/mul_mmq.comp		diff \| blob \| history
src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp		diff \| blob \| history