]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
metal: optimise `GGML_OP_SUM` (#16559)
authorSam/Samuel <redacted>
Wed, 15 Oct 2025 14:05:56 +0000 (23:05 +0900)
committerGitHub <redacted>
Wed, 15 Oct 2025 14:05:56 +0000 (17:05 +0300)
commitf4ce81c45e7bd910e36bf44c253fc5255c49b1e4
treef6336240ef27e4d7ac8f396704f7cdb5ca99b3c3
parent17304cbcc1dd24de7741cbe57925d58e90a98ac1
metal: optimise `GGML_OP_SUM` (#16559)

* optimise GGML_OP_SUM

* add non-contiguous tests by permuting the input

* change tests to require full contiguity of OP_SUM

* cuda : add check GGML_OP_SUM

---------

Co-authored-by: Georgi Gerganov <redacted>
ggml/src/ggml-cuda/ggml-cuda.cu
ggml/src/ggml-metal/ggml-metal-device.m
ggml/src/ggml-metal/ggml-metal-ops.cpp
ggml/src/ggml-metal/ggml-metal.metal
tests/test-backend-ops.cpp