]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit
metal : fuse NORM + MUL + ADD, support non-multiples of 4 (llama/16220)
authorGeorgi Gerganov <redacted>
Thu, 25 Sep 2025 08:30:16 +0000 (11:30 +0300)
committerGeorgi Gerganov <redacted>
Thu, 25 Sep 2025 08:56:34 +0000 (11:56 +0300)
commit9adc411ca81d8b3f7ed7cc6f0656ef6d545b318c
tree45db376c58d7c7beef6be4373b9ae760599d1812
parent58a60fbb970dda343da1a18c7ba7987993f35567
metal : fuse NORM + MUL + ADD, support non-multiples of 4 (llama/16220)

* metal : fuse NORM + MUL + ADD

* metal : support norms of non-multiple of 4

* cont : fix comment [no ci]
src/ggml-metal/ggml-metal-common.cpp
src/ggml-metal/ggml-metal-device.cpp
src/ggml-metal/ggml-metal-device.h
src/ggml-metal/ggml-metal-device.m
src/ggml-metal/ggml-metal-impl.h
src/ggml-metal/ggml-metal-ops.cpp
src/ggml-metal/ggml-metal-ops.h
src/ggml-metal/ggml-metal.metal
tests/test-backend-ops.cpp