]> git.djapps.eu Git - pkg/ggml/sources/whisper.cpp/commit
CUDA: add conv_2d_dw (llama/14265)
authorAman Gupta <redacted>
Fri, 20 Jun 2025 01:50:24 +0000 (09:50 +0800)
committerGeorgi Gerganov <redacted>
Sat, 21 Jun 2025 04:34:17 +0000 (07:34 +0300)
commit5efd43c956cf5abe371c6f0f12b6b1a2447818b8
tree322436cee3d8953578db33519df1291fd221dbb7
parent71adde9203b77251e6b2063bbbff35e19ff5025d
CUDA: add conv_2d_dw (llama/14265)

* CUDA: add conv_2d_dw

* better naming

* simplify using template

* Review: fix operation ordering in ggml-cuda, use __forceinline__, use more const
ggml/src/ggml-cuda/conv2d-dw.cu [new file with mode: 0644]
ggml/src/ggml-cuda/conv2d-dw.cuh [new file with mode: 0644]
ggml/src/ggml-cuda/ggml-cuda.cu