]> git.djapps.eu Git - pkg/ggml/sources/whisper.cpp/commit
CUDA: faster non-contiguous concat (llama/10760)
authora3sh <redacted>
Thu, 12 Dec 2024 18:09:50 +0000 (02:09 +0800)
committerGeorgi Gerganov <redacted>
Wed, 18 Dec 2024 10:52:16 +0000 (12:52 +0200)
commitabe3102cb77891f88ca23946f5a19622f7f52306
tree3550dfc805ac7fd342fb1ec9835f001631de8d75
parent1193e494a9db2eec2028ddca47c71efb8e50c21f
CUDA: faster non-contiguous concat (llama/10760)

* faster uncontiguous concat

* Use a lambda to avoid code duplication

Co-authored-by: Diego Devesa <redacted>
* Update ggml/src/ggml-cuda/concat.cu

* add constexpr  and static assert

---------

Co-authored-by: Diego Devesa <redacted>
ggml/src/ggml-cuda/concat.cu