]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 5a4126a)
CUDA: add fast walsh-hadamard transform (#23615)
authorAman Gupta <redacted>
Mon, 25 May 2026 13:12:10 +0000 (21:12 +0800)
committerGitHub <redacted>
Mon, 25 May 2026 13:12:10 +0000 (21:12 +0800)
commitc1f1e28d29948892b6aec04b53053a3fd1dc4df3
tree19778caf3c9a5a58dc922d83867f6c902e23eddctree
parent5a4126adc1b99d17162caf845afca98c2cf72b36commit | diff
CUDA: add fast walsh-hadamard transform (#23615)

* CUDA: add fast walsh-hadamard transform

* review: add unrolls + change size_t -> int

* warp size 64

---------

Co-authored-by: Johannes Gäßler <redacted>
ggml/src/ggml-cuda/fwht.cu [new file with mode: 0644] blob
ggml/src/ggml-cuda/fwht.cuh [new file with mode: 0644] blob
ggml/src/ggml-cuda/ggml-cuda.cu diff | blob | history
tests/test-backend-ops.cpp diff | blob | history
Packaging of ggml-org/llama.cpp