]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: dd4623a)
ggml-webgpu: improve i-quants mul_mat performance and speed up prefill (#24530)
authorMasashi Yoshimura <redacted>
Mon, 15 Jun 2026 01:15:30 +0000 (10:15 +0900)
committerGitHub <redacted>
Mon, 15 Jun 2026 01:15:30 +0000 (18:15 -0700)
commit6e9007ae61f4e994c27484759caac6ef2aa32b30
tree13386e7d5a0cd5014079b3dd9275f473de4aaba3tree
parentdd4623a74f0c85e6b1dd9ee99a92b9c67cac3708commit | diff
ggml-webgpu: improve i-quants mul_mat performance and speed up prefill (#24530)

* Improve prefill speeds for i-quants

* Fix #if defined() usage in preprocessor guards.
ggml/src/ggml-webgpu/wgsl-shaders/mul_mat_decls.tmpl diff | blob | history
Packaging of ggml-org/llama.cpp