]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 90cc59d)
llama : prevent usage of k-quants when tensor size is not a multiple of 256 (#1921)
authorKawrakow <redacted>
Sun, 18 Jun 2023 08:13:43 +0000 (11:13 +0300)
committerGitHub <redacted>
Sun, 18 Jun 2023 08:13:43 +0000 (11:13 +0300)
commit8ab8ba62eb27cc340be2edf3418e051b1d967416
tree8ea76c75824b2e15cf42afd8040f20b885826bcdtree
parent90cc59d6ab1363a5c69c60c4b94db647d3a54a18commit | diff
llama : prevent usage of k-quants when tensor size is not a multiple of 256 (#1921)

* Fix examples/metal

* k-quants: prevent usage when tensor size is not divisible by 256

---------

Co-authored-by: Iwan Kawrakow <redacted>
llama.cpp diff | blob | history
Packaging of ggml-org/llama.cpp