]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: b08f22c)
CUDA: mul_mat_vec_q max. batch size 8 -> 4 (#5370)
authorJohannes Gäßler <redacted>
Tue, 6 Feb 2024 17:43:06 +0000 (18:43 +0100)
committerGitHub <redacted>
Tue, 6 Feb 2024 17:43:06 +0000 (19:43 +0200)
commit17c97fb0620448b37516a3f53fea6c482b0a30a4
tree20e2fc07a74c88cce62a626808f0a1ae0e54be8btree
parentb08f22c882a1443e6b97081f3ce718a4d1a741f8commit | diff
CUDA: mul_mat_vec_q max. batch size 8 -> 4 (#5370)
ggml-cuda.cu diff | blob | history
Packaging of ggml-org/llama.cpp