]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: eb542d3)
ggml : allocate graphs in a context (#2392)
authorslaren <redacted>
Wed, 26 Jul 2023 13:56:53 +0000 (15:56 +0200)
committerGitHub <redacted>
Wed, 26 Jul 2023 13:56:53 +0000 (15:56 +0200)
commit5488fb789ea5692268309baa76f67598155060be
tree0609924b82a47ce97806b9201b75ec49bdf76c31tree
parenteb542d39324574a6778fad9ba9e34ba7a14a82a3commit | diff
ggml : allocate graphs in a context (#2392)

* ggml : graph allocation in contexts

* allocate work buffer as a ggml_object in ggml_graph_compute_with_ctx

* llama.cpp : allocate graph in the context

* add GGML_PAD

---------

Co-authored-by: Georgi Gerganov <redacted>
ggml.c diff | blob | history
ggml.h diff | blob | history
llama.cpp diff | blob | history
Packaging of ggml-org/llama.cpp