]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: fa84c4b)
metal : fix failure to load model (#1817)
authorKawrakow <redacted>
Mon, 12 Jun 2023 11:31:36 +0000 (14:31 +0300)
committerGitHub <redacted>
Mon, 12 Jun 2023 11:31:36 +0000 (14:31 +0300)
commit8c0a10e64dbf60fd9946c0cd5e6f59690800b123
treefb69c71663c6c987160caf1f17f2bf9c254fe01etree
parentfa84c4b3e80199a5683438f062009c031a06c4facommit | diff
metal : fix failure to load model (#1817)

The number of buffers in the ggml context was left unitialized.
This leads to sporadic failures to load the model on
startup. It is actually strange that the failure occurred so
infrequantly.

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