]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 636fc17)
Make graph_max_nodes vary by ubatch size (#17794)
authorPiotr Wilkin (ilintar) <redacted>
Mon, 8 Dec 2025 13:32:41 +0000 (14:32 +0100)
committerGitHub <redacted>
Mon, 8 Dec 2025 13:32:41 +0000 (14:32 +0100)
commite4e9c4329c088d3aa97b8c242e18ff79bfe66248
tree1929709d09c0d8f89def8a5ba0209de309cf5b71tree
parent636fc17a376dacc01da20d508e6986a299b1f819commit | diff
Make graph_max_nodes vary by ubatch size (#17794)

* Make graph_max_nodes vary by ubatch size for models where chunking might explode the graph

* Update src/llama-context.h

Co-authored-by: Georgi Gerganov <redacted>
* Add missing const

---------

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