]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 9008027)
bench : cache the llama_context state at computed depth (#16944)
authorGeorgi Gerganov <redacted>
Fri, 7 Nov 2025 19:23:11 +0000 (21:23 +0200)
committerGitHub <redacted>
Fri, 7 Nov 2025 19:23:11 +0000 (21:23 +0200)
commit7956bb4d7f430f23f3c4726f7e6404b89f6e20a4
tree3276efe4aa3b516951f9f9b0c7580fd12eedc2eatree
parent9008027aa376526819415469f74fb9281136224ecommit | diff
bench : cache the llama_context state at computed depth (#16944)

* bench : cache llama_context state at depth

* cont : handle failures to restore the old state

* cont : print information when the state is being reused
tools/llama-bench/llama-bench.cpp diff | blob | history
Packaging of ggml-org/llama.cpp