]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 00681df)
CANN: Disable acl_graph for prefill stage (#15933)
authorhipudding <redacted>
Thu, 11 Sep 2025 07:59:37 +0000 (15:59 +0800)
committerGitHub <redacted>
Thu, 11 Sep 2025 07:59:37 +0000 (15:59 +0800)
commitc0389dba43d50695f9d3f57dd1f1a14cbefc100c
treea3d52f16c517cbbdd883c1fc54397211824e99f8tree
parent00681dfc16ba4cebb9c7fbd2cf2656e06a0692a4commit | diff
CANN: Disable acl_graph for prefill stage (#15933)

Since the prefill length is not fixed, graphs constructed for the
prefill stage cannot be reused. For this reason, ACL graph
execution is disabled by default during prefill.
docs/backend/CANN.md diff | blob | history
ggml/src/ggml-cann/ggml-cann.cpp diff | blob | history
Packaging of ggml-org/llama.cpp