]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 44a1a4a)
imatrix : offload to GPU support (#4957)
authorGeorgi Gerganov <redacted>
Wed, 17 Jan 2024 16:46:30 +0000 (18:46 +0200)
committerGitHub <redacted>
Wed, 17 Jan 2024 16:46:30 +0000 (18:46 +0200)
commitba69bbc84ced580fe4fdb0713ca2d95634325b7a
tree75628460a0036db7ca1105bf3093f76c71ab3eedtree
parent44a1a4a41a4c0b03afaa7d9e06bcbc7cf95aa1e6commit | diff
imatrix : offload to GPU support (#4957)

* backend : add eval callback

ggml-ci

* backend : group nodes in a single compute when user don't need them

* backend : clean-up the implementation

ggml-ci

* simple : do not perform tensor data copy if not needed

* simple : fix

* imatrix : offload to GPU support

* imatrix : fix ggml_mul_mat_id hanlding

ggml-ci

* ci : add imatrix test

ggml-ci

* ci : rearrange output

ggml-ci
ci/run.sh diff | blob | history
examples/imatrix/imatrix.cpp diff | blob | history
ggml.c diff | blob | history
ggml.h diff | blob | history
Packaging of ggml-org/llama.cpp