]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 91bb39c)
clip : enable gpu backend (#4205)
authorSteward Garcia <redacted>
Fri, 29 Dec 2023 16:52:15 +0000 (11:52 -0500)
committerGitHub <redacted>
Fri, 29 Dec 2023 16:52:15 +0000 (18:52 +0200)
commitce18d727a47f2473ca863a6f78bf3ad480008f72
tree51708e4435c3bad6da1ad4800f908b02428e0dc7tree
parent91bb39cec7e4dfb9e2293509ef60298a67f0b1b7commit | diff
clip : enable gpu backend (#4205)

* clip: enable CUDA backend

* add missing kernels

* add enough padding for alignment

* remove ggml_repeat of clip.cpp

* add metal backend

* llava : fixes

- avoid ggml_repeat
- use GGML_USE_ instead of CLIP_USE_ macros
- remove unused vars

---------

Co-authored-by: Georgi Gerganov <redacted>
examples/llava/CMakeLists.txt diff | blob | history
examples/llava/clip.cpp diff | blob | history
Packaging of ggml-org/llama.cpp