]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 72bd732)
opencl: add kernel to handle mat mul in attention to improve encoding speed (#17181)
authorshaofeiqi <redacted>
Sun, 16 Nov 2025 01:33:10 +0000 (17:33 -0800)
committerGitHub <redacted>
Sun, 16 Nov 2025 01:33:10 +0000 (17:33 -0800)
commit4db56412103502347e000142ffc01b1eca217e27
tree60f6f95a3487b7ee900cc58e01e28e909cbcb04ftree
parent72bd7321a7d7465d371eb2ae46cd5518842c8f44commit | diff
opencl: add kernel to handle mat mul in attention to improve encoding speed (#17181)

* Add mul_mm_f16_f32_kq_kqv kernel

* Add ggml_cl_mul_mat_kq_kqv_adreno func

* fix whitespace

* remove unused variable

* remove redundant

* refactor and clean up

* remove trailing whitespace
ggml/src/ggml-opencl/CMakeLists.txt diff | blob | history
ggml/src/ggml-opencl/ggml-opencl.cpp diff | blob | history
ggml/src/ggml-opencl/kernels/mul_mm_f16_f32_kq_kqv.cl [new file with mode: 0644] blob
Packaging of ggml-org/llama.cpp