]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: c0b4509)
opencl: optimize mxfp4 kernels (#16037)
authorShawn Gu <redacted>
Thu, 18 Sep 2025 19:03:34 +0000 (12:03 -0700)
committerGitHub <redacted>
Thu, 18 Sep 2025 19:03:34 +0000 (12:03 -0700)
commit3edd87cd055a45d885fa914d879d36d33ecfc3e1
tree488fd1bcc3096d84f3486ad79fba3b1d614f0c3dtree
parentc0b45097c33e2667a94444f08cc9e36bec0a5e2ecommit | diff
opencl: optimize mxfp4 kernels (#16037)

- flatten mxfp4 and packed fp4->fp16 bit-wise convert function (replace lut)
- MoE kernel optimizations

---------

Co-authored-by: Li He <redacted>
ggml/src/ggml-opencl/CMakeLists.txt diff | blob | history
ggml/src/ggml-opencl/ggml-opencl.cpp diff | blob | history
ggml/src/ggml-opencl/kernels/cvt.cl diff | blob | history
ggml/src/ggml-opencl/kernels/mul_mv_id_mxfp4_f32_flat.cl [new file with mode: 0644] blob
ggml/src/ggml-opencl/kernels/mul_mv_mxfp4_f32_flat.cl [new file with mode: 0644] blob
Packaging of ggml-org/llama.cpp