]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
opencl: add q8_0 mm support (#16469)
authorlhez <redacted>
Wed, 15 Oct 2025 17:51:04 +0000 (10:51 -0700)
committerGitHub <redacted>
Wed, 15 Oct 2025 17:51:04 +0000 (10:51 -0700)
commit0cb7a0683b0529172472d74d21f05470a607f297
tree178c80c295b5254cb91a9b5fadd38c276e0885d2
parentd93f8439b08c4f35e13a41a7366901fdbe770fc8
opencl: add q8_0 mm support (#16469)

* opencl: add mm_q8_0_f32

* opencl: fix data loading for incomplete tile

* opencl: use q8_0 mm for larger matrix

* opencl: add some tests to cover the path
ggml/src/ggml-opencl/CMakeLists.txt
ggml/src/ggml-opencl/ggml-opencl.cpp
ggml/src/ggml-opencl/kernels/mul_mm_f16_f32_l4_lm.cl
ggml/src/ggml-opencl/kernels/mul_mm_f32_f32_l4_lm.cl
ggml/src/ggml-opencl/kernels/mul_mm_q8_0_f32_l4_lm.cl [new file with mode: 0644]
tests/test-backend-ops.cpp