]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit
ggml : implement set_rows with i32 index (llama/16159)
authorSigbjørn Skjæret <redacted>
Mon, 22 Sep 2025 17:13:00 +0000 (19:13 +0200)
committerGeorgi Gerganov <redacted>
Thu, 25 Sep 2025 08:56:34 +0000 (11:56 +0300)
commitf24cca7c7deef390b700bed454d2284ad9068e84
treed3a6c22d72af843e77846c75b7e47e462a56d594
parent903b1f2f44b5dc7018fd6de3f3410586ed25c624
ggml : implement set_rows with i32 index (llama/16159)

* implement set_rows with i32 index

* template fix

* test quantized path

warnings--

* Apply suggestions from code review

Co-authored-by: Georgi Gerganov <redacted>
* forgotten name change

* deduplicate cuda/sycl and test-fix

* indent++

* vulkan: support set_rows with i32 index type (llama/16162)

* disable i32 index for webgpu for now

---------

Co-authored-by: Georgi Gerganov <redacted>
Co-authored-by: Jeff Bolz <redacted>
17 files changed:
src/ggml-cpu/ops.cpp
src/ggml-cuda/ggml-cuda.cu
src/ggml-cuda/set-rows.cu
src/ggml-metal/ggml-metal-device.cpp
src/ggml-metal/ggml-metal-device.h
src/ggml-metal/ggml-metal-ops.cpp
src/ggml-metal/ggml-metal.metal
src/ggml-opencl/ggml-opencl.cpp
src/ggml-opencl/kernels/set_rows.cl
src/ggml-sycl/ggml-sycl.cpp
src/ggml-sycl/set_rows.cpp
src/ggml-vulkan/ggml-vulkan.cpp
src/ggml-vulkan/vulkan-shaders/copy_to_quant.comp
src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp
src/ggml-webgpu/ggml-webgpu.cpp
src/ggml.c
tests/test-backend-ops.cpp