]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit
ggml webgpu: minor set rows optimization (llama/16810)
authorReese Levine <redacted>
Sun, 9 Nov 2025 12:44:39 +0000 (14:44 +0200)
committerGeorgi Gerganov <redacted>
Sun, 9 Nov 2025 16:30:22 +0000 (18:30 +0200)
commit977f81b779452584600642f431d19544e6eea264
treeea27c3ac47c3bbb323314947e93e2a56b607ca9a
parent9abaab9204f80ef7f2b04aa6f01c4b6d7f213a95
ggml webgpu: minor set rows optimization (llama/16810)

* Add buffer label and enable dawn-specific toggles to turn off some checks

* Minor set_rows optimization (#4)

* updated optimization, fixed errors

* non vectorized version now dispatches one thread per element

* Simplify

* Change logic for set_rows pipelines

---------

Co-authored-by: Neha Abbas <redacted>
Co-authored-by: Neha Abbas <redacted>
Co-authored-by: Reese Levine <redacted>
* Comment on dawn toggles

* Remove some comments

* Implement overlap binary operators

* Revert "Implement overlap binary operators"

This reverts commit ed710b36f51ab3f53fa13db15c1685dc8678a32a.

* Disable support for non-contiguous binary_op tensors and leave note for future support

---------

Co-authored-by: neha-ha <redacted>
Co-authored-by: Neha Abbas <redacted>
Co-authored-by: Neha Abbas <redacted>
src/ggml-webgpu/ggml-webgpu.cpp
src/ggml-webgpu/wgsl-shaders/set_rows.tmpl.wgsl [new file with mode: 0644]