Back to Freedom.Tech Back
All llama.cpp releasesAll versions
Release Wed, Jul 22, 2026 1 min read

llama.cpp b10090

Original release notes

webgpu : add CONV_2D_DW (depthwise conv2d) kernel (#25847)

  • webgpu : add CONV_2D_DW (depthwise conv2d) kernel

Implement GGML_OP_CONV_2D_DW for the WebGPU backend, ported from the Vulkan backend's conv2d_dw.comp.

Assisted-by: Claude Opus-4.8

  • Remove unnecessary comments in webgpu support
  • update supported ops tables, triggered by adding webgpu CONV_2D_DW

Website: -

UI: