Public
  1. Public

onnxruntime

Public

Commits

Jiajia QinGitHub
e58319ebfc3
Jiajia Qin authored and GitHub committed fd6bab4250c
[js/webgpu] Provide a vectorized algorithm for GroupedConv (#18884)

### Description
This PR provides a vectorized algorithm for NHWC GroupedConv to improve
performance.

The aggregate time of GroupedConv in mobilenetv2-12 becomes ~1ms from
~4ms on Intel Alder Lake machine. About 20% improvement for the whole
model.