[ET-VK][qconv] Fix depthwise weight_sums sum dimension#17573

Merged

SS-JIA merged 6 commits intomainfrom

gh/SS-JIA/432/orig

Feb 20, 2026

Commits on Feb 19, 2026

[ET-VK][qconv] Fix depthwise weight_sums sum dimension
ssjia
committed

Commits on Feb 20, 2026

[ET-VK][qconv] Pad weight_sums buffer to multiple-of-4 alignment

ssjia
authored and
SS-JIA
committed
[ET-VK][qconv] Add apply_relu support to q8ta conv operators

ssjia
authored and
SS-JIA
committed
[ET-VK] Add fused q8ta_relu unary operator for int8x4 tensors

ssjia
authored and
SS-JIA
committed
[ET-VK][ez] Always partition batch norm as it will be fused

ssjia
authored and
SS-JIA
committed
[ET-VK] Support different input layouts in q8ta_binary operator

ssjia
authored and
SS-JIA
committed