【KernelGen】Add cudnn_convolution operator - #1729

Merged

w1120029931-bit merged 13 commits into

flagos-ai:masterfrom

Schopenhauer-loves-Hegel:auto-gen/cudnn_convolution

May 14, 2026

Commits on May 12, 2026

Add cudnn_convolution operator implementation, tests and benchmark

factnn
and
claude
committed
b116740
Apply suggestions from code review

factnn
and
tengqm
committed
524a49a
fix: apply black formatting to test_cudnn_convolution.py

factnn
and
claude
committed
e49f496
fix: remove to_reference for cudnn_convolution tests (CUDA-only op)

factnn
and
claude
committed
4b8b812
fix: rename benchmark file to remove _perf suffix

factnn
and
claude
committed
83297d5
fix: move ref tensor to CPU before gems_assert_close

factnn
and
claude
committed
56cb917
fix: move both res and ref to CPU before comparison

factnn
and
claude
committed
8924dc2
fix: move cudnn_convolution to correct alphabetical position in operators.yaml

factnn
and
claude
committed
cd0e5d9

Commits on May 13, 2026

Merge branch 'master' into auto-gen/cudnn_convolution
factnn
authored
4d55a90
Merge branch 'master' into auto-gen/cudnn_convolution
factnn
authored
edce041
Merge branch 'master' into auto-gen/cudnn_convolution
factnn
authored
58ea712

Commits on May 14, 2026