ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-13 01:36:06 +00:00
Code Issues Packages Projects Releases Wiki Activity

Post-merge cleanup for WMMA grouped conv fwd (#3468)

Browse Source 
* remove duplicate aliases

* Split scaleadd_ab instances for WMMA grouped conv fwd

* removed big shape from the test
This commit is contained in:
Wojciech Laskowski
2025-12-22 15:57:45 +01:00
committed by GitHub
parent 44f1b5c5de
commit a8aebb7a8e
13 changed files with 570 additions and 62 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
profiler/src/profile_grouped_conv_fwd_bilinear.cpp
View File

@@ -108,10 +108,6 @@ int grouped_conv_fwd_bilinear(int argc, char* argv[])
using NDHWGC = ck::tensor_layout::convolution::NDHWGC;
using NDHWGK = ck::tensor_layout::convolution::NDHWGK;
using GKZYXC = ck::tensor_layout::convolution::GKZYXC;
using NDHWGC = ck::tensor_layout::convolution::NDHWGC;
using NDHWGK = ck::tensor_layout::convolution::NDHWGK;
constexpr auto I3 = ck::Number<3>{};
auto profile = [&](auto num_dim_spatial_tmp,