ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-04-20 06:49:15 +00:00
Code Issues Packages Projects Releases Wiki Activity

[CK TILE] Refactor grouped conv fwd large tensor (#3144)

Browse Source
This commit is contained in:
Bartłomiej Kocot
2025-11-04 00:34:48 +01:00
committed by GitHub
parent c7ded76cc7
commit 99f38e4d9b
5 changed files with 161 additions and 167 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
include/ck_tile/ops/grouped_convolution/utils/grouped_convolution_utils.hpp
View File

@@ -63,7 +63,8 @@ template <index_t NDimSpatial_,
index_t VectorSizeB_ = 1,
index_t VectorSizeC_ = 1,
index_t NumGroupsToMerge_ = 1,
typename CDElementwise_ = PassThrough>
typename CDElementwise_ = PassThrough,
bool EnableSplitImage_ = false>
struct GroupedConvTraits
{
private:
@@ -74,6 +75,7 @@ struct GroupedConvTraits
}
public:
static constexpr bool EnableSplitImage = EnableSplitImage_;
static constexpr index_t NumGroupsToMerge = NumGroupsToMerge_;
static constexpr index_t NDimSpatial = NDimSpatial_;
static constexpr ConvolutionSpecialization ConvSpecialization = ConvSpecialization_;