ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-01 12:11:19 +00:00
Code Issues Packages Projects Releases Wiki Activity

[CK TILE] Refactor Conv configs and Conv Elementwise (#3151)

Browse Source 
* [CK TILE] Refactor Conv configs and Conv Elementwise

* fix
This commit is contained in:
Bartłomiej Kocot
2025-11-04 15:04:53 +01:00
committed by GitHub
parent 99f38e4d9b
commit 8681ced962
14 changed files with 230 additions and 219 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
include/ck_tile/ops/grouped_convolution/kernel/grouped_convolution_forward_kernel.hpp
View File

@@ -19,7 +19,7 @@
namespace ck_tile {
/// @brief The Grouped Convolution kernel device arguments.
template <typename GroupedConvTraitsType_>
template <typename GroupedConvTraitsType_, typename CDElementwise_>
struct GroupedConvFwdKernelArgs
{
@@ -31,7 +31,7 @@ struct GroupedConvFwdKernelArgs
GroupedConvTraitsType_::VectorSizeC,
GroupedConvTraitsType_::NumGroupsToMerge,
true>; // Split N enabled
using CDElementwise = typename GroupedConvTraitsType_::CDElementwise;
using CDElementwise = CDElementwise_;
static constexpr index_t NumDTensor = GroupedConvTraitsType_::NumDTensor;
template <
@@ -469,7 +469,8 @@ struct GroupedConvolutionForwardKernel
using CDElementwise = typename EpiloguePipeline::CDElementwise;
using GroupedConvFwdKernelArgsSpecialized = GroupedConvFwdKernelArgs<GroupedConvTraitsType_>;
using GroupedConvFwdKernelArgsSpecialized =
GroupedConvFwdKernelArgs<GroupedConvTraitsType_, CDElementwise>;
static constexpr bool IsSplitKSupported = false;