ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-18 20:09:25 +00:00
Code Issues Packages Projects Releases Wiki Activity

[CK][CONV] Support NCHW in class DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1 (#2459)

Browse Source 
1. Port NCHW support from ConvFwd (#2375) to conv bwd data
2. Add new instance device_grouped_conv_bwd_data_xdl_f16_nchw_instances for nchw

Co-authored-by: azhuang <anzhong.huang@amd.com>

[ROCm/composable_kernel commit: fbd9f32abe]
This commit is contained in:
linqunAMD
2025-07-17 08:19:57 +08:00
committed by GitHub
parent f77d704980
commit 8319789fe1
6 changed files with 509 additions and 156 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
library/src/tensor_operation_instance/gpu/grouped_conv2d_bwd_data/xdl/device_grouped_conv2d_bwd_data_xdl_ngchw_gkcyx_ngkhw_f16_instance.cpp
View File

@@ -32,6 +32,14 @@ void add_device_grouped_conv2d_bwd_data_xdl_ngkhw_gkcyx_ngchw_f16_instances(
Empty_Tuple,
NGCHW,
ConvBwdDataDefault>{});
add_device_operation_instances(
instances,
device_grouped_conv_bwd_data_xdl_f16_nchw_instances<2,
NGKHW,
GKCYX,
Empty_Tuple,
NGCHW,
ConvBwdDataFilter1x1Stride1Pad0>{});
}
} // namespace instance

8
library/src/tensor_operation_instance/gpu/grouped_conv3d_bwd_data/xdl/device_grouped_conv3d_bwd_data_xdl_ngcdhw_gkczyx_ngkdhw_f16_instance.cpp
View File

@@ -32,6 +32,14 @@ void add_device_grouped_conv3d_bwd_data_xdl_ngkdhw_gkczyx_ngcdhw_f16_instances(
Empty_Tuple,
NGCDHW,
ConvBwdDataDefault>{});
add_device_operation_instances(
instances,
device_grouped_conv_bwd_data_xdl_f16_nchw_instances<3,
NGKDHW,
GKCZYX,
Empty_Tuple,
NGCDHW,
ConvBwdDataFilter1x1Stride1Pad0>{});
}
} // namespace instance