ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-03 13:11:25 +00:00
Code Issues Packages Projects Releases Wiki Activity

fixed issue164 (#165)

Browse Source 
* fixed issue164

* removed prints
This commit is contained in:
Jianfeng Yan
2022-03-31 08:50:30 -05:00
committed by GitHub
parent f015c77687
commit ecf337bab5
1 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
include/ck/tensor_operation/gpu/device/device_batched_gemm_xdl.hpp
View File

@@ -129,7 +129,7 @@ struct DeviceBatchedGemmXdl
}
else if constexpr(is_same<tensor_layout::gemm::ColumnMajor, ALayout>::value)
{
return make_naive_tensor_descriptor(make_tuple(M, K), make_tuple(I1, M));
return make_naive_tensor_descriptor(make_tuple(M, K), make_tuple(I1, StrideA));
}
}();
@@ -158,7 +158,7 @@ struct DeviceBatchedGemmXdl
}
else if constexpr(is_same<tensor_layout::gemm::ColumnMajor, BLayout>::value)
{
return make_naive_tensor_descriptor(make_tuple(K, N), make_tuple(I1, K));
return make_naive_tensor_descriptor(make_tuple(K, N), make_tuple(I1, StrideB));
}
}();
@@ -183,7 +183,7 @@ struct DeviceBatchedGemmXdl
}
else if constexpr(is_same<tensor_layout::gemm::ColumnMajor, CLayout>::value)
{
return make_naive_tensor_descriptor(make_tuple(M, N), make_tuple(I1, M));
return make_naive_tensor_descriptor(make_tuple(M, N), make_tuple(I1, StrideC));
}
}();