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

Add grouped conv fwd bias relu instances (#2179)

Browse Source 
* Add grouped conv fwd bias relu instances

* fixes

* fix
This commit is contained in:
Bartłomiej Kocot
2025-05-09 22:52:34 +02:00
committed by GitHub
parent 6b1a339b6f
commit 6fddb5708c
33 changed files with 2477 additions and 550 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
include/ck/tensor_operation/gpu/element/binary_element_wise_operation.hpp
View File

@@ -404,6 +404,14 @@ struct AddRelu
y = a > type_convert<bhalf_t>(0.0f) ? a : type_convert<bhalf_t>(0.0f);
};
template <>
__host__ __device__ constexpr void
operator()<bhalf_t, bhalf_t, bhalf_t>(bhalf_t& y, const bhalf_t& x0, const bhalf_t& x1) const
{
const float a = type_convert<float>(x0) + type_convert<float>(x1);
y = a > type_convert<bhalf_t>(0.0f) ? a : type_convert<bhalf_t>(0.0f);
};
template <>
__host__ __device__ constexpr void
operator()<int, int, int8_t>(int& y, const int& x0, const int8_t& x1) const