ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-26 16:04:58 +00:00
Code Issues Packages Projects Releases Wiki Activity

backward data (#7)

Browse Source 
* enabled atomic add in tensor copy
* added gridwise GEMM
* added backward data conv using GEMM + atomic
* added backward data conv using GEMM, no atomic


[ROCm/composable_kernel commit: 8f5f64960e]
This commit is contained in:
Chao Liu
2019-12-03 01:16:12 -06:00
committed by GitHub
parent 2cfc943628
commit 3799741fee
51 changed files with 3563 additions and 570 deletions
Download Patch File Download Diff File Expand all files Collapse all files

28
driver/include/device_tensor.hpp Normal file
View File

@@ -0,0 +1,28 @@
#pragma once
#include "tensor.hpp"
#include "common_header.hpp"
#include "ConstantTensorDescriptor_deprecated.hpp"
#include "tensor_descriptor.hpp"
template <typename ConstTensorDesc, std::size_t... Is>
auto make_TensorDescriptor_impl(ConstTensorDesc, std::integer_sequence<std::size_t, Is...>)
{
std::initializer_list<std::size_t> lengths = {ConstTensorDesc::GetLengths()[Is]...};
std::initializer_list<std::size_t> strides = {ConstTensorDesc::GetStrides()[Is]...};
return TensorDescriptor(lengths, strides);
}
template <typename ConstTensorDesc>
auto make_TensorDescriptor(ConstTensorDesc)
{
return make_TensorDescriptor_impl(
ConstTensorDesc{},
std::make_integer_sequence<std::size_t, ConstTensorDesc::GetNumOfDimension()>{});
}
template <typename ConstTensorDesc>
void ostream_ConstantTensorDescriptor(ConstTensorDesc, std::ostream& os = std::cout)
{
ostream_TensorDescriptor(make_TensorDescriptor(ConstTensorDesc{}), os);
}