/home/docs/checkouts/readthedocs.org/user_builds/advanced-micro-devices-composable-kernel/checkouts/docs-6.4.3/include/ck/tensor_operation/gpu/device/impl/device_normalization_bwd_data_impl.hpp File Reference

• docs-6.4.3
• include
• ck
• tensor_operation
• gpu
• device
• impl

Classes | Namespaces | Functions

device_normalization_bwd_data_impl.hpp File Reference

#include <iostream>
#include <vector>
#include "ck/tensor_operation/gpu/device/device_normalization_bwd_data.hpp"
#include "ck/tensor_operation/gpu/grid/normalization/gridwise_normalization_bwd_data.hpp"
#include "ck/tensor_description/tensor_descriptor.hpp"
#include "ck/tensor_description/tensor_descriptor_helper.hpp"
#include "ck/tensor_operation/gpu/device/impl/device_reduce_common.hpp"
#include "ck/host_utility/device_prop.hpp"
#include "ck/host_utility/kernel_launch.hpp"

Go to the source code of this file.

Classes
struct	ck::tensor_operation::device::DeviceNormalizationBwdDataImpl< DYDataType, XDataType, GammaDataType, MeanInvStdDataType, ComputeDataType, DXDataType, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, IsDYFastestDimReduced, DYSrcVectorSize, IsXFastestDimReduced, XSrcVectorSize, IsGammaFastestDimReduced, GammaSrcVectorSize, IsMeanInvStdFastestDimReduced, MeanInvStdSrcVectorSize, IsDxFastestDimReduced, DXDstVectorSize >
struct	ck::tensor_operation::device::DeviceNormalizationBwdDataImpl< DYDataType, XDataType, GammaDataType, MeanInvStdDataType, ComputeDataType, DXDataType, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, IsDYFastestDimReduced, DYSrcVectorSize, IsXFastestDimReduced, XSrcVectorSize, IsGammaFastestDimReduced, GammaSrcVectorSize, IsMeanInvStdFastestDimReduced, MeanInvStdSrcVectorSize, IsDxFastestDimReduced, DXDstVectorSize >::Argument
struct	ck::tensor_operation::device::DeviceNormalizationBwdDataImpl< DYDataType, XDataType, GammaDataType, MeanInvStdDataType, ComputeDataType, DXDataType, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, IsDYFastestDimReduced, DYSrcVectorSize, IsXFastestDimReduced, XSrcVectorSize, IsGammaFastestDimReduced, GammaSrcVectorSize, IsMeanInvStdFastestDimReduced, MeanInvStdSrcVectorSize, IsDxFastestDimReduced, DXDstVectorSize >::Invoker

Namespaces
	ck
	ck::tensor_operation
	ck::tensor_operation::device

Functions
template<typename GridwiseNormalizationBwd , typename DYDataType , typename XDataType , typename GammaDataType , typename MeanInvStdDataType , typename DXDataType , typename GridDesc_M_K >
__global__ void	ck::tensor_operation::device::kernel_normalization_bwd_data (const GridDesc_M_K dy_grid_desc_m_k, const GridDesc_M_K x_grid_desc_m_k, const GridDesc_M_K gamma_grid_desc_m_k, const GridDesc_M_K mean_grid_desc_m_k, const GridDesc_M_K inv_std_grid_desc_m_k, const GridDesc_M_K dx_grid_desc_m_k, index_t num_k_block_tile_iteration, const DYDataType *const __restrict__ p_dy_global, const XDataType *const __restrict__ p_x_global, const GammaDataType *const __restrict__ p_gamma_global, const MeanInvStdDataType *const __restrict__ p_mean_global, const MeanInvStdDataType *const __restrict__ p_inv_std_global, DXDataType *const __restrict__ p_dx_global)