GetReduceCountPerThreadForBlockwiseWelford< K_BlockTileSize, KThreadSliceSize > Struct Template Reference

• ck
• tensor_operation
• device
• GetReduceCountPerThreadForBlockwiseWelford

Public Member Functions | Public Attributes | List of all members

ck::tensor_operation::device::GetReduceCountPerThreadForBlockwiseWelford< K_BlockTileSize, KThreadSliceSize > Struct Template Reference

#include <welford_helper.hpp>

Public Member Functions
	GetReduceCountPerThreadForBlockwiseWelford (index_t numBlockTileIteration, long_index_t reduce_length)
__device__ index_t	operator() (index_t thread_k_cluster_id) const

Public Attributes
index_t	numBlockTileIteration_
index_t	count_in_last_tile_

Constructor & Destructor Documentation

◆ GetReduceCountPerThreadForBlockwiseWelford()

template<index_t K_BlockTileSize, index_t KThreadSliceSize>

ck::tensor_operation::device::GetReduceCountPerThreadForBlockwiseWelford< K_BlockTileSize, KThreadSliceSize >::GetReduceCountPerThreadForBlockwiseWelford ( index_t numBlockTileIteration, long_index_t reduce_length )

inline

Member Function Documentation

◆ operator()()

template<index_t K_BlockTileSize, index_t KThreadSliceSize>

__device__ index_t ck::tensor_operation::device::GetReduceCountPerThreadForBlockwiseWelford< K_BlockTileSize, KThreadSliceSize >::operator() ( index_t thread_k_cluster_id ) const

inline

Member Data Documentation

◆ count_in_last_tile_

template<index_t K_BlockTileSize, index_t KThreadSliceSize>

index_t ck::tensor_operation::device::GetReduceCountPerThreadForBlockwiseWelford< K_BlockTileSize, KThreadSliceSize >::count_in_last_tile_

◆ numBlockTileIteration_

template<index_t K_BlockTileSize, index_t KThreadSliceSize>

index_t ck::tensor_operation::device::GetReduceCountPerThreadForBlockwiseWelford< K_BlockTileSize, KThreadSliceSize >::numBlockTileIteration_

---

The documentation for this struct was generated from the following file:

• welford_helper.hpp