#include <threadwise_welford.hpp>
|
constexpr __device__ | ThreadwiseWelford () |
|
__device__ void | Update (T &mean, T &var, T x) |
|
template<typename XBufferType , typename MeanBufferType , typename VarBufferType > |
__device__ void | Run (const XBufferType &x_buf_m_k, MeanBufferType &mean_buf_m, VarBufferType &var_buf_m) |
|
◆ ThreadwiseWelford()
template<typename T , typename XThreadDesc_M_K , typename MeanVarThreadDesc_M >
◆ Run()
template<typename T , typename XThreadDesc_M_K , typename MeanVarThreadDesc_M >
template<typename XBufferType , typename MeanBufferType , typename VarBufferType >
__device__ void ck::ThreadwiseWelford< T, XThreadDesc_M_K, MeanVarThreadDesc_M >::Run |
( |
const XBufferType & |
x_buf_m_k, |
|
|
MeanBufferType & |
mean_buf_m, |
|
|
VarBufferType & |
var_buf_m |
|
) |
| |
|
inline |
◆ Update()
template<typename T , typename XThreadDesc_M_K , typename MeanVarThreadDesc_M >
__device__ void ck::ThreadwiseWelford< T, XThreadDesc_M_K, MeanVarThreadDesc_M >::Update |
( |
T & |
mean, |
|
|
T & |
var, |
|
|
T |
x |
|
) |
| |
|
inline |
◆ cur_count_
template<typename T , typename XThreadDesc_M_K , typename MeanVarThreadDesc_M >
◆ max_count_
template<typename T , typename XThreadDesc_M_K , typename MeanVarThreadDesc_M >
◆ mean_var_thread_desc_m
template<typename T , typename XThreadDesc_M_K , typename MeanVarThreadDesc_M >
constexpr auto ck::ThreadwiseWelford< T, XThreadDesc_M_K, MeanVarThreadDesc_M >::mean_var_thread_desc_m = MeanVarThreadDesc_M{} |
|
staticconstexpr |
◆ thread_mean_var_length_m
template<typename T , typename XThreadDesc_M_K , typename MeanVarThreadDesc_M >
constexpr auto ck::ThreadwiseWelford< T, XThreadDesc_M_K, MeanVarThreadDesc_M >::thread_mean_var_length_m = mean_var_thread_desc_m.GetLength(Number<0>{}) |
|
staticconstexpr |
◆ thread_x_length_k
template<typename T , typename XThreadDesc_M_K , typename MeanVarThreadDesc_M >
constexpr auto ck::ThreadwiseWelford< T, XThreadDesc_M_K, MeanVarThreadDesc_M >::thread_x_length_k = x_thread_desc_m_k.GetLength(Number<1>{}) |
|
staticconstexpr |
◆ thread_x_length_m
template<typename T , typename XThreadDesc_M_K , typename MeanVarThreadDesc_M >
constexpr auto ck::ThreadwiseWelford< T, XThreadDesc_M_K, MeanVarThreadDesc_M >::thread_x_length_m = x_thread_desc_m_k.GetLength(Number<0>{}) |
|
staticconstexpr |
◆ x_thread_desc_m_k
template<typename T , typename XThreadDesc_M_K , typename MeanVarThreadDesc_M >
constexpr auto ck::ThreadwiseWelford< T, XThreadDesc_M_K, MeanVarThreadDesc_M >::x_thread_desc_m_k = XThreadDesc_M_K{} |
|
staticconstexpr |
The documentation for this struct was generated from the following file:
- /home/docs/checkouts/readthedocs.org/user_builds/advanced-micro-devices-composable-kernel/checkouts/docs-6.4.3/include/ck/tensor_operation/gpu/thread/threadwise_welford.hpp