Apache SINGA
A distributed deep learning platform .
|
reduction to 1 dimension tensor input: Tensor<Device,k>: ishape output: Tensor<Device,1> shape[0] = ishape[dimkeep]; More...
#include <tensor_expr_ext.h>
Public Member Functions | |
ReduceTo1DExp (const EType &src, real_t scale) | |
construct a repmat expression from src and nrow | |
Public Member Functions inherited from mshadow::expr::Exp< ReduceTo1DExp< EType, Reducer, dimkeep >, type::kComplex > | |
const ReduceTo1DExp< EType, Reducer, dimkeep > & | self (void) const |
ReduceTo1DExp< EType, Reducer, dimkeep > & | refself (void) |
Public Attributes | |
const EType & | src_ |
source operand | |
real_t | scale_ |
source operand, scale of the | |
reduction to 1 dimension tensor input: Tensor<Device,k>: ishape output: Tensor<Device,1> shape[0] = ishape[dimkeep];
EType | type of expression to be reduced |
Reducer | which reducer to use |
srcdim | dimension of source |
dimkeep | which dimension to be kept, |