Developer Guide for Intel® Data Analytics Acceleration Library 2018 Update 2

Batch Processing

Layer Input

The backward element-wise sum layer accepts the input described below. Pass the Input ID as a parameter to the methods that provide input for your algorithm. For more details, see Algorithms.

Input ID

Input

inputGradient

Pointer to the tensor of size n 1 x ... х n p that stores the input gradient computed on the preceding layer. This input can be an object of any class derived from Tensor.

inputFromForward

Element ID

Element

auxCoefficients

Pointer to the tensor of size K that stores the coefficients . This input can be an object of any class derived from Tensor. If you pass a null pointer, you must specify the input auxNumberOfCoefficients, and the algorithm sets the coefficients to 1.

auxNumberOfCoefficients

If the input auxNumberOfCoefficients is a null pointer, this input must be a numeric table of size 1 × 1 that contains the number of coefficients K; otherwise the algorithm ignores the input.

Layer Parameters

The backward element-wise sum layer has the following parameters:

Parameter

Default Value

Description

algorithmFPType

float

The floating-point type that the algorithm uses for intermediate computations. Can be float or double.

method

defaultDense

Performance-oriented computation method, the only method supported by the layer.

Layer Output

The backward element-wise sum layer calculates the result described below. Pass the Result ID as a parameter to the methods that access the results of your algorithm. For more details, see Algorithms.

Result ID

Result

resultLayerData

The collection K of tensors Z (i) of size n 1 x n p that stores the gradients computed on the backward element-wise sum layer. This result can be an object of any class derived from Tensor.

Examples

C++: eltwise_sum_layer_dense_batch.cpp

Java*: EltwiseSumLayerDenseBatch.java

Python*: eltwise_sum_layer_dense_batch.py

See Also