Batch Processing

Layer Input

Backward ELU 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

that stores the input gradient computed on the preceding layer. This input can be an object of any class derived fromTensor.

inputFromForward

Collection of data needed for the backward ELU layer.

Element ID

Element

auxData

Pointer to the tensor of size

that stores the input data for the forward ELU layer. This input can be an object of any class derived from Tensor.

auxIntermediateValue

Pointer to the tensor of size

that stores an intermediate value of the forward ELU layer. The tensor can be an object of any class derived from Tensor. This parameter is optional, if you provide it, the layer will use values from that tensor to compute the result of the backward operation, otherwise the layer will use only auxData. Note that parameter auxData must be provided in both cases.

Layer Parameters

Backward ELU layer has the following parameters:

Parameter

Default Value

Description

algorithmFPType

float

The floating-point type used for intermediate computations within algorithm. Can be either float or double.

method

defaultDense

The default performance-oriented computation method, the only method supported by the algorithm.

alpha

1.0

Parameter α of the function.

Layer Output

Backward ELU 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

gradient

Pointer to the tensor of

stores the result of the backward ELU layer. This input can be an object of any class derived from Tensor.

Examples

C++: elu_layer_dense_batch.cpp

Java*: ELULayerDenseBatch.java

Python*: elu_layer_dense_batch.py

For more complete information about compiler optimizations, see our Optimization Notice.
Select sticky button color: 
Orange (only for download buttons)