Interface Activation

All Known Implementing Classes:

AbstractActivation, ELU, Exponential, GELU, HardSigmoid, Linear, ReLU, SELU, Sigmoid, Softmax, Softplus, Softsign, Swish, Tanh

Functional Interface:

This is a functional interface and can therefore be used as the assignment target for a lambda expression or method reference.

@FunctionalInterface
public interface Activation

Interface for Activations

Method Summary

Modifier and Type	Method	Description
<T extends TNumber> Operand<T>	call(Ops tf, Operand<T> input)	Gets the calculation operation for the activation.
static Activation	create(String name)	Creates an Activation instance based on the name as known to the TensorFlow engine.
static Activation	create(Map<String,Object> config)	Creates an Activation getInstance based on a configuration as produced by TensorFlow.

Method Details

create

static Activation create(String name)

Creates an Activation instance based on the name as known to the TensorFlow engine.

Parameters:

name - the activation name

Returns:

the Activation

Throws:

NullPointerException - if name is null

IllegalArgumentException - if the name is not a known ActivationType

create

static Activation create(Map<String,Object> config)

Creates an Activation getInstance based on a configuration as produced by TensorFlow.

Parameters:

config - a Map object containing the Activation's state. This Map object must contain at least a name key.

"name" : String - this is the TensorFlow Engine's Activation name

}

Returns:

the Activation

Throws:

NullPointerException - if config is null, or the activation name is missing from the Map.

IllegalArgumentException - if the name contained in the config map is not a known ActivationType

call

<T extends TNumber> Operand<T> call(Ops tf,
 Operand<T> input)

Gets the calculation operation for the activation.

Type Parameters:

T - the data type of the input and the result

Parameters:

tf - the TensorFlow Ops

input - the input tensor

Returns:

The operand for the activation