Have a question? Connect with the community at the TensorFlow Forum Visit Forum



Activation <T extends TNumber > Abstract base class for Activations

Note: The ERROR(/#tf) attribute must be set prior to invoking the call method.

ELU <T extends TFloating > Exponential linear unit.
Exponential <T extends TFloating > Exponential activation function.
HardSigmoid <T extends TFloating > Hard sigmoid activation.
Linear <U extends TNumber > Linear activation function (pass-through).
ReLU <T extends TNumber > Rectified Linear Unit(ReLU) activation.
SELU <T extends TFloating > Scaled Exponential Linear Unit (SELU).
Sigmoid <T extends TFloating > Sigmoid activation.
Softmax <T extends TFloating > Softmax converts a real vector to a vector of categorical probabilities.
Softplus <T extends TFloating > Softplus activation function, softplus(x) = log(exp(x) + 1) .
Softsign <T extends TFloating > Softsign activation function, softsign(x) = x / (abs(x) + 1) .
Swish <T extends TFloating > Swish activation function.
Tanh <T extends TFloating > Hyperbolic tangent activation function.