org.tensorflow.framework.initializers

Interfaces

Initializer<T extends TType> An interface for Initializers 

Classes

BaseInitializer<T extends TType> Abstract base class for all Initializers  
Constant<T extends TType> Initializer that generates tensors with a constant value. 
Glorot<T extends TFloating> The Glorot initializer, also called Xavier initializer. 
He<T extends TFloating> He initializer. 
Identity<T extends TFloating> Initializer that generates the identity matrix. 
LeCun<T extends TFloating> LeCun normal initializer. 
Ones<T extends TType> Initializer that generates tensors initialized to 1. 
Orthogonal<T extends TFloating> Initializer that generates an orthogonal matrix. 
RandomNormal<T extends TFloating> Initializer that generates tensors with a normal distribution. 
RandomUniform<T extends TNumber> Initializer that generates tensors with a uniform distribution. 
TruncatedNormal<T extends TFloating> Initializer that generates a truncated normal distribution. 
VarianceScaling<T extends TFloating> Initializer capable of adapting its scale to the shape of weights tensors. 
Zeros<T extends TType> Creates an Initializer that sets all values to zero. 

Enums

VarianceScaling.Distribution The random distribution to use when initializing the values. 
VarianceScaling.Mode The mode to use for calculating the fan values.