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.