VarianceScalingInitializer

Initializer capable of adapting its scale to the shape of weights tensors.

With the Normal distribution option, samples are drawn from a truncated Normal distribution centered on zero, and with standard deviation equal to sqrt(initialScale / n), where n is:

the number of input units in the weight tensor, if mode == FanInScalingMode,
the number of output units, if mode == FanOutScalingMode, or
the average of the numbers of input and output units, if mode == FanAverageScalingMode

With uniform distribution option, samples are drawn from a uniform distribution within [-limit, limit], where limit = sqrt(3 * initialScale / n).

Linear Supertypes

Initializer, AnyRef, Any

Known Subclasses

GlorotNormalInitializer, GlorotUniformInitializer

Instance Constructors

new VarianceScalingInitializer(initialScale: Float = 1.0f, scalingMode: ScalingMode = FanInScalingMode, distribution: Distribution = ..., seed: Option[Int] = None)

initialScale
Initial variance scale.
scalingMode
Variance scaling mode.
distribution
Distribution to use when sampling.
seed
Optional random seed, used to generate a random seed pair for the random number generator, when combined with the graph-level seed.

Value Members

final def !=(arg0: Any): Boolean

Definition Classes
AnyRef → Any
final def ##(): Int

Definition Classes
AnyRef → Any
final def ==(arg0: Any): Boolean

Definition Classes
AnyRef → Any
def apply(dataType: types.DataType, shape: core.Shape, partitionInfo: PartitionInformation): Output

Definition Classes
Initializer
final def asInstanceOf[T0]: T0

Definition Classes
Any
def clone(): AnyRef

Attributes
protected[java.lang]
Definition Classes
AnyRef
Annotations
@throws( ... )
val dataType: types.DataType

Data type of the values produced by this initializer.
Data type of the values produced by this initializer. If null, then the initializer may produce values of any data type.

Definition Classes
Initializer
val distribution: Distribution

Distribution to use when sampling.
final def eq(arg0: AnyRef): Boolean

Definition Classes
AnyRef
def equals(arg0: Any): Boolean

Definition Classes
AnyRef → Any
def finalize(): Unit

Attributes
protected[java.lang]
Definition Classes
AnyRef
Annotations
@throws( classOf[java.lang.Throwable] )
final def getClass(): Class[_]

Definition Classes
AnyRef → Any
def hashCode(): Int

Definition Classes
AnyRef → Any
val initialScale: Float

Initial variance scale.
def initialValue(dataType: types.DataType, shape: core.Shape, partitionInfo: PartitionInformation): Output

Generates an initial value op.
Generates an initial value op.
dataType
Data type for the output tensor.
shape
Shape for the output tensor.
partitionInfo
PartitionInformation object holding additional information about how the variable is partitioned. May be null if the variable is not partitioned.
returns
Created op output.

Definition Classes
VarianceScalingInitializer → Initializer
Annotations
@throws( ... )
Exceptions thrown
ShapeMismatchException If the initializer cannot produce a value with the requested shape.
final def isInstanceOf[T0]: Boolean

Definition Classes
Any
final def ne(arg0: AnyRef): Boolean

Definition Classes
AnyRef
final def notify(): Unit

Definition Classes
AnyRef
final def notifyAll(): Unit

Definition Classes
AnyRef
val scalingMode: ScalingMode

Variance scaling mode.
val seed: Option[Int]

Optional random seed, used to generate a random seed pair for the random number generator, when combined with the graph-level seed.
val shape: core.Shape

Shape of the values produced by this initializer.
Shape of the values produced by this initializer. If null, then the initializer may produce values of any shape.

Definition Classes
Initializer
final def synchronized[T0](arg0: ⇒ T0): T0

Definition Classes
AnyRef
def toString(): String

Definition Classes
AnyRef → Any
final def wait(): Unit

Definition Classes
AnyRef
Annotations
@throws( ... )
final def wait(arg0: Long, arg1: Int): Unit

Definition Classes
AnyRef
Annotations
@throws( ... )
final def wait(arg0: Long): Unit

Definition Classes
AnyRef
Annotations
@throws( ... )

Related Docs: object VarianceScalingInitializer | package variables

class VarianceScalingInitializer extends Initializer

Instance Constructors

new VarianceScalingInitializer(initialScale: Float = 1.0f, scalingMode: ScalingMode = FanInScalingMode, distribution: Distribution = ..., seed: Option[Int] = None)

Value Members

final def !=(arg0: Any): Boolean

final def ##(): Int

final def ==(arg0: Any): Boolean

def apply(dataType: types.DataType, shape: core.Shape, partitionInfo: PartitionInformation): Output

final def asInstanceOf[T0]: T0

def clone(): AnyRef

val dataType: types.DataType

val distribution: Distribution

final def eq(arg0: AnyRef): Boolean

def equals(arg0: Any): Boolean

def finalize(): Unit

final def getClass(): Class[_]

def hashCode(): Int

val initialScale: Float

def initialValue(dataType: types.DataType, shape: core.Shape, partitionInfo: PartitionInformation): Output

final def isInstanceOf[T0]: Boolean

final def ne(arg0: AnyRef): Boolean

final def notify(): Unit

final def notifyAll(): Unit

val scalingMode: ScalingMode

val seed: Option[Int]

val shape: core.Shape

final def synchronized[T0](arg0: ⇒ T0): T0

def toString(): String

final def wait(): Unit

final def wait(arg0: Long, arg1: Int): Unit

final def wait(arg0: Long): Unit

Inherited from Initializer

Inherited from AnyRef

Inherited from Any

Ungrouped