CustomOpProp

Instance Constructors

new CustomOpProp(needTopGrad: Boolean = false)

needTopGrad
: Boolean The default declareBackwardDependency function use this value to determine whether this operator needs gradient input for above.

Abstract Value Members

abstract def createOperator(ctx: String, inShapes: Array[Array[Int]], inDtypes: Array[Int]): CustomOp

Create an operator that carries out the real computation given the context, input shapes, and input data types.

Concrete 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
final def asInstanceOf[T0]: T0

Definition Classes
Any
def clone(): AnyRef

Attributes
protected[java.lang]
Definition Classes
AnyRef
Annotations
@throws( ... )
def declareBackwardDependency(outGrad: Array[Int], inData: Array[Int], outData: Array[Int]): Array[Int]

Declare dependencies of this operator for backward pass.
Declare dependencies of this operator for backward pass.
outGrad
: array of Int ids of outGrad blobs.
inData
: array of Int ids of inData blobs.
outData
: array of Int ids of outData blobs.
returns
deps : array of Int ids of the needed blobs.
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
def inferShape(inShape: Array[Shape]): (Array[Shape], Array[Shape], Array[Shape])

inferShape interface.
inferShape interface. override to create new operators
inShape
: array of Shape list of argument shapes in the same order as declared in listArguments().
returns
inShapes : array of Shape array of argument shapes. Can be modified from inShape. outShapes : array of Shape array of output shapes calculated from inShape, in the same order as declared in listOutputs(). auxShapes : Optional, array of Shape array of aux shapes calculated from inShape, in the same order as declared in listAuxiliaryStates().
def inferType(inType: Array[DType.DType]): (Array[DType.DType], Array[DType.DType], Array[DType.DType])

inferType interface.
inferType interface. override to create new operators
inType
: array of DType list of argument types in the same order as declared in listArguments().
returns
inTypes : array of DType array of argument types. Can be modified from inType. outTypes : array of DType array of output types calculated from inType, in the same order as declared in listOutputs(). auxTypes : Optional, array of DType array of aux types calculated from inType, in the same order as declared in listAuxiliaryStates().
final def isInstanceOf[T0]: Boolean

Definition Classes
Any
var kwargs: Map[String, String]

Attributes
protected
def listArguments(): Array[String]

listArguments interface.
listArguments interface. override to create new operators
returns
arguments : array of String list of argument blob names.
def listAuxiliaryStates(): Array[String]

listAuxiliaryStates interface.
listAuxiliaryStates interface. override to create new operators
returns
auxs : array of String list of auxiliary state blob names.
def listOutputs(): Array[String]

listOutputs interface.
listOutputs interface. override to create new operators
returns
outputs : array of String list of output blob names.
final def ne(arg0: AnyRef): Boolean

Definition Classes
AnyRef
final def notify(): Unit

Definition Classes
AnyRef
final def notifyAll(): Unit

Definition Classes
AnyRef
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 Doc: package mxnet

abstract class CustomOpProp extends AnyRef

Instance Constructors

new CustomOpProp(needTopGrad: Boolean = false)

Abstract Value Members

abstract def createOperator(ctx: String, inShapes: Array[Array[Int]], inDtypes: Array[Int]): CustomOp

Concrete Value Members

final def !=(arg0: Any): Boolean

final def ##(): Int

final def ==(arg0: Any): Boolean

final def asInstanceOf[T0]: T0

def clone(): AnyRef

def declareBackwardDependency(outGrad: Array[Int], inData: Array[Int], outData: Array[Int]): Array[Int]

final def eq(arg0: AnyRef): Boolean

def equals(arg0: Any): Boolean

def finalize(): Unit

final def getClass(): Class[_]

def hashCode(): Int

def inferShape(inShape: Array[Shape]): (Array[Shape], Array[Shape], Array[Shape])

def inferType(inType: Array[DType.DType]): (Array[DType.DType], Array[DType.DType], Array[DType.DType])

final def isInstanceOf[T0]: Boolean

var kwargs: Map[String, String]

def listArguments(): Array[String]

def listAuxiliaryStates(): Array[String]

def listOutputs(): Array[String]

final def ne(arg0: AnyRef): Boolean

final def notify(): Unit

final def notifyAll(): Unit

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 AnyRef

Inherited from Any

Ungrouped