PerDeviceVariable

Instance Constructors

new PerDeviceVariable(primaryVariable: variables.Variable, index: Map[DeviceSpecification, variables.Variable], reduction: Reduction)

primaryVariable
Primary variable.
index
Index map from devices to variables.
reduction
Reduction to use when saving the variables.

Attributes
protected

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 assign(value: Output, name: String = "Assign")(implicit context: DistributionContext): Output

Creates an op that assigns the provided value to this variable and returns its value.
Creates an op that assigns the provided value to this variable and returns its value.
value
Value to assign the variable to.
name
Name for created op.
returns
Variable value read op, after the assignment.

Annotations
@throws( ... )
def assignAdd(value: Output, name: String = "AssignAdd")(implicit context: DistributionContext): Output

Creates an op that adds the provided value to the current value of the variable and returns its value.
Creates an op that adds the provided value to the current value of the variable and returns its value.
value
Value to add to the current variable value.
name
Name for created op.
returns
Variable value read op, after the addition.

Annotations
@throws( ... ) @throws( ... )
def assignScatter(indices: Output, values: Output, name: String = "AssignScatter")(implicit context: DistributionContext): Output

Creates an op that applies updates the provided sparse value updates to this variable and returns its value.
Creates an op that applies updates the provided sparse value updates to this variable and returns its value.
indices
Indices corresponding to the values used for the update.
values
Values to use for updating, corresponding to the provided indices.
name
Name for created op.
returns
Variable value read op, after the addition.

Annotations
@throws( ... ) @throws( ... )
def assignScatterAdd(indices: Output, values: Output, name: String = "AssignScatterAdd")(implicit context: DistributionContext): Output

Creates an op that adds the provided sparse value to the current value of the variable and returns its value.
Creates an op that adds the provided sparse value to the current value of the variable and returns its value.
indices
Indices corresponding to the values being added.
values
Values to be added, corresponding to the provided indices.
name
Name for created op.
returns
Variable value read op, after the addition.

Annotations
@throws( ... ) @throws( ... )
def assignScatterSub(indices: Output, values: Output, name: String = "AssignScatterSub")(implicit context: DistributionContext): Output

Creates an op that subtracts the provided sparse value from the current value of the variable and returns its value.
Creates an op that subtracts the provided sparse value from the current value of the variable and returns its value.
indices
Indices corresponding to the values being subtracted.
values
Values to be subtracted, corresponding to the provided indices.
name
Name for created op.
returns
Variable value read op, after the subtraction.

Annotations
@throws( ... ) @throws( ... )
def assignSub(value: Output, name: String = "AssignAdd")(implicit context: DistributionContext): Output

Creates an op that subtracts the provided value from the current value of the variable and returns its value.
Creates an op that subtracts the provided value from the current value of the variable and returns its value.
value
Value to subtract from the current variable value.
name
Name for created op.
returns
Variable value read op, after the subtraction.

Annotations
@throws( ... ) @throws( ... )
def clone(): AnyRef

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

Data type of this variable.
Data type of this variable.

Definition Classes
DistributedVariable
def devices: Seq[DeviceSpecification]

Returns the devices on which this value is distributed.
Returns the devices on which this value is distributed.

Definition Classes
DistributedValue
val distributionType: Type

Type of this distributed value (e.g., per-device or mirrored).
Type of this distributed value (e.g., per-device or mirrored).

Definition Classes
DistributedValue
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] )
def gather(indices: Output, name: String = "Gather")(implicit context: DistributionContext): Output

Creates an op that reads the value of this variable sparsely, using the provided indices.
Creates an op that reads the value of this variable sparsely, using the provided indices.
This method should be used when there are multiple reads, or when it is desirable to read the value only after some condition is true.
indices
Indices to use for the sparse read.
name
Name for the created op.
returns
Created op.

Definition Classes
DistributedVariable
Annotations
@throws( ... )
def get(device: String = "current")(implicit context: DistributionContext): variables.Variable

Returns the value on the specified device (defaults to the current device, if not provided.
Returns the value on the specified device (defaults to the current device, if not provided.

Definition Classes
DistributedValue
final def getClass(): Class[_]

Definition Classes
AnyRef → Any
val graph: core.Graph

Graph where this variable is defined.
Graph where this variable is defined.

Definition Classes
DistributedVariable
def hashCode(): Int

Definition Classes
AnyRef → Any
val index: Map[DeviceSpecification, variables.Variable]

Index map from devices to variables.
Index map from devices to variables.

Definition Classes
PerDeviceVariable → DistributedVariable → PerDeviceValue → DistributedValue

def initializedValue(implicit context: DistributionContext): Output

Value of the initialized variable.

Value of the initialized variable. You should use this instead of the variable itself to initialize another variable with a value that depends on the value of this variable.

Example:

// Initialize `v` with random values, and then use `initializedValue` to guarantee that `v` has been initialized
// before its value is used to initialize `w`. The random tensor will only be sampled once.
val v = tf.variable("v", FLOAT32, Shape(10, 40), tf.RandomTruncatedNormalInitializer())
val w = tf.variable("w", initializer = tf.ConstantInitializer(v.initializedValue * 2.0))

Definition Classes: DistributedVariable

val initializer: Op

Returns the initializer for this variable.
Returns the initializer for this variable.

Definition Classes
DistributedVariable
val isInitialized: Output

Op output that is true when the variable has been initialized and false otherwise.
Op output that is true when the variable has been initialized and false otherwise.

Definition Classes
DistributedVariable
final def isInstanceOf[T0]: Boolean

Definition Classes
Any
val name: String

Name of this variable.
Name of this variable.

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

Definition Classes
AnyRef
final def notify(): Unit

Definition Classes
AnyRef
final def notifyAll(): Unit

Definition Classes
AnyRef
def onDevice(device: String): Boolean

Returns true if the values are distributed on the provided device.
Returns true if the values are distributed on the provided device.

Definition Classes
DistributedValue
def op(implicit context: DistributionContext): Op

Returns the op of this variable.
Returns the op of this variable.

Definition Classes
DistributedVariable
val primaryVariable: variables.Variable

Primary variable.
Primary variable.

Definition Classes
PerDeviceVariable → DistributedVariable
def read(name: String = "Read")(implicit context: DistributionContext): Output

Creates an op that reads the value of this variable.
Creates an op that reads the value of this variable.
This method should be used when there are multiple reads, or when it is desirable to read the value only after some condition is true.
The returned value may be different from that of value depending on the device being used, the control dependencies, etc.
returns
Created op.

Definition Classes
DistributedVariable
val reduction: Reduction

Reduction to use when saving the variables.
val shape: core.Shape

Shape of this variable.
Shape of this variable.

Definition Classes
DistributedVariable
val sharedName: String

Attributes
protected
Definition Classes
DistributedVariable
final def synchronized[T0](arg0: ⇒ T0): T0

Definition Classes
AnyRef
def toProto(exportScope: String): VariableDef

Alias for toVariableDef.
Alias for toVariableDef.

Definition Classes
DistributedVariable
def toProto: VariableDef

Converts this object to its corresponding ProtoBuf object.
Converts this object to its corresponding ProtoBuf object.
returns
ProtoBuf object corresponding to this object.

Definition Classes
DistributedVariable → Serializable
def toString(): String

Definition Classes
AnyRef → Any
def toVariableDef(exportScope: String): VariableDef

Convert this object to its corresponding ProtoBuf object.
Convert this object to its corresponding ProtoBuf object.
exportScope
Optional string specifying the name scope to remove. Only the ops within this name scope will be included in the resulting ProtoBuf object and the export scope will be stripped from their names to allow for easy import into new name scopes.
returns
ProtoBuf object corresponding to this object.

Definition Classes
DistributedVariable
def value(implicit context: DistributionContext): Output

Returns a cached op which reads the last value of this partitioned variable.
Returns a cached op which reads the last value of this partitioned variable.
You can not assign a new value to the returned tensor as it is not a reference to the variable.
The returned op output will not inherit the control dependencies from the scope where the value is used, which is equivalent behavior to that of getting the value of a variable.
NOTE: You usually do not need to call this method directly, as all ops that use variables do so by internally converting them to tensors.

Definition Classes
DistributedVariable
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 PerDeviceVariable | package values

class PerDeviceVariable extends PerDeviceValue[variables.Variable] with DistributedVariable

Instance Constructors

new PerDeviceVariable(primaryVariable: variables.Variable, index: Map[DeviceSpecification, variables.Variable], reduction: Reduction)

Value Members

final def !=(arg0: Any): Boolean

final def ##(): Int

final def ==(arg0: Any): Boolean

final def asInstanceOf[T0]: T0

def assign(value: Output, name: String = "Assign")(implicit context: DistributionContext): Output

def assignAdd(value: Output, name: String = "AssignAdd")(implicit context: DistributionContext): Output

def assignScatter(indices: Output, values: Output, name: String = "AssignScatter")(implicit context: DistributionContext): Output

def assignScatterAdd(indices: Output, values: Output, name: String = "AssignScatterAdd")(implicit context: DistributionContext): Output

def assignScatterSub(indices: Output, values: Output, name: String = "AssignScatterSub")(implicit context: DistributionContext): Output

def assignSub(value: Output, name: String = "AssignAdd")(implicit context: DistributionContext): Output

def clone(): AnyRef

val dataType: types.DataType

def devices: Seq[DeviceSpecification]

val distributionType: Type

final def eq(arg0: AnyRef): Boolean

def equals(arg0: Any): Boolean

def finalize(): Unit

def gather(indices: Output, name: String = "Gather")(implicit context: DistributionContext): Output

def get(device: String = "current")(implicit context: DistributionContext): variables.Variable

final def getClass(): Class[_]

val graph: core.Graph

def hashCode(): Int

val index: Map[DeviceSpecification, variables.Variable]

def initializedValue(implicit context: DistributionContext): Output

val initializer: Op

val isInitialized: Output

final def isInstanceOf[T0]: Boolean

val name: String

final def ne(arg0: AnyRef): Boolean

final def notify(): Unit

final def notifyAll(): Unit

def onDevice(device: String): Boolean

def op(implicit context: DistributionContext): Op

val primaryVariable: variables.Variable

def read(name: String = "Read")(implicit context: DistributionContext): Output

val reduction: Reduction

val shape: core.Shape

val sharedName: String

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

def toProto(exportScope: String): VariableDef

def toProto: VariableDef

def toString(): String

def toVariableDef(exportScope: String): VariableDef

def value(implicit context: DistributionContext): Output

final def wait(): Unit

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

final def wait(arg0: Long): Unit

Inherited from DistributedVariable

Inherited from Serializable

Inherited from PerDeviceValue[variables.Variable]

Inherited from DistributedValue[variables.Variable]

Inherited from AnyRef

Inherited from Any

Ungrouped