object DistriOptimizer extends AbstractOptimizer
Linear Supertypes
Ordering
- Alphabetic
- By Inheritance
Inherited
- DistriOptimizer
- AbstractOptimizer
- AnyRef
- Any
- Hide All
- Show All
Visibility
- Public
- All
Type Members
-
abstract
class
Cache[T] extends AnyRef
Optimizer cache some metadata on each executor
Optimizer cache some metadata on each executor
- T
Tensor element type
- case class CacheV1[T](localModels: Array[Module[T]], modelWeights: Array[Tensor[T]], modelGradients: Array[Tensor[T]], localCriterions: Array[Criterion[T]], localStates: Array[Table], moduleTimeList: Array[Long] = null, localMethods: Array[Option[Array[ValidationMethod[T]]]], optimMethods: Map[String, OptimMethod[T]], parameterSynchronizer: DistriParameterSynchronizer[T] = null) extends Cache[T] with Product with Serializable
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
checkpoint[T](cacheTrigger: Option[Trigger], cachePath: Option[String], isOverWrite: Boolean, wallClockTime: Long, models: RDD[Cache[T]], state: Table, parameters: AllReduceParameter[T], optimMethods: Map[String, OptimMethod[T]], trainingModel: Module[T])(implicit arg0: ClassTag[T], ev: TensorNumeric[T]): Unit
* Create checkpoint.
* Create checkpoint.
- cacheTrigger
cache trigger
- cachePath
cache path
- isOverWrite
whether over write
- wallClockTime
wall clock time
- models
cached models
- state
state table
- parameters
all reduce parameters
- optimMethods
all optim methods
- trainingModel
training model
- Attributes
- protected
- Definition Classes
- AbstractOptimizer
-
def
clone(): AnyRef
- Attributes
- protected[java.lang]
- Definition Classes
- AnyRef
- Annotations
- @native() @throws( ... )
-
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
- Annotations
- @native()
-
def
getModel[T](models: RDD[Cache[T]], parameters: AllReduceParameter[T], trainingModel: Module[T])(implicit arg0: ClassTag[T], ev: TensorNumeric[T]): Module[T]
Fetch current model parameters to driver, and copy to trainingModel.
Fetch current model parameters to driver, and copy to trainingModel.
- models
cached models
- parameters
AllReduceParameter
- trainingModel
the model is trained by optimizer
- returns
trained model
- Attributes
- protected
- Definition Classes
- DistriOptimizer → AbstractOptimizer
-
def
hashCode(): Int
- Definition Classes
- AnyRef → Any
- Annotations
- @native()
-
final
def
isInstanceOf[T0]: Boolean
- Definition Classes
- Any
- val logger: Logger
-
final
def
ne(arg0: AnyRef): Boolean
- Definition Classes
- AnyRef
-
final
def
notify(): Unit
- Definition Classes
- AnyRef
- Annotations
- @native()
-
final
def
notifyAll(): Unit
- Definition Classes
- AnyRef
- Annotations
- @native()
-
def
saveSummary[T](trainSummary: TrainSummary, models: RDD[Cache[T]], driverState: Table, parameters: AllReduceParameter[T], trainingModel: Module[T])(implicit arg0: ClassTag[T], ev: TensorNumeric[T]): Unit
Save train summaries.
Save train summaries.
- trainSummary
train logger
- models
cached models
- driverState
driver state
- parameters
AllReduceParameter
- Attributes
- protected
- Definition Classes
- AbstractOptimizer
-
final
def
synchronized[T0](arg0: ⇒ T0): T0
- Definition Classes
- AnyRef
-
def
toString(): String
- Definition Classes
- AnyRef → Any
-
def
validate[T](validationTrigger: Option[Trigger], validationDataSet: Option[DataSet[MiniBatch[T]]], validationMethods: Option[Array[ValidationMethod[T]]], coresPerNode: Int, models: RDD[Cache[T]], state: Table, validationSummary: Option[ValidationSummary], header: String, parameters: AllReduceParameter[T] = null): Unit
Validate current model and save the result.
Validate current model and save the result.
- validationTrigger
validation trigger
- validationDataSet
validation dataset
- validationMethods
validation methods
- coresPerNode
cores per node
- models
cached models
- state
state table
- validationSummary
validation logger.
- header
log header string
- Attributes
- protected
- Definition Classes
- AbstractOptimizer
-
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
- @native() @throws( ... )