org.deeplearning4j.spark.api

Interface TrainingMaster<R extends TrainingResult,W extends TrainingWorker<R>>

All Known Implementing Classes:

ParameterAveragingTrainingMaster

public interface TrainingMaster<R extends TrainingResult,W extends TrainingWorker<R>>

A TrainingMaster controls how distributed training is executed in practice
In principle, a large number of different approches can be used in distributed training (synchronous vs. asynchronous, parameter vs. gradient averaging, etc). Each of these different approaches would be implemented as a TrainingMaster; this allows SparkDl4jMultiLayer and SparkComputationGraph to be used with different training methods.

Method Summary

Modifier and Type	Method and Description
void	addHook(TrainingHook trainingHook) Add a hook for the master for pre and post training
boolean	deleteTempFiles(org.apache.spark.api.java.JavaSparkContext sc) Attempt to delete any temporary files generated by this TrainingMaster.
boolean	deleteTempFiles(org.apache.spark.SparkContext sc) Attempt to delete any temporary files generated by this TrainingMaster.
void	executeTraining(SparkComputationGraph network, org.apache.spark.api.java.JavaPairRDD<java.lang.String,org.apache.spark.input.PortableDataStream> trainingData) Deprecated. Deprecated due to poor performance
void	executeTraining(SparkComputationGraph graph, org.apache.spark.api.java.JavaRDD<org.nd4j.linalg.dataset.DataSet> trainingData) Train the SparkComputationGraph with the specified data set
void	executeTraining(SparkDl4jMultiLayer network, org.apache.spark.api.java.JavaPairRDD<java.lang.String,org.apache.spark.input.PortableDataStream> trainingData) Deprecated. Deprecated due to poor performance
void	executeTraining(SparkDl4jMultiLayer network, org.apache.spark.api.java.JavaRDD<org.nd4j.linalg.dataset.DataSet> trainingData) Train the SparkDl4jMultiLayer with the specified data set
void	executeTrainingMDS(SparkComputationGraph network, org.apache.spark.api.java.JavaPairRDD<java.lang.String,org.apache.spark.input.PortableDataStream> trainingData) Deprecated. Deprecated due to poor performance
void	executeTrainingMDS(SparkComputationGraph graph, org.apache.spark.api.java.JavaRDD<org.nd4j.linalg.dataset.api.MultiDataSet> trainingData) Train the SparkComputationGraph with the specified data set
void	executeTrainingPaths(SparkComputationGraph network, org.apache.spark.api.java.JavaRDD<java.lang.String> trainingDataPaths) EXPERIMENTAL method, may be removed in a future release. Fit the network using a list of paths for serialized DataSet objects.
void	executeTrainingPaths(SparkDl4jMultiLayer network, org.apache.spark.api.java.JavaRDD<java.lang.String> trainingDataPaths) EXPERIMENTAL method, may be removed in a future release. Fit the network using a list of paths for serialized DataSet objects.
void	executeTrainingPathsMDS(SparkComputationGraph network, org.apache.spark.api.java.JavaRDD<java.lang.String> trainingMultiDataSetPaths) EXPERIMENTAL method, may be removed in a future release. Fit the network using a list of paths for serialized MultiDataSet objects.
boolean	getIsCollectTrainingStats() Get the current setting for collectTrainingStats
SparkTrainingStats	getTrainingStats() Return the training statistics.
W	getWorkerInstance(SparkComputationGraph graph) Get the worker instance for this training master
W	getWorkerInstance(SparkDl4jMultiLayer network) Get the worker instance for this training master
void	removeHook(TrainingHook trainingHook) Remove a training hook from the worker
void	setCollectTrainingStats(boolean collectTrainingStats) Set whether the training statistics should be collected.
void	setListeners(java.util.Collection<IterationListener> listeners) Set the iteration listeners.
void	setListeners(StatsStorageRouter router, java.util.Collection<IterationListener> listeners) Set the iteration listeners and the StatsStorageRouter.
java.lang.String	toJson() Get the TrainingMaster configuration as JSON
java.lang.String	toYaml() Get the TrainingMaster configuration as YAML

Method Detail

removeHook

void removeHook(TrainingHook trainingHook)

Remove a training hook from the worker

Parameters:

trainingHook - the training hook to remove

addHook

void addHook(TrainingHook trainingHook)

Add a hook for the master for pre and post training

Parameters:

trainingHook - the training hook to add

toJson

java.lang.String toJson()

Get the TrainingMaster configuration as JSON

toYaml

java.lang.String toYaml()

Get the TrainingMaster configuration as YAML

getWorkerInstance

W getWorkerInstance(SparkDl4jMultiLayer network)

Get the worker instance for this training master

Parameters:

network - Current SparkDl4jMultiLayer

Returns:

Worker instance

getWorkerInstance

W getWorkerInstance(SparkComputationGraph graph)

Get the worker instance for this training master

Parameters:

graph - Current SparkComputationGraph

Returns:

Worker instance

executeTraining

void executeTraining(SparkDl4jMultiLayer network,
                     org.apache.spark.api.java.JavaRDD<org.nd4j.linalg.dataset.DataSet> trainingData)

Train the SparkDl4jMultiLayer with the specified data set

Parameters:

network - Current network state

trainingData - Data to train on

executeTraining

@Deprecated
void executeTraining(SparkDl4jMultiLayer network,
                                 org.apache.spark.api.java.JavaPairRDD<java.lang.String,org.apache.spark.input.PortableDataStream> trainingData)

Deprecated. Deprecated due to poor performance

Train the SparkDl4jMultiLayer with the specified serialized DataSet objects. The assumption here is that the PortableDataStreams are for DataSet objects, one per file.

Parameters:

network - Current network state

trainingData - Data to train on

executeTrainingPaths

void executeTrainingPaths(SparkDl4jMultiLayer network,
                          org.apache.spark.api.java.JavaRDD<java.lang.String> trainingDataPaths)

EXPERIMENTAL method, may be removed in a future release.
Fit the network using a list of paths for serialized DataSet objects.

Parameters:

network - Current network state

trainingDataPaths - Data to train on

executeTraining

void executeTraining(SparkComputationGraph graph,
                     org.apache.spark.api.java.JavaRDD<org.nd4j.linalg.dataset.DataSet> trainingData)

Train the SparkComputationGraph with the specified data set

Parameters:

graph - Current network state

trainingData - Data to train on

executeTraining

@Deprecated
void executeTraining(SparkComputationGraph network,
                                 org.apache.spark.api.java.JavaPairRDD<java.lang.String,org.apache.spark.input.PortableDataStream> trainingData)

Deprecated. Deprecated due to poor performance

Train the SparkComputationGraph with the specified serialized DataSet objects. The assumption here is that the PortableDataStreams are for DataSet objects, one per file, and that these have been serialized using DataSet.save(OutputStream)

Parameters:

network - Current network state

trainingData - Data to train on

executeTrainingPaths

void executeTrainingPaths(SparkComputationGraph network,
                          org.apache.spark.api.java.JavaRDD<java.lang.String> trainingDataPaths)

EXPERIMENTAL method, may be removed in a future release.
Fit the network using a list of paths for serialized DataSet objects.

Parameters:

network - Current network state

trainingDataPaths - Data to train on

executeTrainingPathsMDS

void executeTrainingPathsMDS(SparkComputationGraph network,
                             org.apache.spark.api.java.JavaRDD<java.lang.String> trainingMultiDataSetPaths)

EXPERIMENTAL method, may be removed in a future release.
Fit the network using a list of paths for serialized MultiDataSet objects.

Parameters:

network - Current network state

trainingMultiDataSetPaths - Data to train on

executeTrainingMDS

void executeTrainingMDS(SparkComputationGraph graph,
                        org.apache.spark.api.java.JavaRDD<org.nd4j.linalg.dataset.api.MultiDataSet> trainingData)

Train the SparkComputationGraph with the specified data set

Parameters:

graph - Current network state

trainingData - Data to train on

executeTrainingMDS

@Deprecated
void executeTrainingMDS(SparkComputationGraph network,
                                    org.apache.spark.api.java.JavaPairRDD<java.lang.String,org.apache.spark.input.PortableDataStream> trainingData)

Deprecated. Deprecated due to poor performance

Train the SparkComputationGraph with the specified serialized MultiDataSet objects. The assumption here is that the PortableDataStreams are for MultiDataSet objects, one per file.

Parameters:

network - Current network state

trainingData - Data to train on

setCollectTrainingStats

void setCollectTrainingStats(boolean collectTrainingStats)

Set whether the training statistics should be collected. Training statistics may include things like per-epoch run times, time spent waiting for data, etc.

These statistics are primarily used for debugging and optimization, in order to gain some insight into what aspects of network training are taking the most time.

Parameters:

collectTrainingStats - If true: collecting training statistics will be

getIsCollectTrainingStats

boolean getIsCollectTrainingStats()

Get the current setting for collectTrainingStats

getTrainingStats

SparkTrainingStats getTrainingStats()

Return the training statistics. Note that this may return null, unless setCollectTrainingStats has been set first

Returns:

Training statistics

setListeners

void setListeners(java.util.Collection<IterationListener> listeners)

Set the iteration listeners. These should be called after every averaging (or similar) operation in the TrainingMaster, though the exact behaviour may be dependent on each IterationListener

Parameters:

listeners - Listeners to set

setListeners

void setListeners(StatsStorageRouter router,
                  java.util.Collection<IterationListener> listeners)

Set the iteration listeners and the StatsStorageRouter. This is typically used for UI functionality: for example, setListeners(new FileStatsStorage(myFile), Collections.singletonList(new StatsListener(null))). This will pass a StatsListener to each worker, and then shuffle the results back to the specified FileStatsStorage instance (which can then be attached to the UI or loaded later)

Parameters:

router - StatsStorageRouter in which to place the results

listeners - Listeners

deleteTempFiles

boolean deleteTempFiles(org.apache.spark.api.java.JavaSparkContext sc)

Attempt to delete any temporary files generated by this TrainingMaster. Depending on the configuration, no temporary files may be generated.

Parameters:

sc - JavaSparkContext (used to access HDFS etc file systems, when required)

Returns:

True if deletion was successful (or, no files to delete); false otherwise.

deleteTempFiles

boolean deleteTempFiles(org.apache.spark.SparkContext sc)

Attempt to delete any temporary files generated by this TrainingMaster. Depending on the configuration, no temporary files may be generated.

Parameters:

sc - SparkContext (used to access HDFS etc file systems, when required)

Returns:

True if deletion was successful (or, no files to delete); false otherwise.