org.deeplearning4j.arbiter.optimize.api

Interface TaskCreator

All Known Implementing Classes:

ComputationGraphTaskCreator, MultiLayerNetworkTaskCreator

public interface TaskCreator

The TaskCreator is used to take a candidate configuration, data provider and score function, and create something that can be executed as a Callable

Author:

Alex Black

Method Summary

Modifier and Type	Method and Description
Callable<OptimizationResult>	create(Candidate candidate, Class<? extends DataSource> dataSource, Properties dataSourceProperties, ScoreFunction scoreFunction, List<StatusListener> statusListeners, IOptimizationRunner runner) Generate a callable that can be executed to conduct the training of this model (given the model configuration)
Callable<OptimizationResult>	create(Candidate candidate, DataProvider dataProvider, ScoreFunction scoreFunction, List<StatusListener> statusListeners, IOptimizationRunner runner) Deprecated.

Method Detail

create

@Deprecated
Callable<OptimizationResult> create(Candidate candidate,
                                                 DataProvider dataProvider,
                                                 ScoreFunction scoreFunction,
                                                 List<StatusListener> statusListeners,
                                                 IOptimizationRunner runner)

Deprecated.

Generate a callable that can be executed to conduct the training of this model (given the model configuration)

Parameters:

candidate - Candidate (model) configuration to be trained

dataProvider - DataProvider, for the data

scoreFunction - Score function to be used to evaluate the model

statusListeners - Status listeners, that can be used for callbacks (to UI, for example)

Returns:

A callable that returns an OptimizationResult, once optimization is complete

create

Callable<OptimizationResult> create(Candidate candidate,
                                    Class<? extends DataSource> dataSource,
                                    Properties dataSourceProperties,
                                    ScoreFunction scoreFunction,
                                    List<StatusListener> statusListeners,
                                    IOptimizationRunner runner)

Generate a callable that can be executed to conduct the training of this model (given the model configuration)

Parameters:

candidate - Candidate (model) configuration to be trained

dataSource - Data source

dataSourceProperties - Properties (may be null) for the data source

scoreFunction - Score function to be used to evaluate the model

statusListeners - Status listeners, that can be used for callbacks (to UI, for example)

Returns:

A callable that returns an OptimizationResult, once optimization is complete