Package opennlp.tools.monitoring

Interface TrainingProgressMonitor

All Known Implementing Classes:

DefaultTrainingProgressMonitor

public interface TrainingProgressMonitor

An interface to capture training progress of an AbstractModel.

Method Summary

Modifier and Type	Method	Description
void	display(boolean clear)	Displays the training progress and optionally clears the recorded progress (to save memory).
void	finishedIteration(int iteration, int numberCorrectEvents, int totalEvents, TrainingMeasure measure, double measureValue)	Captures the Iteration progress.
void	finishedTraining(int iterations, StopCriteria<?> stopCriteria)	Captures the training completion progress.
boolean	isTrainingFinished()	Checks whether the training has finished.

Method Details

finishedIteration

void finishedIteration(int iteration,
 int numberCorrectEvents,
 int totalEvents,
 TrainingMeasure measure,
 double measureValue)

Captures the Iteration progress.

Parameters:

iteration - The completed iteration number.

numberCorrectEvents - Number of correctly predicted events in this iteration.

totalEvents - Total count of events processed in this iteration.

measure - TrainingMeasure.

measureValue - measure value corresponding to the applicable TrainingMeasure.

finishedTraining

void finishedTraining(int iterations,
 StopCriteria<?> stopCriteria)

Captures the training completion progress.

Parameters:

iterations - Total number of iterations configured for the training.

stopCriteria - StopCriteria for the training.

isTrainingFinished

boolean isTrainingFinished()

Checks whether the training has finished.

Returns:

true if the training has finished, false if the training is not yet completed.

display

void display(boolean clear)

Displays the training progress and optionally clears the recorded progress (to save memory). Callers of this method can invoke it periodically during training, to avoid holding too much progress related data in memory.

Parameters:

clear - Set to true to clear the recorded progress.