Fix earlystopping metric issue #3722 #3736
Merged
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
This PR fixes issue #3722 by extending the
EarlyStoppingListener
to support"trainLoss"
as a valid monitored metric in addition to the default"validateLoss"
.Previously, if a user specified
"trainLoss"
via.optMonitoredMetric("trainLoss")
, the early stopping logic would not behave as expected becauseTrainingResult.getTrainLoss()
was not considered in the metric selection. This update introduces an explicit check for"trainLoss"
in thegetMetric(...)
method, ensuring it is properly used for early stopping decisions.Changes Summary
getMetric(...)
method now explicitly supports"trainLoss"
in addition to"validateLoss"
.Result
.optMonitoredMetric("trainLoss")
, training is now stopped early if the training loss fails to improve as specified.Closes #3722