Package ai.djl.training.optimizer
Class AdamW.Builder
java.lang.Object
ai.djl.training.optimizer.Optimizer.OptimizerBuilder<AdamW.Builder>
ai.djl.training.optimizer.AdamW.Builder
- Enclosing class:
- AdamW
The Builder to construct an
AdamW object.-
Method Summary
Modifier and TypeMethodDescriptionbuild()Builds aAdamWblock.optBeta1(float beta1) Sets the decay rate for the first moment estimates.optBeta2(float beta2) Sets the decay rate for the second moment estimates.optEpsilon(float epsilon) Sets \(epsilon\) - a small quantity for numerical stability.optLearningRateTracker(ParameterTracker learningRateTracker) Sets theParameterTrackerfor this optimizer.protected AdamW.Builderself()Methods inherited from class ai.djl.training.optimizer.Optimizer.OptimizerBuilder
optBeginNumUpdate, optClipGrad, optWeightDecays, setRescaleGrad
-
Method Details
-
self
- Specified by:
selfin classOptimizer.OptimizerBuilder<AdamW.Builder>
-
optLearningRateTracker
Sets theParameterTrackerfor this optimizer.- Parameters:
learningRateTracker- theParameterTrackerto be set- Returns:
- this
Builder
-
optBeta1
Sets the decay rate for the first moment estimates.- Parameters:
beta1- the deacay rate for the the first moment estimates- Returns:
- this
Builder
-
optBeta2
Sets the decay rate for the second moment estimates.- Parameters:
beta2- the decay rate for the the second moment estimates- Returns:
- this
Builder
-
optEpsilon
Sets \(epsilon\) - a small quantity for numerical stability.- Parameters:
epsilon- a small quantity for numerical stability- Returns:
- this
Builder
-
build
Builds aAdamWblock.- Returns:
- the
AdamWblock
-