Overview > klyn.math.nn > Sequential > compile
Version 0.2.2 Overview Index Tree Deprecated
klyn.math.nn.Sequential.compile
method
public compile(optimizer as Optimizer<T>, loss as RegressionLoss<T>, metrics as IList<String> = []) as Void
Description

Configures a typed floating-target regression objective.

Parameters
ParameterDescription
optimizerParameter update rule; weights are retained.
lossObjective accepting targets in T.
metricsEmpty for regression; classification accuracy is rejected.
Throws
  • ValueException for null components or unsupported metrics.
Example

model.compile(Adam<Float>(), MeanSquaredError<Float>())