SITrainerOptions<T>
Options & Configuration · AiDotNet.ContinualLearning.Trainers
Configuration options for the SI trainer.
For Beginners
These options control how the SI trainer operates. SI tracks parameter importance along the optimization trajectory.
Constructors
| Constructor | Summary |
|---|---|
SITrainerOptions() |
Properties
| Property | Summary |
|---|---|
bool? ComputeValidationMetrics { get; set; } |
Whether to compute validation metrics after each epoch. |
bool? NormalizeImportance { get; set; } |
Whether to normalize importance values after each task. |
bool? UseExperienceReplay { get; set; } |
Whether to use experience replay alongside SI regularization. |
double? DampingParameter { get; set; } |
Damping parameter (ξ) to prevent division by zero in importance computation. |
double? GradientClipThreshold { get; set; } |
Gradient clipping threshold. |
double? PathIntegralDecay { get; set; } |
Decay factor for path integral accumulation. |
double? RegularizationWeight { get; set; } |
Weight for the regularization loss relative to task loss. |
double? ReplayLearningRateFactor { get; set; } |
Fraction of learning rate to use for replay samples. |