IDistillationStrategy<T, TInput, TOutput>

Interfaces · AiDotNet.ContinualLearning.Interfaces

Extended strategy interface for knowledge distillation-based strategies.

Implements: IContinualLearningStrategy<T, TInput, TOutput>

Known implementations: LearningWithoutForgetting<T, TInput, TOutput>

Properties

Property Summary
IFullModel<T, TInput, TOutput> TeacherModel { get; } Gets the teacher model (frozen copy from before current task).
T DistillationWeight { get; } Gets the weight for distillation loss vs task loss.
T Temperature { get; } Gets the distillation temperature.

Methods

Method Summary
T ComputeDistillationLoss(TOutput teacherOutput, TOutput studentOutput) Computes the distillation loss between teacher and student outputs.