Table of Contents

Namespace AiDotNet.KnowledgeDistillation.Strategies

Classes

AccuracyBasedAdaptiveStrategy<T>

Adaptive distillation strategy that adjusts temperature based on student accuracy.

AdaptiveDistillationStrategyBase<T>

Abstract base class for adaptive distillation strategies with performance tracking.

AttentionDistillationStrategy<T>

Implements attention-based knowledge distillation for transformer models. Transfers knowledge through attention patterns rather than just final outputs.

ConfidenceBasedAdaptiveStrategy<T>

Adaptive distillation strategy that adjusts temperature based on student confidence.

ContrastiveDistillationStrategy<T>

Implements Contrastive Representation Distillation (CRD) which transfers knowledge through contrastive learning of sample relationships rather than just matching outputs.

CurriculumDistillationStrategyBase<T>

Abstract base class for curriculum distillation strategies with progressive difficulty adjustment.

EasyToHardCurriculumStrategy<T>

Curriculum distillation strategy that progresses from easy to hard samples.

EntropyBasedAdaptiveStrategy<T>

Adaptive distillation strategy that adjusts temperature based on prediction entropy.

FactorTransferDistillationStrategy<T>
FlowBasedDistillationStrategy<T>

Flow-based distillation that matches the information flow between layers.

HardToEasyCurriculumStrategy<T>

Curriculum distillation strategy that progresses from hard to easy samples.

HybridDistillationStrategy<T>

Hybrid distillation strategy that combines multiple distillation strategies with configurable weights.

NeuronSelectivityDistillationStrategy<T>

Neuron selectivity distillation that transfers the activation patterns and selectivity of individual neurons.

ProbabilisticDistillationStrategy<T>

Probabilistic distillation that transfers distributional knowledge by matching statistical properties.

RelationalDistillationStrategy<T>
SimilarityPreservingStrategy<T>

Similarity-preserving distillation that preserves pairwise similarity structure.

VariationalDistillationStrategy<T>

Variational distillation based on variational inference principles and information theory.

Interfaces

IAdaptiveDistillationStrategy<T>

Interface for adaptive distillation strategies that adjust temperature based on student performance.

ICurriculumDistillationStrategy<T>

Interface for curriculum distillation strategies that progressively adjust training difficulty.

Enums

AttentionMatchingMode

Defines how to match attention patterns between teacher and student.

ContrastiveMode

Defines the contrastive learning mode.

FactorMode
ProbabilisticMode
RelationalDistanceMetric

Distance metrics for relational knowledge distillation.

SelectivityMetric
VariationalMode