IMemoryBasedStrategy<T, TInput, TOutput>

Interfaces · AiDotNet.ContinualLearning.Interfaces

Extended strategy interface for strategies that store task examples.

Implements: IContinualLearningStrategy<T, TInput, TOutput>

Known implementations: IGradientConstraintStrategy<T, TInput, TOutput>

Properties

Property Summary
int MaxExamples { get; } Gets the maximum number of examples that can be stored.
int StoredExampleCount { get; } Gets the number of examples stored in memory.

Methods

Method Summary
IReadOnlyList<ValueTuple<TInput, TOutput, int>> SampleExamples(int batchSize) Samples a batch of stored examples for replay.
void ClearMemory() Clears all stored examples.
void StoreTaskExamples(IDataset<T, TInput, TOutput> taskData) Stores examples from a completed task.