ICurriculumLearnerConfig<T>

Interfaces · AiDotNet.CurriculumLearning.Interfaces

Configuration interface for curriculum learning.

Known implementations: CurriculumLearnerConfig<T>

For Beginners

This configuration controls how curriculum learning behaves. You can set things like how many epochs to train, how to schedule the introduction of harder samples, and when to stop training.

How It Works

Key Configuration Areas:

Properties

Property Summary
Action<string> LogAction { get; } Gets the custom logging action.
CurriculumScheduleType ScheduleType { get; } Gets the type of curriculum schedule to use.
CurriculumVerbosity Verbosity { get; } Gets the verbosity level for logging.
T EarlyStoppingMinDelta { get; } Gets the minimum improvement required to reset patience.
T FinalDataFraction { get; } Gets the final data fraction (at final phase).
T InitialDataFraction { get; } Gets the initial data fraction (at phase 0).
T LearningRate { get; } Gets the learning rate.
bool NormalizeDifficulties { get; } Gets whether to normalize difficulty scores to [0, 1].
bool RecalculateDifficulties { get; } Gets whether to recalculate difficulty scores during training.
bool ShuffleWithinPhase { get; } Gets whether to shuffle within curriculum phases.
bool UseDifficultyWeighting { get; } Gets whether to apply sample weighting based on difficulty.
bool UseEarlyStopping { get; } Gets whether to use early stopping.
int BatchSize { get; } Gets the batch size for training.
int DifficultyRecalculationFrequency { get; } Gets how often to recalculate difficulties (in epochs).
int EarlyStoppingPatience { get; } Gets the patience for early stopping (epochs without improvement).
int EpochsPerPhase { get; } Gets the number of epochs per curriculum phase.
int NumPhases { get; } Gets the number of curriculum phases.
int TotalEpochs { get; } Gets the total number of training epochs.
int? RandomSeed { get; } Gets the random seed for reproducibility.