CurriculumEpochMetrics<T>

Models & Types · AiDotNet.CurriculumLearning.Interfaces

Metrics from a curriculum learning epoch.

Constructors

Constructor Summary
CurriculumEpochMetrics()

Properties

Property Summary
T LossImprovement { get; set; } Gets or sets the improvement in loss from previous epoch.
T TrainingAccuracy { get; set; } Gets or sets the training accuracy (for classification).
T TrainingLoss { get; set; } Gets or sets the average training loss.
T ValidationAccuracy { get; set; } Gets or sets the validation accuracy (if applicable).
T ValidationLoss { get; set; } Gets or sets the validation loss (if validation data provided).
Vector<T> SampleLosses { get; set; } Gets or sets per-sample losses.
bool Improved { get; set; } Gets or sets whether this epoch showed improvement.
int Epoch { get; set; } Gets or sets the current epoch number.
int Phase { get; set; } Gets or sets the current phase number.
int SamplesUsed { get; set; } Gets or sets the number of samples used in this epoch.