IEpisode<T, TInput, TOutput>

Interfaces · AiDotNet.Interfaces

Represents a single episode in meta-learning, wrapping an IMetaLearningTask with additional metadata such as domain, difficulty, and timing information.

Known implementations: Episode<T, TInput, TOutput>

For Beginners

An episode is one complete “mini-learning session” in meta-learning. Each episode contains a task (support set + query set) along with extra information about that task, like how hard it is and what domain it came from.

Properties

Property Summary
Dictionary<string, object> EpisodeMetadata { get; } Gets optional key-value metadata associated with the episode.
IMetaLearningTask<T, TInput, TOutput> Task { get; } Gets the underlying meta-learning task containing support and query sets.
double? Difficulty { get; set; } Gets an optional difficulty score for this episode, typically in [0, 1].
double? LastLoss { get; set; } Gets or sets the loss observed on this episode during the most recent evaluation.
int EpisodeId { get; } Gets the unique identifier for this episode.
int SampleCount { get; set; } Gets or sets the number of times this episode has been sampled.
long CreatedTimestamp { get; } Gets the timestamp when this episode was created.
string Domain { get; } Gets an optional domain or category label for this episode.