MetaAdaptationResult<T>

Models & Types · AiDotNet.Models.Results

Results from adapting a meta-learner to a single task.

For Beginners

Meta-learning has two phases for each task:

  1. Adaptation (Inner Loop): The model quickly learns from the support set (a few examples)
  2. Evaluation: We test on the query set (held-out examples) to see if it really learned

This result tracks:

  • Support metrics: How well the model fits the training examples (should be very good)
  • Query metrics: How well it generalizes (the real test!)
  • Adaptation details: How many steps, how long it took, etc.

For example, in 5-way 1-shot learning:

  • Support set: 5 examples (1 per class) - model adapts on these
  • Query set: 15 examples (3 per class) - model is evaluated on these
  • Good meta-learning: High query accuracy even with tiny support set

How It Works

This class captures detailed metrics from the inner loop adaptation process for a single task. It tracks both support set performance (where the model adapts) and query set performance (where we measure true generalization), along with adaptation details.

Constructors

Constructor Summary
MetaAdaptationResult(T queryAccuracy, T queryLoss, T supportAccuracy, T supportLoss, int adaptationSteps, double adaptationTimeMs, List<T> perStepLosses, Dictionary<string, T> additionalMetrics) Initializes a new instance with adaptation metrics.

Properties

Property Summary
Dictionary<string, T> AdditionalMetrics { get; } Gets algorithm-specific metrics for this adaptation.
List<T> PerStepLosses { get; } Gets the loss values recorded at each adaptation step (optional).
T QueryAccuracy { get; } Gets the accuracy on the query set after adaptation.
T QueryLoss { get; } Gets the loss on the query set after adaptation.
T SupportAccuracy { get; } Gets the accuracy on the support set after adaptation.
T SupportLoss { get; } Gets the loss on the support set after adaptation.
double AdaptationTimeMs { get; } Gets the time taken for adaptation in milliseconds.
int AdaptationSteps { get; } Gets the number of gradient steps taken during adaptation.

Methods

Method Summary
T CalculateOverfittingGap() Calculates the overfitting gap between support and query performance.
bool DidConverge(double convergenceThreshold) Checks if the adaptation converged based on loss reduction.
string GenerateReport() Generates a formatted summary of adaptation results.