JMPAlgorithm<T, TInput, TOutput>
Models & Types · AiDotNet.MetaLearning.Algorithms
Implementation of JMP: Joint Multi-Phase meta-learning.
Inheritance: ModelBase<T, TInput, TOutput> → MetaLearnerBase<T, TInput, TOutput> → JMPAlgorithm<T, TInput, TOutput>
How It Works
JMP divides the inner loop into two phases with distinct optimization characteristics. Phase 1 (coarse) uses a higher learning rate for rapid, broad adaptation. Phase 2 (fine) uses a lower learning rate with L2 regularization toward the Phase 1 result, enabling careful refinement without deviating too far from the coarse solution.
Algorithm:
Constructors
| Constructor | Summary |
|---|---|
JMPAlgorithm(JMPOptions<T, TInput, TOutput> options) |
Properties
| Property | Summary |
|---|---|
MetaLearningAlgorithmType AlgorithmType { get; } |
Methods
| Method | Summary |
|---|---|
IModel<TInput, TOutput, ModelMetadata<T>> Adapt(IMetaLearningTask<T, TInput, TOutput> task) |
|
T MetaTrain(TaskBatch<T, TInput, TOutput> taskBatch) |