SelfSupervisedLearningConfig<T>
Options & Configuration · AiDotNet.SelfSupervisedLearning
Unified configuration for self-supervised learning with industry-standard defaults.
For Beginners
Self-supervised learning (SSL) learns useful representations from unlabeled data. This configuration controls how SSL pretraining works, including the method to use, batch size, training epochs, and method-specific settings.
How It Works
Key features:
Example - Simple usage with defaults:
Example - Custom configuration:
Constructors
| Constructor | Summary |
|---|---|
SelfSupervisedLearningConfig() |
Creates a new SSL configuration with industry-standard defaults. |
Properties
| Property | Summary |
|---|---|
BYOLConfig BYOL { get; set; } |
Gets or sets BYOL-specific configuration. |
BarlowTwinsConfig BarlowTwins { get; set; } |
Gets or sets Barlow Twins-specific configuration. |
DINOConfig DINO { get; set; } |
Gets or sets DINO-specific configuration. |
ISelfSupervisedLearningMethod<T> Method { get; set; } |
Gets or sets the SSL method to use. |
MAEConfig MAE { get; set; } |
Gets or sets MAE-specific configuration. |
MoCoConfig MoCo { get; set; } |
Gets or sets MoCo-specific configuration. |
SelfSupervisedLearningDistributedConfig Distributed { get; set; } |
Gets or sets the distributed training configuration. |
SelfSupervisedLearningOptimizerType? OptimizerType { get; set; } |
Gets or sets the optimizer type for SSL training. |
bool? EnableKNNEvaluation { get; set; } |
Gets or sets whether to run k-NN evaluation. |
bool? EnableLinearEvaluation { get; set; } |
Gets or sets whether to run linear evaluation during/after pretraining. |
bool? UseCosineDecay { get; set; } |
Gets or sets whether to use cosine learning rate decay. |
bool? UseTemperatureScheduling { get; set; } |
Gets or sets whether to use temperature scheduling. |
double? LearningRate { get; set; } |
Gets or sets the base learning rate. |
double? Temperature { get; set; } |
Gets or sets the temperature parameter for contrastive loss. |
double? WeightDecay { get; set; } |
Gets or sets the weight decay (L2 regularization). |
int? BatchSize { get; set; } |
Gets or sets the batch size for pretraining. |
int? CheckpointFrequency { get; set; } |
Gets or sets the checkpoint save frequency (epochs). |
int? KNNNeighbors { get; set; } |
Gets or sets the number of neighbors for k-NN evaluation. |
int? LinearEvaluationFrequency { get; set; } |
Gets or sets the frequency of linear evaluation (epochs). |
int? PretrainingEpochs { get; set; } |
Gets or sets the number of pretraining epochs. |
int? ProjectorHiddenDimension { get; set; } |
Gets or sets the hidden dimension of the projection head MLP. |
int? ProjectorLayers { get; set; } |
Gets or sets the number of layers in the projection head. |
int? ProjectorOutputDimension { get; set; } |
Gets or sets the output dimension of the projection head. |
int? Seed { get; set; } |
Gets or sets the random seed for reproducibility. |
int? WarmupEpochs { get; set; } |
Gets or sets the number of warmup epochs. |
string CheckpointPath { get; set; } |
Gets or sets the path to save checkpoints. |
Methods
| Method | Summary |
|---|---|
IDictionary<string, object> GetConfiguration() |
Gets the configuration as a dictionary for logging or serialization. |