MLObjectDetector.ModelParameters
Parameters that affect the process of training an object detection model.
Declaration
struct ModelParametersOverview
Customize the training process of an object detector by creating an MLObjectDetector.ModelParameters instance and passing it to an object detector’s initializer. You can explicitly set values for maxIterations and batchSize. You can also explicitly define the validation dataset to override the default behavior, which uses a random selection of your training dataset for validation.
Topics
Creating object detector parameters
init(validation:batchSize:maxIterations:)init(validation:batchSize:maxIterations:gridSize:algorithm:)init(validationData:batchSize:maxIterations:)
Accessing the training parameters
Describing the model parameters
Supporting types
MLObjectDetector.ModelParameters.ValidationDataMLObjectDetector.ModelParameters.ModelAlgorithmTypeMLObjectDetector.ModelParameters.FeatureExtractorType