Class

MLParameterKey

A key for model and update parameter dictionaries.

Declaration

class MLParameterKey : MLKey

Topics

Model Parameter Keys

class var linkedModelFileName: MLParameterKey

The key you use to access the name of the linked model.

class var linkedModelSearchPath: MLParameterKey

The key you use to access the search path for the linked model.

Fixed Update Parameter Keys

class var epochs: MLParameterKey

The key you use to access the epochs parameter for the optimizer (an Int64 value).

class var miniBatchSize: MLParameterKey

The key you use to access the mini batch-size parameter for the optimizer (an Int64 value).

Dynamic Update Parameter Keys

class var beta1: MLParameterKey

The key you use to access the beta1 parameter for the Adam optimizer (a Double value).

class var beta2: MLParameterKey

The key you use to access the beta2 parameter for the Adam optimizer (a Double value).

class var eps: MLParameterKey

The key you use to access the eps parameter for the Adam optimizer (a Double value).

class var learningRate: MLParameterKey

The key you use to access the parameter for the optimizer’s learning rate (a Double value).

class var momentum: MLParameterKey

The key you use to access the parameter for the optimizer’s momentum gradient (a Double value).

Supporting Types

class MLKey

An abstract base class for machine learning key types.

Relationships

Inherits From

Conforms To

See Also

Starting and Resuming an Update

func resume(withParameters: [MLParameterKey : Any])

Resumes a model update with updated parameter values.