OnlineGradientDescentTrainer.Options Class
Definition
Important
Some information relates to prerelease product that may be substantially modified before it’s released. Microsoft makes no warranties, express or implied, with respect to the information provided here.
Options for the OnlineGradientDescentTrainer as used in OnlineGradientDescent(Options).
public sealed class OnlineGradientDescentTrainer.Options : Microsoft.ML.Trainers.AveragedLinearOptionstype OnlineGradientDescentTrainer.Options = class
inherit AveragedLinearOptionsPublic NotInheritable Class OnlineGradientDescentTrainer.Options
Inherits AveragedLinearOptions- Inheritance
-
OnlineGradientDescentTrainer.Options
Constructors
| OnlineGradientDescentTrainer.Options() |
Create a new OnlineGradientDescentTrainer.Options object with default values. |
Fields
| Averaged |
Determines whether to do averaging or not. (Inherited from AveragedLinearOptions) |
| DecreaseLearningRate |
Determine whether to decrease the LearningRate or not. (Inherited from AveragedLinearOptions) |
| FeatureColumnName |
Column to use for features. (Inherited from TrainerInputBase) |
| InitialWeightsDiameter |
Initial weights and bias scale. (Inherited from OnlineLinearOptions) |
| L2Regularization |
The L2 weight for regularization. (Inherited from AveragedLinearOptions) |
| LabelColumnName |
Column to use for labels. (Inherited from TrainerInputBaseWithLabel) |
| LazyUpdate |
Determines when to update averaged weights. (Inherited from AveragedLinearOptions) |
| LearningRate |
Learning rate. (Inherited from AveragedLinearOptions) |
| NumberOfIterations |
Number of passes through the training dataset. (Inherited from OnlineLinearOptions) |
| RecencyGain |
Extra weight given to more recent updates. (Inherited from AveragedLinearOptions) |
| RecencyGainMultiplicative |
Determines whether RecencyGain is multiplicative or additive. (Inherited from AveragedLinearOptions) |
| ResetWeightsAfterXExamples |
Number of examples after which weights will be reset to the current average. (Inherited from AveragedLinearOptions) |
| Shuffle |
Determines whether to shuffle data for each training iteration. (Inherited from OnlineLinearOptions) |
Properties
| LossFunction |
A custom loss. |