SymbolicSgdLogisticRegressionBinaryTrainer.Options.Tolerance Field

  • Reference

Definition

Namespace:
Microsoft.ML.Trainers
Assembly:
Microsoft.ML.Mkl.Components.dll
Package:
Microsoft.ML.Mkl.Components v3.0.1
Package:
Microsoft.ML.Mkl.Components v1.0.0
Package:
Microsoft.ML.Mkl.Components v1.1.0
Package:
Microsoft.ML.Mkl.Components v1.2.0
Package:
Microsoft.ML.Mkl.Components v1.3.1
Package:
Microsoft.ML.Mkl.Components v1.4.0
Package:
Microsoft.ML.Mkl.Components v1.5.5
Package:
Microsoft.ML.Mkl.Components v1.6.0
Package:
Microsoft.ML.Mkl.Components v1.7.0
Package:
Microsoft.ML.Mkl.Components v2.0.0

Tolerance for difference in average loss in consecutive passes. If the reduction on loss is smaller than the specified tolerance in one iteration, the training process will be terminated.

public float Tolerance;
val mutable Tolerance : single
Public Tolerance As Single

Field Value

Single

Applies to