---
title: "init(learningRate:gradientRescale:appliesGradientClipping:gradientClipMax:gradientClipMin:regularizationType:regularizationScale:)"
framework: ML Compute
role: symbol
role_heading: Initializer
platforms: [iOS 14.0+, iPadOS 14.0+, Mac Catalyst 14.0+, macOS 11.0+, tvOS 14.0+]
path: "mlcompute/mlcoptimizerdescriptor/init(learningrate:gradientrescale:appliesgradientclipping:gradientclipmax:gradientclipmin:regularizationtype:regularizationscale:)"
---

# init(learningRate:gradientRescale:appliesGradientClipping:gradientClipMax:gradientClipMin:regularizationType:regularizationScale:)

Creates a descriptor with the learning rate, gradient rescale, clipping option and values, and regularization type and scale that you specify.

## Declaration

```swift
convenience init(learningRate: Float, gradientRescale: Float, appliesGradientClipping: Bool, gradientClipMax: Float, gradientClipMin: Float, regularizationType: MLCRegularizationType, regularizationScale: Float)
```

## Parameters

- `learningRate`: The learning rate.
- `gradientRescale`: The gradient rescale value.
- `appliesGradientClipping`: A Boolean value that indicates whether you apply gradient clipping.
- `gradientClipMax`: The maximum gradient value before the optimizer rescales the gradient, if you enable gradient clipping.
- `gradientClipMin`: The minimum gradient value before the optimizer rescales the gradient, if you enable gradient clipping.
- `regularizationType`: The regularization type.
- `regularizationScale`: The regularization scale.

## See Also

### Creating an Optimizer Descriptor

- [init(learningRate:gradientRescale:regularizationType:regularizationScale:)](mlcompute/mlcoptimizerdescriptor/init(learningrate:gradientrescale:regularizationtype:regularizationscale:).md)
- [init(learningRate:gradientRescale:appliesGradientClipping:gradientClippingType:gradientClipMax:gradientClipMin:maximumClippingNorm:customGlobalNorm:regularizationType:regularizationScale:)](mlcompute/mlcoptimizerdescriptor/init(learningrate:gradientrescale:appliesgradientclipping:gradientclippingtype:gradientclipmax:gradientclipmin:maximumclippingnorm:customglobalnorm:regularizationtype:regularizationscale:).md)