# BNNS.AdamOptimizer An optimizer that uses the Adam optimization algorithm. ``` struct AdamOptimizer ``` ## Topics ### Creating an Adam Optimizer [`init(learningRate: Float, beta1: Float, beta2: Float, timeStep: Float, epsilon: Float, gradientScale: Float, regularizationScale: Float, clipsGradientsTo: ClosedRange?, regularizationFunction: BNNSOptimizerRegularizationFunction)`](/documentation/Accelerate/BNNS/AdamOptimizer/init(learningRate:beta1:beta2:timeStep:epsilon:gradientScale:regularizationScale:clipsGradientsTo:regularizationFunction:)) Returns a new Adam optimizer object. [`init(learningRate: Float, beta1: Float, beta2: Float, timeStep: Float, epsilon: Float, gradientScale: Float, regularizationScale: Float, gradientClipping: BNNS.GradientClipping, regularizationFunction: BNNSOptimizerRegularizationFunction, usesAMSGrad: Bool)`](/documentation/Accelerate/BNNS/AdamOptimizer/init(learningRate:beta1:beta2:timeStep:epsilon:gradientScale:regularizationScale:gradientClipping:regularizationFunction:usesAMSGrad:)) Returns a new Adam optimizer object with gradient clipped by value or clipped by norm. ### Inspecting the Properties of an Adam Optimizer [`var learningRate: Float`](/documentation/Accelerate/BNNS/AdamOptimizer/learningRate) A value that specifies the learning rate. [`var beta1: Float`](/documentation/Accelerate/BNNS/AdamOptimizer/beta1) A value that specifies the first-moment constant, in the range `0` to `1`. [`var beta2: Float`](/documentation/Accelerate/BNNS/AdamOptimizer/beta2) A value that specifies the second-moment constant, in the range `0` to `1`. [`var timeStep: Float`](/documentation/Accelerate/BNNS/AdamOptimizer/timeStep) A value that’s at least `1` and represents the optimizer’s current time. [`var epsilon: Float`](/documentation/Accelerate/BNNS/AdamOptimizer/epsilon) The epsilon value you use to improve numerical stability. [`var gradientScale: Float`](/documentation/Accelerate/BNNS/AdamOptimizer/gradientScale) A value that specifies the gradient scaling factor. [`var regularizationScale: Float`](/documentation/Accelerate/BNNS/AdamOptimizer/regularizationScale) A value that specifies the regularization scaling factor. [`var gradientBounds: ClosedRange?`](/documentation/Accelerate/BNNS/AdamOptimizer/gradientBounds) The values for the minimum and maximum gradients. [`var gradientClipping: BNNS.GradientClipping`](/documentation/Accelerate/BNNS/AdamOptimizer/gradientClipping) The gradient clipping. [`var regularizationFunction: BNNSOptimizerRegularizationFunction`](/documentation/Accelerate/BNNS/AdamOptimizer/regularizationFunction) The variable that specifies the regularization function. [`var usesAMSGrad: Bool`](/documentation/Accelerate/BNNS/AdamOptimizer/usesAMSGrad) A Boolean value that specifies whether to use the AMSGrad variant. [`var accumulatorCountMultiplier: Int`](/documentation/Accelerate/BNNS/AdamOptimizer/accumulatorCountMultiplier) The number of accumulators required for each parameter. [`enum GradientClipping`](/documentation/Accelerate/BNNS/GradientClipping) Constants that describe clipping functions. ### Applying an Optimization Step ## Relationships ### Conforms To [`BNNSOptimizer`](/documentation/Accelerate/BNNSOptimizer) --- Copyright © 2026 Apple Inc. All rights reserved. | [Terms of Use](https://www.apple.com/legal/internet-services/terms/site.html) | [Privacy Policy](https://www.apple.com/privacy/privacy-policy)