순차

public struct Sequential<Layer1: Module, Layer2: Layer>: Module
    where Layer1.Output == Layer2.Input,
          Layer1.TangentVector.VectorSpaceScalar == Layer2.TangentVector.VectorSpaceScalar

extension Sequential: Layer where Layer1: Layer

두 개 이상의 다른 레이어를 순차적으로 구성하는 레이어입니다.

예:

• MNIST를 위한 간단한 2계층 퍼셉트론 모델을 구축합니다.

let inputSize = 28 * 28
let hiddenSize = 300
var classifier = Sequential {
     Dense<Float>(inputSize: inputSize, outputSize: hiddenSize, activation: relu)
     Dense<Float>(inputSize: hiddenSize, outputSize: 3, activation: identity)
 }

• MNIST용 자동 인코더를 빌드합니다.

var autoencoder = Sequential {
    // The encoder.
    Dense<Float>(inputSize: 28 * 28, outputSize: 128, activation: relu)
    Dense<Float>(inputSize: 128, outputSize: 64, activation: relu)
    Dense<Float>(inputSize: 64, outputSize: 12, activation: relu)
    Dense<Float>(inputSize: 12, outputSize: 3, activation: relu)
    // The decoder.
    Dense<Float>(inputSize: 3, outputSize: 12, activation: relu)
    Dense<Float>(inputSize: 12, outputSize: 64, activation: relu)
    Dense<Float>(inputSize: 64, outputSize: 128, activation: relu)
    Dense<Float>(inputSize: 128, outputSize: imageHeight * imageWidth, activation: tanh)
}

• 레이어1
  선언

  public var layer1: Layer1

• 레이어2
  선언

  public var layer2: Layer2

• 초기화(_:_:)
  선언

  public init(_ layer1: Layer1, _ layer2: Layer2)

• callAsFunction(_:)
  선언

  @differentiable(wrt: self)
  public func callAsFunction(_ input: Layer1.Input) -> Layer2.Output

• 초기화(레이어:)
  선언

  public init(@LayerBuilder layers: () -> `Self`)

'Layer1': 'Layer'에서 사용 가능

• callAsFunction(_:)
  선언

  @differentiable
  public func callAsFunction(_ input: Layer1.Input) -> Layer2.Output