LSTMCell

public struct LSTMCell<Scalar> : RecurrentLayerCell where Scalar : TensorFlowFloatingPoint

An LSTM cell.

• fusedWeight

  Declaration

  public var fusedWeight: Tensor<Scalar>

• fusedBias

  Declaration

  public var fusedBias: Tensor<Scalar>

• inputWeight

  Declaration

  public var inputWeight: Tensor<Scalar> { get }

• updateWeight

  Declaration

  public var updateWeight: Tensor<Scalar> { get }

• forgetWeight

  Declaration

  public var forgetWeight: Tensor<Scalar> { get }

• outputWeight

  Declaration

  public var outputWeight: Tensor<Scalar> { get }

• inputBias

  Declaration

  public var inputBias: Tensor<Scalar> { get }

• updateBias

  Declaration

  public var updateBias: Tensor<Scalar> { get }

• forgetBias

  Declaration

  public var forgetBias: Tensor<Scalar> { get }

• outputBias

  Declaration

  public var outputBias: Tensor<Scalar> { get }

• TimeStepInput

  Declaration

  public typealias TimeStepInput = Tensor<Scalar>

• TimeStepOutput

  Declaration

  public typealias TimeStepOutput = State

• Input

  Declaration

  public typealias Input = RNNCellInput<TimeStepInput, State>

• Output

  Declaration

  public typealias Output = RNNCellOutput<TimeStepOutput, State>

• init(inputSize:hiddenSize:)

  Creates a LSTMCell with the specified input size and hidden state size.

  Declaration

  public init(inputSize: Int, hiddenSize: Int)

  Parameters

  inputSize	The number of features in 2-D input tensors.
  hiddenSize	The number of features in 2-D hidden states.

• State

  Declaration

  public struct State : Equatable, Differentiable, VectorProtocol, KeyPathIterable, Mergeable

• zeroState(for:)

  Returns a zero-valued state with shape compatible with the provided input.

  Declaration

  public func zeroState(for input: Tensor<Scalar>) -> State

• callAsFunction(_:)

  Returns the output obtained from applying the layer to the given input.

  Declaration

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

  Parameters

  input

  The input to the layer.

  Return Value

  The hidden state.