Options
All
  • Public
  • Public/Protected
  • All
Menu

Class ConductorDecoder

Hierarchical decoder that produces intermediate embeddings to pass to lower-level Decoder objects. The outputs from different decoders are concatenated depth-wise (axis 3), and the outputs from different steps of the conductor are concatenated across time (axis 1).

Hierarchy

Index

Constructors

constructor

  • Decoder contructor.

    Parameters

    • coreDecoders: Decoder[]

      Lower-level Decoder objects to pass the conductor LSTM output embeddings to for futher decoding.

    • lstmCellVars: LayerVars[]

      The LayerVars for each layer of the conductor LSTM.

    • zToInitStateVars: LayerVars

      The LayerVars for projecting from the latent variable z to the initial states of the conductor LSTM layers.

    • numSteps: number

      The number of embeddings the conductor LSTM should produce and pass to the lower-level decoder.

    Returns ConductorDecoder

Properties

coreDecoders

coreDecoders: Decoder[]

lstmCellVars

lstmCellVars: LayerVars[]

numSteps

numSteps: number

outputDims

outputDims: number

zDims

zDims: number

zToInitStateVars

zToInitStateVars: LayerVars

Methods

decode

  • decode(z: tf.Tensor2D, length: number, initialInput?: tf.Tensor2D, temperature?: number, controls?: tf.Tensor2D): Tensor<R3>
  • Hierarchically decodes a batch of latent vectors, z.

    Parameters

    • z: tf.Tensor2D

      A batch of latent vectors to decode, sized [batchSize, zDims].

    • length: number

      The length of decoded sequences.

    • Optional initialInput: tf.Tensor2D
    • Optional temperature: number

      (Optional) The softmax temperature to use when sampling from the logits. Argmax is used if not provided.

    • Optional controls: tf.Tensor2D

      (Optional) Control tensors to use for conditioning, sized [length, controlDepth].

    Returns Tensor<R3>

    A boolean tensor containing the decoded sequences, shaped [batchSize, length, depth].

Generated using TypeDoc