TFGENZOO.layers.flowtts_coupling module¶

TFGENZOO.layers.flowtts_coupling.CouplingBlock(x: tensorflow.python.framework.ops.Tensor, cond: tensorflow.python.framework.ops.Tensor, depth, **kwargs)[source]¶

Parameters

x (tf.Tensor) – input contents tensor [B, T, C]
c (tf.Tensor) – input conditional tensor [B, T, C’] where C’ can be different with C

Returns

CouplingBlock: reference: Flow-TTS

Return type

tf.keras.Model

Examples

>>> import tensorflow as tf
>>> from utils.coupling_block import CouplingBlock
>>> x = tf.keras.layers.Input([None, 32])
>>> c = tf.keras.layers.Input([None, 128])
>>> cp = CouplingBlock(x, c, depth=256)
>>> cp.summary()
Model: "model"
__________________________________________________________________________________________________
Layer (type)                    Output Shape         Param #     Connected to
==================================================================================================
input_1 (InputLayer)            [(None, None, 32)]   0
__________________________________________________________________________________________________
conv1d (Conv1D)                 (None, None, 256)    8192        input_1[0][0]
__________________________________________________________________________________________________
input_2 (InputLayer)            [(None, None, 128)]  0
__________________________________________________________________________________________________
gtu (GTU)                       (None, None, 256)    98816       conv1d[0][0]
__________________________________________________________________________________________________
tf_op_layer_AddV2 (TensorFlowOp [(None, None, 256)]  0           gtu[0][0]
                                                                conv1d[0][0]
__________________________________________________________________________________________________
conv1d_1 (Conv1D)               (None, None, 64)     16448       tf_op_layer_AddV2[0][0]
==================================================================================================
Total params: 123,456
Trainable params: 123,456
Non-trainable params: 0
__________________________________________________________________________________________________
>>> cp([x, c])
<tf.Tensor 'model_1/Identity:0' shape=(None, None, 64) dtype=float32>

class TFGENZOO.layers.flowtts_coupling.GTU(**kwargs)[source]¶

Bases: tensorflow.python.keras.engine.base_layer.Layer

GTU layer proposed in Flow-TTS

Note

formula
\[z = tanh(W_{f, k} \star y) \odot sigmoid(W_{g, k} \star c)\]

build(input_shape: tensorflow.python.framework.tensor_shape.TensorShape)[source]¶

Creates the variables of the layer (optional, for subclass implementers).

This is a method that implementers of subclasses of Layer or Model can override if they need a state-creation step in-between layer instantiation and layer call.

This is typically used to create the weights of Layer subclasses.

Parameters: input_shape – Instance of TensorShape, or list of instances of TensorShape if the layer expects a list of inputs (one instance per input).

call(y: tensorflow.python.framework.ops.Tensor, c: tensorflow.python.framework.ops.Tensor, **kwargs)[source]¶

Parameters

y (tf.Tensor) – input contents tensor [B, T, C]
c (tf.Tensor) – input conditional tensor [B, T, C’] where C’ can be different with C

Returns

[B, T, C]

Return type

tf.Tensor

get_config()[source]¶

Returns the config of the layer.

A layer config is a Python dictionary (serializable) containing the configuration of a layer. The same layer can be reinstantiated later (without its trained weights) from this configuration.

The config of a layer does not include connectivity information, nor the layer class name. These are handled by Network (one layer of abstraction above).

Returns: Python dictionary.