Args: inputs: A 3-D tensor with shape of [batch, time, depth]. filters: An int. Number of outputs (=activation maps) size: An i

Applies batch normalization. Args: inputs: A tensor with 2 or more dimensions, where the first dimension has `batch_size`. If type

Applies a GRU. Args: inputs: A 3d tensor with shape of [N, T, C]. num_units: An int. The number of hidden units. bidirectio

Applies a series of conv1d separately. Args: inputs: A 3d tensor with shape of [N, T, C] K: An int. The size of conv1d banks. Tha

Highway networks, see https://arxiv.org/abs/1505.00387 Args: inputs: A 3D tensor of shape [N, T, W]. num_units: An int or `None`. Spe

spectrogram: [f, t]

Prenet for Encoder and Decoder1. Args: inputs: A 2D or 3D tensor. num_units: A list of two integers. or None. is_training: A pyt

Applies a GRU to `inputs`, while attending `memory`. Args: inputs: A 3d tensor with shape of [N, T', C']. Decoder inputs. memory: A 3d

Args: inputs: A 3d tensor with shape of [N, T_y/r, n_mels(*r)]. Shifted log melspectrogram of sound files. memory: A 3d tensor with s

Decoder Post-processing net = CBHG Args: inputs: A 3d tensor with shape of [N, T_y/r, n_mels*r]. Log magnitude spectrogram of sound files.

Embeds a given tensor. Args: inputs: A `Tensor` with type `int32` or `int64` containing the ids to be looked up in `lookup ta

Args: inputs: A 2d tensor with shape of [N, T_x, E], with dtype of int32. Encoder inputs. is_training: Whether or not the layer is in

Loads training data and put them in queues

Returns normalized log(melspectrogram) and log(magnitude) from `sound_file`. Args: sound_file: A string. The full path of a sound file.

Applies Griffin-Lim's raw.

Noam scheme from tensor2tensor

Plots the alignment alignments: A list of (numpy) matrix of shape (encoder_steps, decoder_steps) gs : (int) global step

# Generate wave file from spectrogram