args: padded_input: B x T x D input_lengths: B padded_target: B x T output: bat

args: padded_input: B x T encoder_padded_outputs: B x T x H encoder_input_lengths: B returns

args: inputs: B x H output: predictions: B x C

args: padded_input: B x T x D input_lengths: B return: output: B x T x H

args: input: B x T x D output: tensor: B x T

args: x: tensor output: y: tensor

args: x: tensor output: y: tensor

query: B x T_Q x H, key: B x T_K x H, value: B x T_V x H mask: B x T x T (attention mask)

query: B x T_Q x H, key: B x T_K x H, value: B x T_V x H mask: B x T x T (attention mask)

Args: queries: N x To x H values : N x Ti x H Returns: output: N x To x H atten

args: padded_input: B x 1 (channel for spectrogram=1) x (freq) x T padded_input: B x T x D input_leng

Get word segments args: seq: String output: word_segments: list of String

Load stanford core NLP toolkit object args: path: String output: Stanford core NLP objects

Args: logits: A Variable containing a FloatTensor of size (batch, max_len, num_classes) which contains the un

:param transcript_path: Path where transcript is stored from the manifest file :return: Transcript in training/testing format

Takes positive (target) samples, negative (generator) samples and prepares inp and target data for discriminator. Inputs: pos_samples, ne

Preprocess function args: seq: String output: seq: String

Recursively iterate directories and read all the data

Recursively iterate phase 1 directories and read all the data

Remove emojis args: seq: String output: seq: String

Remove return characters args: seq: String output: seq: String

Training args: model: Model object train_data_list: DataLoader object of the training set va

Training args: model: Model object train_data_list: DataLoader object of the training set va

Training args: model: Model object train_data_list: DataLoader object of the training set va

Training args: model: Model object train_data_list: DataLoader object of the training set va