espnet.nets.mt_interface.MTInterface

source

class espnet.nets.mt_interface.MTInterface

Bases: object

MT Interface for ESPnet model implementation.

static add_arguments(parser)

Add arguments to parser.

property attention_plot_class

Get attention plot class.

classmethod build(idim: int, odim: int, **kwargs)

Initialize this class with python-level args.

• Parameters:
  • idim (int) – The number of an input feature dim.
  • odim (int) – The number of output vocab.
• Returns: A new instance of ASRInterface.
• Return type: ASRinterface

calculate_all_attentions(xs, ilens, ys)

Calculate attention.

• Parameters:
  • xs (list) – list of padded input sequences [(T1, idim), (T2, idim), …]
  • ilens (ndarray) – batch of lengths of input sequences (B)
  • ys (list) – list of character id sequence tensor [(L1), (L2), (L3), …]
• Returns: attention weights (B, Lmax, Tmax)
• Return type: float ndarray

forward(xs, ilens, ys)

Compute loss for training.

• Parameters:
  • xs – For pytorch, batch of padded source sequences torch.Tensor (B, Tmax, idim) For chainer, list of source sequences chainer.Variable
  • ilens – batch of lengths of source sequences (B) For pytorch, torch.Tensor For chainer, list of int
  • ys – For pytorch, batch of padded source sequences torch.Tensor (B, Lmax) For chainer, list of source sequences chainer.Variable
• Returns: loss value
• Return type: torch.Tensor for pytorch, chainer.Variable for chainer

translate(x, trans_args, char_list=None, rnnlm=None)

Translate x for evaluation.

• Parameters:
  • x (ndarray) – input acouctic feature (B, T, D) or (T, D)
  • trans_args (namespace) – argment namespace contraining options
  • char_list (list) – list of characters
  • rnnlm (torch.nn.Module) – language model module
• Returns: N-best decoding results
• Return type: list

translate_batch(x, trans_args, char_list=None, rnnlm=None)

Beam search implementation for batch.

• Parameters:
  • x (torch.Tensor) – encoder hidden state sequences (B, Tmax, Henc)
  • trans_args (namespace) – argument namespace containing options
  • char_list (list) – list of characters
  • rnnlm (torch.nn.Module) – language model module
• Returns: N-best decoding results
• Return type: list