espnet.nets.pytorch_backend.lm.seq_rnn.SequentialRNNLM

Less than 1 minute

class espnet.nets.pytorch_backend.lm.seq_rnn.SequentialRNNLM(n_vocab, args)

Bases: LMInterface, Module

Sequential RNNLM.

Initialize class.

Parameters:
- n_vocab (int) – The size of the vocabulary
- args (argparse.Namespace) – configurations. see py:method:add_arguments

static add_arguments(parser)

Add arguments to command line argument parser.

forward(x, t)

Compute LM loss value from buffer sequences.

Parameters:
- x (torch.Tensor) – Input ids. (batch, len)
- t (torch.Tensor) – Target ids. (batch, len)
Returns: Tuple of : loss to backward (scalar), negative log-likelihood of t: -log p(t) (scalar) and the number of elements in x (scalar)
Return type: tuple[torch.Tensor, torch.Tensor, torch.Tensor]

Notes

The last two return values are used in perplexity: p(t)^{-n} = exp(-log p(t) / n)

init_state(x)

Get an initial state for decoding.

Returns: initial state

score(y, state, x)

Score new token.

Parameters:
- y (torch.Tensor) – 1D torch.int64 prefix tokens.
- state – Scorer state for prefix tokens
- x (torch.Tensor) – 2D encoder feature that generates ys.
Returns: Tuple of : torch.float32 scores for next token (n_vocab) and next state for ys
Return type: tuple[torch.Tensor, Any]