espnet2.legacy.nets.pytorch_backend.transducer.utils.select_k_expansions

source

espnet2.legacy.nets.pytorch_backend.transducer.utils.select_k_expansions(hyps: List[ExtendedHypothesis], topk_idxs: Tensor, topk_logps: Tensor, gamma: float) → List[ExtendedHypothesis]

Return K hypotheses candidates for expansion from a list of hypothesis.

K candidates are selected according to the extended hypotheses probabilities and a prune-by-value method. Where K is equal to beam_size + beta.

• Parameters:
  • hyps – Hypotheses.
  • topk_idxs – Indices of candidates hypothesis.
  • topk_logps – Log-probabilities for hypotheses expansions.
  • gamma – Allowed logp difference for prune-by-value method.
• Returns: Best K expansion hypotheses candidates.
• Return type: k_expansions