espnet2.text.whisper_token_id_converter.OpenAIWhisperTokenIDConverter

source

class espnet2.text.whisper_token_id_converter.OpenAIWhisperTokenIDConverter(model_type: str, language: str | None = 'en', task: str = 'transcribe', added_tokens_txt: str | None = None, sot: bool = False, speaker_change_symbol: str = '<sc>')

Bases: object

get_num_vocabulary_size() → int

ids2tokens(integers: ndarray | Iterable[int]) → List[str]

tokens2ids(tokens: Iterable[str]) → List[int]