funasr2

游雁

2023-12-19 00ea1186f96e6732e2edb4fab6c0ed6896e3b352

 funasr/tokenizer/char_tokenizer.py

@@ -4,17 +4,19 @@
from typing import Union
import warnings

from funasr.tokenizer.abs_tokenizer import BaseTokenizer
from funasr.utils.register import register_class

from funasr.tokenizer.abs_tokenizer import AbsTokenizer


class CharTokenizer(AbsTokenizer):
@register_class("tokenizer_classes", "CharTokenizer")
class CharTokenizer(BaseTokenizer):
    def __init__(
        self,
        non_linguistic_symbols: Union[Path, str, Iterable[str]] = None,
        space_symbol: str = "<space>",
        remove_non_linguistic_symbols: bool = False,
        **kwargs,
    ):
        super().__init__(**kwargs)
        self.space_symbol = space_symbol
        if non_linguistic_symbols is None:
            self.non_linguistic_symbols = set()