langchain_text_splitters.base
.Tokenizer¶
- class langchain_text_splitters.base.Tokenizer(chunk_overlap: int, tokens_per_chunk: int, decode: Callable[[List[int]], str], encode: Callable[[str], List[int]])[source]¶
分词器数据类。
Attributes
chunk_overlap
在块之间存在重叠的标记
tokens_per_chunk
每个块的最大令牌数
decode
将token id列表解码为字符串的函数
encode
将字符串编码为标记ID列表的函数
Methods
__init__
(chunk_overlap, tokens_per_chunk, ...)- Parameters
chunk_overlap (int) –
tokens_per_chunk (int) –
decode (Callable[[List[int]], str]) –
encode (Callable[[str], List[int]]) –
- Return type
None
- __init__(chunk_overlap: int, tokens_per_chunk: int, decode: Callable[[List[int]], str], encode: Callable[[str], List[int]]) None ¶
- Parameters
chunk_overlap (int) –
tokens_per_chunk (int) –
decode (Callable[[List[int]], str]) –
encode (Callable[[str], List[int]]) –
- Return type
None