28 lines
662 B
JSON
28 lines
662 B
JSON
|
{
|
||
|
|
"backend": "tokenizers",
|
||
|
|
"bos_token": "<s>",
|
||
|
|
"clean_up_tokenization_spaces": false,
|
||
|
|
"eos_token": "</s>",
|
||
|
|
"extra_special_tokens": [
|
||
|
|
"<unk>",
|
||
|
|
"<s>",
|
||
|
|
"</s>"
|
||
|
|
],
|
||
|
|
"is_local": false,
|
||
|
|
"legacy": true,
|
||
|
|
"max_length": 512,
|
||
|
|
"model_max_length": 1000000000000000019884624838656,
|
||
|
|
"pad_to_multiple_of": null,
|
||
|
|
"pad_token": "</s>",
|
||
|
|
"pad_token_type_id": 0,
|
||
|
|
"padding_side": "right",
|
||
|
|
"sp_model_kwargs": {},
|
||
|
|
"spaces_between_special_tokens": false,
|
||
|
|
"stride": 0,
|
||
|
|
"tokenizer_class": "TokenizersBackend",
|
||
|
|
"truncation_side": "right",
|
||
|
|
"truncation_strategy": "longest_first",
|
||
|
|
"unk_token": "<unk>",
|
||
|
|
"use_default_system_prompt": true
|
||
|
|
}
|