ConfliBERT-Spanish-Beto-Cased-v1 / tokenizer_config.json
salsarra's picture
Upload 13 files
9380076 verified
{
"cls_token": "[CLS]",
"do_basic_tokenize": true,
"do_lower_case": false,
"mask_token": "[MASK]",
"model_max_length": 512,
"name_or_path": "../models/bert-base-spanish-wwm-cased",
"never_split": null,
"pad_token": "[PAD]",
"sep_token": "[SEP]",
"special_tokens_map_file": "../models/bert-base-spanish-wwm-cased/special_tokens_map.json",
"strip_accents": false,
"tokenize_chinese_chars": true,
"tokenizer_class": "BertTokenizer",
"unk_token": "[UNK]"
}