bert-base-spanish-wwm-cased-cantemist / tokenizer_config.json
GuillemGSubies's picture
Upload the tokenizer
f8d624c
raw
history blame
661 Bytes
{
"cls_token": "[CLS]",
"do_basic_tokenize": true,
"do_lower_case": false,
"keep_accents": true,
"mask_token": "[MASK]",
"model_max_len": 512,
"model_max_length": 512,
"name_or_path": "dccuchile/bert-base-spanish-wwm-cased-64-4e-05-0.1-cantemist_best",
"never_split": null,
"pad_token": "[PAD]",
"sep_token": "[SEP]",
"special_tokens_map_file": "/home/guillem.garcia/.cache/huggingface/hub/models--dccuchile--bert-base-spanish-wwm-cased/snapshots/56a7647b957a4230fc3f80dafbe80f2ba9b0de73/special_tokens_map.json",
"strip_accents": false,
"tokenize_chinese_chars": true,
"tokenizer_class": "BertTokenizer",
"unk_token": "[UNK]"
}