lenu_SE / tokenizer_config.json
aarimond's picture
added the tokenizer
b9bb562
raw
history blame
576 Bytes
{
"cls_token": "[CLS]",
"do_basic_tokenize": true,
"do_lower_case": false,
"mask_token": "[MASK]",
"model_max_length": 1000000000000000019884624838656,
"name_or_path": "KB/bert-base-swedish-cased",
"never_split": null,
"pad_token": "[PAD]",
"sep_token": "[SEP]",
"special_tokens_map_file": "/root/.cache/huggingface/hub/models--KB--bert-base-swedish-cased/snapshots/81c7baa04742a30cb6732c181e678721868cb42e/special_tokens_map.json",
"strip_accents": false,
"tokenize_chinese_chars": true,
"tokenizer_class": "BertTokenizer",
"unk_token": "[UNK]"
}