LegalBERT_GDPR / tokenizer_config.json
AndreaSimeri's picture
Upload 6 files
105f60f verified
raw
history blame contribute delete
561 Bytes
{
"cls_token": "[CLS]",
"do_basic_tokenize": true,
"do_lower_case": true,
"mask_token": "[MASK]",
"model_max_length": 512,
"name_or_path": "nlpaueb/legal-bert-base-uncased",
"never_split": null,
"pad_token": "[PAD]",
"sep_token": "[SEP]",
"special_tokens_map_file": "/home/user/.cache/huggingface/hub/models--nlpaueb--legal-bert-base-uncased/snapshots/15b570cbf88259610b082a167dacc190124f60f6/special_tokens_map.json",
"strip_accents": null,
"tokenize_chinese_chars": true,
"tokenizer_class": "BertTokenizer",
"unk_token": "[UNK]"
}