LAMP-SCOUT / tokenizer_config.json
palusi's picture
Upload 12 files
fd97629 verified
{
"add_prefix_space": true,
"added_tokens_decoder": {
"0": {
"content": "[PAD]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"100": {
"content": "[UNK]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"101": {
"content": "[CLS]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"102": {
"content": "[SEP]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"103": {
"content": "[MASK]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30522": {
"content": "[ENTER]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30523": {
"content": "[EXIT]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30524": {
"content": "[NEXT]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30525": {
"content": "[FINISH]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30526": {
"content": "[TASK]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30527": {
"content": "[MEMORY]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30528": {
"content": "[CURRENT PAGE]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30529": {
"content": "[TOTAL PAGES]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30530": {
"content": "[N FILLABLES]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30531": {
"content": "[N CLICKABLES]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30532": {
"content": "[CLICKABLES]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30533": {
"content": "[FILLABLES]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30534": {
"content": "[CONTEXT]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30535": {
"content": "[NEW]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30536": {
"content": "[EXPANSIONBEGIN]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30537": {
"content": "[EXPANSIONEND]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30538": {
"content": "[LABEL]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30539": {
"content": "[TEXT]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30540": {
"content": "[ARIA]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30541": {
"content": "[INFO]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30542": {
"content": "[PLACEHOLDER]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30543": {
"content": "[FILLED WITH]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30544": {
"content": "[CLICKABLE]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30545": {
"content": "[FILLABLE]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30546": {
"content": "[END]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30547": {
"content": "[CHECKBOX][UNCHECKED]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30548": {
"content": "[CHECKBOX][CHECKED]",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30549": {
"content": "- ",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30550": {
"content": "**Placeholder:** ",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30551": {
"content": "**Text:** ",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30552": {
"content": "**Aria-Label:** ",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30553": {
"content": "**Clickable**: ",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30554": {
"content": "**Fillable**: ",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"30555": {
"content": "**Context:** ",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
}
},
"clean_up_tokenization_spaces": false,
"cls_token": "[CLS]",
"do_lower_case": true,
"extra_special_tokens": {},
"mask_token": "[MASK]",
"model_max_length": 512,
"pad_token": "[PAD]",
"sep_token": "[SEP]",
"strip_accents": null,
"tokenize_chinese_chars": true,
"tokenizer_class": "DistilBertTokenizer",
"unk_token": "[UNK]"
}