bert-beatrix-2048 / tokenizer_config.json
AbstractPhil's picture
Update tokenizer_config.json
a7fb0f6 verified
{
"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": "<subject>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30523": {
"content": "<pose>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30524": {
"content": "<emotion>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30525": {
"content": "<surface>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30526": {
"content": "<lighting>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30527": {
"content": "<material>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30528": {
"content": "<accessory>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30529": {
"content": "<footwear>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30530": {
"content": "<upper_body_clothing>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30531": {
"content": "<hair_style>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30532": {
"content": "<hair_length>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30533": {
"content": "<headwear>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30534": {
"content": "<texture>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30535": {
"content": "<pattern>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30536": {
"content": "<grid>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30537": {
"content": "<zone>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30538": {
"content": "<offset>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30539": {
"content": "<object_left>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30540": {
"content": "<object_right>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30541": {
"content": "<relation>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30542": {
"content": "<intent>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30543": {
"content": "<style>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30544": {
"content": "<fabric>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30545": {
"content": "<jewelry>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30546": {
"content": "<subject1>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30547": {
"content": "<subject2>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30548": {
"content": "[SHUNT_1000000]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30549": {
"content": "[SHUNT_1000001]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30550": {
"content": "[SHUNT_1000002]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30551": {
"content": "[SHUNT_1000003]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30552": {
"content": "[SHUNT_1000004]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30553": {
"content": "[SHUNT_1000005]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30554": {
"content": "[SHUNT_1000006]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30555": {
"content": "[SHUNT_1000007]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30556": {
"content": "[SHUNT_1000008]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30557": {
"content": "[SHUNT_1000009]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30558": {
"content": "[SHUNT_1000010]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30559": {
"content": "[SHUNT_1000011]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30560": {
"content": "[SHUNT_1000012]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30561": {
"content": "[SHUNT_1000013]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30562": {
"content": "[SHUNT_1000014]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30563": {
"content": "[SHUNT_1000015]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30564": {
"content": "[SHUNT_1000016]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30565": {
"content": "[SHUNT_1000017]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30566": {
"content": "[SHUNT_1000018]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30567": {
"content": "[SHUNT_1000019]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30568": {
"content": "[SHUNT_1000020]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30569": {
"content": "[SHUNT_1000021]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30570": {
"content": "[SHUNT_1000022]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30571": {
"content": "[SHUNT_1000023]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30572": {
"content": "[SHUNT_1000024]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"30573": {
"content": "[SHUNT_1000025]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
}
},
"additional_special_tokens": [
"<subject>",
"<pose>",
"<emotion>",
"<surface>",
"<lighting>",
"<material>",
"<accessory>",
"<footwear>",
"<upper_body_clothing>",
"<hair_style>",
"<hair_length>",
"<headwear>",
"<texture>",
"<pattern>",
"<grid>",
"<zone>",
"<offset>",
"<object_left>",
"<object_right>",
"<relation>",
"<intent>",
"<style>",
"<fabric>",
"<jewelry>",
"<subject1>",
"<subject2>",
"[SHUNT_1000000]",
"[SHUNT_1000001]",
"[SHUNT_1000002]",
"[SHUNT_1000003]",
"[SHUNT_1000004]",
"[SHUNT_1000005]",
"[SHUNT_1000006]",
"[SHUNT_1000007]",
"[SHUNT_1000008]",
"[SHUNT_1000009]",
"[SHUNT_1000010]",
"[SHUNT_1000011]",
"[SHUNT_1000012]",
"[SHUNT_1000013]",
"[SHUNT_1000014]",
"[SHUNT_1000015]",
"[SHUNT_1000016]",
"[SHUNT_1000017]",
"[SHUNT_1000018]",
"[SHUNT_1000019]",
"[SHUNT_1000020]",
"[SHUNT_1000021]",
"[SHUNT_1000022]",
"[SHUNT_1000023]",
"[SHUNT_1000024]",
"[SHUNT_1000025]"
],
"clean_up_tokenization_spaces": true,
"cls_token": "[CLS]",
"do_lower_case": true,
"extra_special_tokens": {},
"mask_token": "[MASK]",
"model_max_length": 8192,
"pad_to_multiple_of": null,
"pad_token": "[PAD]",
"pad_token_type_id": 0,
"padding_side": "right",
"sep_token": "[SEP]",
"stride": 0,
"strip_accents": null,
"tokenize_chinese_chars": true,
"tokenizer_class": "BertTokenizer",
"truncation_side": "right",
"truncation_strategy": "longest_first",
"unk_token": "[UNK]"
}