|
{ |
|
"added_tokens_decoder": { |
|
"0": { |
|
"content": "<s>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"1": { |
|
"content": "<pad>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"2": { |
|
"content": "</s>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"3": { |
|
"content": "<unk>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57521": { |
|
"content": "<mask>", |
|
"lstrip": true, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57522": { |
|
"content": "<sep/>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57523": { |
|
"content": "<s_iitcdip>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57524": { |
|
"content": "<s_synthdog>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57525": { |
|
"content": "<s_date>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57526": { |
|
"content": "</s_date>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57527": { |
|
"content": "<s_invoice_number>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57528": { |
|
"content": "</s_invoice_number>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57529": { |
|
"content": "<s_amount>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57530": { |
|
"content": "</s_amount>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57531": { |
|
"content": "<s_desc>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57532": { |
|
"content": "</s_desc>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57533": { |
|
"content": "<s_supplier>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57534": { |
|
"content": "</s_supplier>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57535": { |
|
"content": "<s_name>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57536": { |
|
"content": "</s_name>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57537": { |
|
"content": "<s_address>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57538": { |
|
"content": "</s_address>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57539": { |
|
"content": "<s_vat_number>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57540": { |
|
"content": "</s_vat_number>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57541": { |
|
"content": "<s_customer>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57542": { |
|
"content": "</s_customer>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57543": { |
|
"content": "<s_lines>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57544": { |
|
"content": "</s_lines>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57545": { |
|
"content": "<s_price>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57546": { |
|
"content": "</s_price>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57547": { |
|
"content": "<s_pos>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57548": { |
|
"content": "</s_pos>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57549": { |
|
"content": "<s_qty>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57550": { |
|
"content": "</s_qty>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57551": { |
|
"content": "<s_tax>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57552": { |
|
"content": "</s_tax>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57553": { |
|
"content": "<s_rate>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57554": { |
|
"content": "</s_rate>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57555": { |
|
"content": "<s_contact>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57556": { |
|
"content": "</s_contact>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57557": { |
|
"content": "<s_phone>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57558": { |
|
"content": "</s_phone>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57559": { |
|
"content": "<s_fax>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57560": { |
|
"content": "</s_fax>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57561": { |
|
"content": "<s_email>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57562": { |
|
"content": "</s_email>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57563": { |
|
"content": "<s_subtotal>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57564": { |
|
"content": "</s_subtotal>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
}, |
|
"57565": { |
|
"content": "<s_cord-v2>", |
|
"lstrip": false, |
|
"normalized": true, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": false |
|
} |
|
}, |
|
"additional_special_tokens": [ |
|
"<s_iitcdip>", |
|
"<s_synthdog>" |
|
], |
|
"bos_token": "<s>", |
|
"clean_up_tokenization_spaces": true, |
|
"cls_token": "<s>", |
|
"eos_token": "</s>", |
|
"mask_token": "<mask>", |
|
"model_max_length": 1000000000000000019884624838656, |
|
"pad_token": "<pad>", |
|
"processor_class": "DonutProcessor", |
|
"sep_token": "</s>", |
|
"sp_model_kwargs": {}, |
|
"tokenizer_class": "XLMRobertaTokenizer", |
|
"unk_token": "<unk>" |
|
} |
|
|