|
{ |
|
"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_TOTAL_WORDS>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57526": { |
|
"content": "</s_TOTAL_WORDS>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57527": { |
|
"content": "<s_TITLE>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57528": { |
|
"content": "</s_TITLE>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57529": { |
|
"content": "<s_SUB_TOTAL>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57530": { |
|
"content": "</s_SUB_TOTAL>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57531": { |
|
"content": "<s_SEND_TO>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57532": { |
|
"content": "</s_SEND_TO>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57533": { |
|
"content": "<s_OTHER>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57534": { |
|
"content": "</s_OTHER>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57535": { |
|
"content": "<s_NUMBER>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57536": { |
|
"content": "</s_NUMBER>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57537": { |
|
"content": "<s_DATE>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57538": { |
|
"content": "</s_DATE>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57539": { |
|
"content": "<s_CONDITIONS>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57540": { |
|
"content": "</s_CONDITIONS>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57541": { |
|
"content": "<s_TOTAL>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57542": { |
|
"content": "</s_TOTAL>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57543": { |
|
"content": "<s_SELLER_SITE>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57544": { |
|
"content": "</s_SELLER_SITE>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57545": { |
|
"content": "<s_SELLER_NAME>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57546": { |
|
"content": "</s_SELLER_NAME>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57547": { |
|
"content": "<s_SELLER_EMAIL>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57548": { |
|
"content": "</s_SELLER_EMAIL>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57549": { |
|
"content": "<s_SELLER_ADDRESS>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57550": { |
|
"content": "</s_SELLER_ADDRESS>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57551": { |
|
"content": "<s_PAYMENT_DETAILS>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57552": { |
|
"content": "</s_PAYMENT_DETAILS>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57553": { |
|
"content": "<s_BUYER>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57554": { |
|
"content": "</s_BUYER>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57555": { |
|
"content": "<s_NOTE>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57556": { |
|
"content": "</s_NOTE>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57557": { |
|
"content": "<s_GSTIN>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57558": { |
|
"content": "</s_GSTIN>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57559": { |
|
"content": "<s_TAX>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57560": { |
|
"content": "</s_TAX>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57561": { |
|
"content": "<s_PO_NUMBER>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57562": { |
|
"content": "</s_PO_NUMBER>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57563": { |
|
"content": "<s_DUE_DATE>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57564": { |
|
"content": "</s_DUE_DATE>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57565": { |
|
"content": "<s_BILL_TO>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57566": { |
|
"content": "</s_BILL_TO>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57567": { |
|
"content": "<s_GSTIN_SELLER>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57568": { |
|
"content": "</s_GSTIN_SELLER>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57569": { |
|
"content": "<s_GSTIN_BUYER>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57570": { |
|
"content": "</s_GSTIN_BUYER>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57571": { |
|
"content": "<s_DISCOUNT>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57572": { |
|
"content": "</s_DISCOUNT>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57573": { |
|
"content": "<s_GST(5%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57574": { |
|
"content": "</s_GST(5%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57575": { |
|
"content": "<s_GST(20%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57576": { |
|
"content": "</s_GST(20%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57577": { |
|
"content": "<s_GST(18%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57578": { |
|
"content": "</s_GST(18%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57579": { |
|
"content": "<s_GST(12%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57580": { |
|
"content": "</s_GST(12%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57581": { |
|
"content": "<s_GST(1%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57582": { |
|
"content": "</s_GST(1%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57583": { |
|
"content": "<s_GST(7%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57584": { |
|
"content": "</s_GST(7%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57585": { |
|
"content": "<s_AMOUNT_DUE>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57586": { |
|
"content": "</s_AMOUNT_DUE>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57587": { |
|
"content": "<s_GST(9%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
}, |
|
"57588": { |
|
"content": "</s_GST(9%)>", |
|
"lstrip": false, |
|
"normalized": false, |
|
"rstrip": false, |
|
"single_word": false, |
|
"special": true |
|
} |
|
}, |
|
"additional_special_tokens": [ |
|
"<s_TOTAL>", |
|
"</s_TOTAL>", |
|
"<s_TITLE>", |
|
"</s_TITLE>", |
|
"<s_SELLER_NAME>", |
|
"</s_SELLER_NAME>", |
|
"<s_SELLER_ADDRESS>", |
|
"</s_SELLER_ADDRESS>", |
|
"<s_PO_NUMBER>", |
|
"</s_PO_NUMBER>", |
|
"<s_OTHER>", |
|
"</s_OTHER>", |
|
"<s_NUMBER>", |
|
"</s_NUMBER>", |
|
"<s_NOTE>", |
|
"</s_NOTE>", |
|
"<s_DUE_DATE>", |
|
"</s_DUE_DATE>", |
|
"<s_DATE>", |
|
"</s_DATE>", |
|
"<s_TOTAL_WORDS>", |
|
"</s_TOTAL_WORDS>", |
|
"<s_SUB_TOTAL>", |
|
"</s_SUB_TOTAL>", |
|
"<s_SELLER_SITE>", |
|
"</s_SELLER_SITE>", |
|
"<s_SELLER_EMAIL>", |
|
"</s_SELLER_EMAIL>", |
|
"<s_PAYMENT_DETAILS>", |
|
"</s_PAYMENT_DETAILS>", |
|
"<s_GST(9%)>", |
|
"</s_GST(9%)>", |
|
"<s_BUYER>", |
|
"</s_BUYER>", |
|
"<s_CONDITIONS>", |
|
"</s_CONDITIONS>", |
|
"<s_SEND_TO>", |
|
"</s_SEND_TO>", |
|
"<s_GSTIN>", |
|
"</s_GSTIN>", |
|
"<s_AMOUNT_DUE>", |
|
"</s_AMOUNT_DUE>", |
|
"<s_TAX>", |
|
"</s_TAX>", |
|
"<s_BILL_TO>", |
|
"</s_BILL_TO>", |
|
"<s_DISCOUNT>", |
|
"</s_DISCOUNT>", |
|
"<s_GSTIN_SELLER>", |
|
"</s_GSTIN_SELLER>", |
|
"<s_GSTIN_BUYER>", |
|
"</s_GSTIN_BUYER>", |
|
"<s_GST(5%)>", |
|
"</s_GST(5%)>", |
|
"<s_GST(20%)>", |
|
"</s_GST(20%)>", |
|
"<s_GST(18%)>", |
|
"</s_GST(18%)>", |
|
"<s_GST(12%)>", |
|
"</s_GST(12%)>", |
|
"<s_GST(1%)>", |
|
"</s_GST(1%)>", |
|
"<s_GST(7%)>", |
|
"</s_GST(7%)>", |
|
"<s>", |
|
"</s>" |
|
], |
|
"bos_token": "<s>", |
|
"clean_up_tokenization_spaces": true, |
|
"cls_token": "<s>", |
|
"eos_token": "</s>", |
|
"mask_token": "<mask>", |
|
"max_length": 512, |
|
"model_max_length": 1000000000000000019884624838656, |
|
"pad_to_multiple_of": null, |
|
"pad_token": "<pad>", |
|
"pad_token_type_id": 0, |
|
"padding_side": "right", |
|
"processor_class": "DonutProcessor", |
|
"sep_token": "</s>", |
|
"sp_model_kwargs": {}, |
|
"stride": 0, |
|
"tokenizer_class": "XLMRobertaTokenizer", |
|
"truncation_side": "right", |
|
"truncation_strategy": "longest_first", |
|
"unk_token": "<unk>" |
|
} |
|
|