Hugging Face's logo

alaeddinehamroun
/
donut-base-custom-invoices

Image-Text-to-Text
Transformers
TensorBoard
Safetensors
vision-encoder-decoder
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
donut-base-custom-invoices / tokenizer_config.json
alaeddinehamroun's picture
alaeddinehamroun
End of training
4fa0988
raw
history blame contribute delete
8.38 kB
{
"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_item_desc>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57526": {
"content": "<s_seller_tax_id>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57527": {
"content": "<s_total_gross_worth>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57528": {
"content": "<s_client_tax_id>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57529": {
"content": "</s_item_net_price>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57530": {
"content": "</s_invoice_date>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57531": {
"content": "<s_item_gross_worth>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57532": {
"content": "</s_seller_tax_id>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57533": {
"content": "<s_item_net_price>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57534": {
"content": "<s_iban>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57535": {
"content": "</s_item_desc>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57536": {
"content": "</s_total_vat>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57537": {
"content": "</s_item_net_worth>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57538": {
"content": "</s_total_net_worth>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57539": {
"content": "</s_iban>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57540": {
"content": "</s_item_qty>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57541": {
"content": "<s_item_vat>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57542": {
"content": "<s_seller>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57543": {
"content": "<s_total_net_worth>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57544": {
"content": "<s_client>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57545": {
"content": "</s_total_gross_worth>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57546": {
"content": "</s_client>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57547": {
"content": "<s_invoice_date>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57548": {
"content": "<s_total_vat>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57549": {
"content": "<s_invoice_no>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57550": {
"content": "</s_seller>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57551": {
"content": "<s_item_qty>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57552": {
"content": "</s_item_vat>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57553": {
"content": "</s_item_gross_worth>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57554": {
"content": "</s_client_tax_id>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57555": {
"content": "</s_invoice_no>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57556": {
"content": "<s_item_net_worth>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
}
},
"additional_special_tokens": [
"<s_item_desc>",
"<s_seller_tax_id>",
"<s_total_gross_worth>",
"<s_client_tax_id>",
"</s_item_net_price>",
"</s_invoice_date>",
"<s_item_gross_worth>",
"</s_seller_tax_id>",
"<s_item_net_price>",
"<s_iban>",
"</s_item_desc>",
"</s_total_vat>",
"</s_item_net_worth>",
"</s_total_net_worth>",
"</s_iban>",
"<s>",
"</s_item_qty>",
"<s_item_vat>",
"<s_seller>",
"<s_total_net_worth>",
"<s_client>",
"</s_total_gross_worth>",
"</s_client>",
"<s_invoice_date>",
"<s_total_vat>",
"<s_invoice_no>",
"</s_seller>",
"</s>",
"<s_item_qty>",
"</s_item_vat>",
"</s_item_gross_worth>",
"</s_client_tax_id>",
"</s_invoice_no>",
"<s_item_net_worth>"
],
"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>"
}