donut-base-ner_t4 / tokenizer_config.json
JandC's picture
End of training
80b44e4
raw
history blame contribute delete
No virus
7.16 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_gross_weight>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57526": {
"content": "</s_purchase_order>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57527": {
"content": "</s_item_index>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57528": {
"content": "</s_part_number>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57529": {
"content": "</s_total_item_value>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57530": {
"content": "</s_item_quantity>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57531": {
"content": "<s_part_number>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57532": {
"content": "<s_hs_code>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57533": {
"content": "</s_origin_country>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57534": {
"content": "<s_item_value>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57535": {
"content": "</s_hs_code>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57536": {
"content": "<s_net_weight>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57537": {
"content": "<s_item_quantity>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57538": {
"content": "</s_description>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57539": {
"content": "<s_origin_country>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57540": {
"content": "<s_gross_weight>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57541": {
"content": "</s_net_weight>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57542": {
"content": "<s_uom>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57543": {
"content": "</s_item_value>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57544": {
"content": "</s_uom>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57545": {
"content": "<s_purchase_order>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57546": {
"content": "<s_description>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57547": {
"content": "<s_serial_number>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57548": {
"content": "<s_total_item_value>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57549": {
"content": "</s_serial_number>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"57550": {
"content": "<s_item_index>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
}
},
"additional_special_tokens": [
"</s_gross_weight>",
"</s_purchase_order>",
"</s_item_index>",
"</s_part_number>",
"</s_total_item_value>",
"</s_item_quantity>",
"<s_part_number>",
"<s_hs_code>",
"</s_origin_country>",
"<s_item_value>",
"</s_hs_code>",
"<s_net_weight>",
"<s>",
"<s_item_quantity>",
"</s_description>",
"<s_origin_country>",
"<s_gross_weight>",
"</s_net_weight>",
"<s_uom>",
"</s_item_value>",
"</s_uom>",
"<s_purchase_order>",
"<s_description>",
"<s_serial_number>",
"</s>",
"<s_total_item_value>",
"</s_serial_number>",
"<s_item_index>"
],
"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>"
}