doshan1250 commited on
Commit
608deae
1 Parent(s): ff9a9e4
added_tokens.json ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</s_MOQ>": 57593,
3
+ "</s_UniformNum>": 57541,
4
+ "</s_address>": 57527,
5
+ "</s_batch>": 57595,
6
+ "</s_connector>": 57547,
7
+ "</s_connectorFax>": 57551,
8
+ "</s_connectorTel>": 57549,
9
+ "</s_count>": 57565,
10
+ "</s_customerName>": 57543,
11
+ "</s_customerNo>": 57599,
12
+ "</s_customerSign>": 57585,
13
+ "</s_date>": 57537,
14
+ "</s_deliverAddress>": 57553,
15
+ "</s_deliverNum>": 57589,
16
+ "</s_dueDate>": 57597,
17
+ "</s_fax>": 57531,
18
+ "</s_index>": 57591,
19
+ "</s_invoiceData>": 57573,
20
+ "</s_invoiceNum>": 57577,
21
+ "</s_invoiceStage>": 57581,
22
+ "</s_invoiceWay>": 57579,
23
+ "</s_listNum>": 57587,
24
+ "</s_memo>": 57583,
25
+ "</s_orderNum>": 57539,
26
+ "</s_page>": 57533,
27
+ "</s_priceTax>": 57545,
28
+ "</s_productContent>": 57561,
29
+ "</s_productID>": 57559,
30
+ "</s_productMemo>": 57569,
31
+ "</s_products>": 57557,
32
+ "</s_sales>": 57555,
33
+ "</s_storage>": 57563,
34
+ "</s_stroage>": 57571,
35
+ "</s_taxTag>": 57575,
36
+ "</s_tel>": 57529,
37
+ "</s_title>": 57525,
38
+ "</s_type>": 57535,
39
+ "</s_unit>": 57567,
40
+ "<s_MOQ>": 57594,
41
+ "<s_UniformNum>": 57542,
42
+ "<s_address>": 57528,
43
+ "<s_batch>": 57596,
44
+ "<s_connector>": 57548,
45
+ "<s_connectorFax>": 57552,
46
+ "<s_connectorTel>": 57550,
47
+ "<s_count>": 57566,
48
+ "<s_customerName>": 57544,
49
+ "<s_customerNo>": 57600,
50
+ "<s_customerSign>": 57586,
51
+ "<s_date>": 57538,
52
+ "<s_deliverAddress>": 57554,
53
+ "<s_deliverNum>": 57590,
54
+ "<s_dueDate>": 57598,
55
+ "<s_fax>": 57532,
56
+ "<s_iitcdip>": 57523,
57
+ "<s_index>": 57592,
58
+ "<s_invoiceData>": 57574,
59
+ "<s_invoiceNum>": 57578,
60
+ "<s_invoiceStage>": 57582,
61
+ "<s_invoiceWay>": 57580,
62
+ "<s_listNum>": 57588,
63
+ "<s_memo>": 57584,
64
+ "<s_orderNum>": 57540,
65
+ "<s_page>": 57534,
66
+ "<s_preparedFinetuneData_Bird>": 57601,
67
+ "<s_priceTax>": 57546,
68
+ "<s_productContent>": 57562,
69
+ "<s_productID>": 57560,
70
+ "<s_productMemo>": 57570,
71
+ "<s_products>": 57558,
72
+ "<s_sales>": 57556,
73
+ "<s_storage>": 57564,
74
+ "<s_stroage>": 57572,
75
+ "<s_synthdog>": 57524,
76
+ "<s_taxTag>": 57576,
77
+ "<s_tel>": 57530,
78
+ "<s_title>": 57526,
79
+ "<s_type>": 57536,
80
+ "<s_unit>": 57568,
81
+ "<sep/>": 57522
82
+ }
artifacts.ckpt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:64b6223b1be5d9faefc75a6446eb40880e6393d650d2d081ddf4c3a917300d54
3
+ size 1609478544
config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "naver-clova-ix/donut-base",
3
+ "align_long_axis": false,
4
+ "architectures": [
5
+ "DonutModel"
6
+ ],
7
+ "decoder_layer": 4,
8
+ "encoder_layer": [
9
+ 2,
10
+ 2,
11
+ 14,
12
+ 2
13
+ ],
14
+ "input_size": [
15
+ 1280,
16
+ 960
17
+ ],
18
+ "max_length": 768,
19
+ "max_position_embeddings": 768,
20
+ "model_type": "donut",
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.25.1",
23
+ "window_size": 10
24
+ }
config.yaml ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ resume_from_checkpoint_path: None
2
+ result_path: 'gdrive/MyDrive/result_Bird'
3
+ pretrained_model_name_or_path: 'naver-clova-ix/donut-base'
4
+ dataset_name_or_paths:
5
+ - 'gdrive/MyDrive/preparedFinetuneData_Bird'
6
+ sort_json_key: False
7
+ train_batch_sizes:
8
+ - 1
9
+ val_batch_sizes:
10
+ - 1
11
+ input_size:
12
+ - 1280
13
+ - 960
14
+ max_length: 768
15
+ align_long_axis: False
16
+ num_nodes: 1
17
+ seed: 2022
18
+ lr: 3e-05
19
+ warmup_steps: 0
20
+ num_training_samples_per_epoch: 800
21
+ max_epochs: 30
22
+ max_steps: -1
23
+ num_workers: 2
24
+ val_check_interval: 1.0
25
+ check_val_every_n_epoch: 10
26
+ gradient_clip_val: 1.0
27
+ verbose: True
28
+ exp_name: 'train_cord_finetune_Bird'
29
+ exp_version: '20240327_130533'
events.out.tfevents.1711544851.c2016526b586.3075.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f24a09d7d984de9dd3c824478a62a46263229a61580a390fab67396a6e78306c
3
+ size 8062
hparams.yaml ADDED
@@ -0,0 +1 @@
 
 
1
+ {}
model_checkpoint.ckpt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:47aeb8ce8ced98c5548b5221fc87b438678c5f1fe0a134eeb018ec0f29757235
3
+ size 1609475934
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:38c0208d9ed22b2aa8b9d869373429bcce0f80b3e5b29e5eb8700f6237140017
3
+ size 858644166
sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb9e3dce4c326195d08fc3dd0f7e2eee1da8595c847bf4c1a9c78b7a82d47e2d
3
+ size 1296245
special_tokens_map.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<s_preparedFinetuneData_Bird>"
4
+ ],
5
+ "bos_token": "<s>",
6
+ "cls_token": "<s>",
7
+ "eos_token": "</s>",
8
+ "mask_token": {
9
+ "content": "<mask>",
10
+ "lstrip": true,
11
+ "normalized": true,
12
+ "rstrip": false,
13
+ "single_word": false
14
+ },
15
+ "pad_token": "<pad>",
16
+ "sep_token": "</s>",
17
+ "unk_token": "<unk>"
18
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "from_slow": true,
6
+ "mask_token": {
7
+ "__type": "AddedToken",
8
+ "content": "<mask>",
9
+ "lstrip": true,
10
+ "normalized": true,
11
+ "rstrip": false,
12
+ "single_word": false
13
+ },
14
+ "model_max_length": 1000000000000000019884624838656,
15
+ "name_or_path": "naver-clova-ix/donut-base",
16
+ "pad_token": "<pad>",
17
+ "processor_class": "DonutProcessor",
18
+ "sep_token": "</s>",
19
+ "sp_model_kwargs": {},
20
+ "special_tokens_map_file": null,
21
+ "tokenizer_class": "XLMRobertaTokenizer",
22
+ "unk_token": "<unk>"
23
+ }