Hugging Face's logo

rizkynindra
/
donut-ktp

TensorBoard
Model card Files Metrics Community
rizkynindra commited on
Commit
a00b85f
•
1 Parent(s): ad340b7

donut-ktp v1.0.0

Browse files

Donut model for KTP

Files changed (15) hide show
  1. 20240902_131908/.ipynb_checkpoints/added_tokens-checkpoint.json +89 -0
  2. 20240902_131908/.ipynb_checkpoints/config-checkpoint.json +24 -0
  3. 20240902_131908/.ipynb_checkpoints/config-checkpoint.yaml +29 -0
  4. 20240902_131908/.ipynb_checkpoints/special_tokens_map-checkpoint.json +18 -0
  5. 20240902_131908/.ipynb_checkpoints/tokenizer_config-checkpoint.json +23 -0
  6. 20240902_131908/added_tokens.json +89 -0
  7. 20240902_131908/artifacts.ckpt +3 -0
  8. 20240902_131908/config.json +24 -0
  9. 20240902_131908/config.yaml +29 -0
  10. 20240902_131908/events.out.tfevents.1725258004.43A-07-APT-5 +3 -0
  11. 20240902_131908/hparams.yaml +1 -0
  12. 20240902_131908/pytorch_model.bin +3 -0
  13. 20240902_131908/sentencepiece.bpe.model +3 -0
  14. 20240902_131908/special_tokens_map.json +18 -0
  15. 20240902_131908/tokenizer_config.json +23 -0
20240902_131908/.ipynb_checkpoints/added_tokens-checkpoint.json ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</s_Alamat>": 57604,
3
+ "</s_Jenis Kelamin>": 57588,
4
+ "</s_Kelurahan>": 57606,
5
+ "</s_NIK>": 57598,
6
+ "</s_Nama>": 57600,
7
+ "</s_RT/RW>": 57592,
8
+ "</s_Tanggal Lahir>": 57602,
9
+ "</s_Tempat Lahir>": 57584,
10
+ "</s_alamat>": 57590,
11
+ "</s_cashprice>": 57549,
12
+ "</s_changeprice>": 57551,
13
+ "</s_cnt>": 57529,
14
+ "</s_creditcardprice>": 57563,
15
+ "</s_discount_price>": 57557,
16
+ "</s_discountprice>": 57567,
17
+ "</s_emoneyprice>": 57569,
18
+ "</s_etc>": 57541,
19
+ "</s_itemsubtotal>": 57577,
20
+ "</s_kecamatan>": 57596,
21
+ "</s_kel_desa>": 57594,
22
+ "</s_menu>": 57525,
23
+ "</s_menuqty_cnt>": 57555,
24
+ "</s_menutype_cnt>": 57553,
25
+ "</s_nama>": 57582,
26
+ "</s_nik>": 57580,
27
+ "</s_nm>": 57527,
28
+ "</s_num>": 57565,
29
+ "</s_othersvc_price>": 57573,
30
+ "</s_price>": 57531,
31
+ "</s_service_price>": 57537,
32
+ "</s_sub>": 57547,
33
+ "</s_sub_total>": 57533,
34
+ "</s_subtotal_price>": 57535,
35
+ "</s_tax_price>": 57539,
36
+ "</s_tgl_lahir>": 57586,
37
+ "</s_total>": 57543,
38
+ "</s_total_etc>": 57561,
39
+ "</s_total_price>": 57545,
40
+ "</s_unitprice>": 57559,
41
+ "</s_vatyn>": 57575,
42
+ "</s_void_menu>": 57571,
43
+ "<s_Alamat>": 57605,
44
+ "<s_Jenis Kelamin>": 57589,
45
+ "<s_Kelurahan>": 57607,
46
+ "<s_NIK>": 57599,
47
+ "<s_Nama>": 57601,
48
+ "<s_RT/RW>": 57593,
49
+ "<s_Tanggal Lahir>": 57603,
50
+ "<s_Tempat Lahir>": 57585,
51
+ "<s_alamat>": 57591,
52
+ "<s_cashprice>": 57550,
53
+ "<s_changeprice>": 57552,
54
+ "<s_cnt>": 57530,
55
+ "<s_cord-v2>": 57579,
56
+ "<s_creditcardprice>": 57564,
57
+ "<s_data>": 57608,
58
+ "<s_discount_price>": 57558,
59
+ "<s_discountprice>": 57568,
60
+ "<s_emoneyprice>": 57570,
61
+ "<s_etc>": 57542,
62
+ "<s_iitcdip>": 57523,
63
+ "<s_itemsubtotal>": 57578,
64
+ "<s_kecamatan>": 57597,
65
+ "<s_kel_desa>": 57595,
66
+ "<s_menu>": 57526,
67
+ "<s_menuqty_cnt>": 57556,
68
+ "<s_menutype_cnt>": 57554,
69
+ "<s_nama>": 57583,
70
+ "<s_nik>": 57581,
71
+ "<s_nm>": 57528,
72
+ "<s_num>": 57566,
73
+ "<s_othersvc_price>": 57574,
74
+ "<s_price>": 57532,
75
+ "<s_service_price>": 57538,
76
+ "<s_sub>": 57548,
77
+ "<s_sub_total>": 57534,
78
+ "<s_subtotal_price>": 57536,
79
+ "<s_synthdog>": 57524,
80
+ "<s_tax_price>": 57540,
81
+ "<s_tgl_lahir>": 57587,
82
+ "<s_total>": 57544,
83
+ "<s_total_etc>": 57562,
84
+ "<s_total_price>": 57546,
85
+ "<s_unitprice>": 57560,
86
+ "<s_vatyn>": 57576,
87
+ "<s_void_menu>": 57572,
88
+ "<sep/>": 57522
89
+ }
20240902_131908/.ipynb_checkpoints/config-checkpoint.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "naver-clova-ix/donut-base-finetuned-cord-v2",
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
+ }
20240902_131908/.ipynb_checkpoints/config-checkpoint.yaml ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ resume_from_checkpoint_path: None
2
+ result_path: '/home/rizkynindra/result'
3
+ pretrained_model_name_or_path: 'naver-clova-ix/donut-base-finetuned-cord-v2'
4
+ dataset_name_or_paths:
5
+ - '/home/rizkynindra/data'
6
+ sort_json_key: False
7
+ train_batch_sizes:
8
+ - 2
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: 222
20
+ num_training_samples_per_epoch: 148
21
+ max_epochs: 30
22
+ max_steps: -1
23
+ num_workers: 4
24
+ val_check_interval: 1.0
25
+ check_val_every_n_epoch: 3
26
+ gradient_clip_val: 1.0
27
+ verbose: True
28
+ exp_name: 'train_id'
29
+ exp_version: '20240902_131908'
20240902_131908/.ipynb_checkpoints/special_tokens_map-checkpoint.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<s_data>"
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
+ }
20240902_131908/.ipynb_checkpoints/tokenizer_config-checkpoint.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-finetuned-cord-v2",
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
+ }
20240902_131908/added_tokens.json ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</s_Alamat>": 57604,
3
+ "</s_Jenis Kelamin>": 57588,
4
+ "</s_Kelurahan>": 57606,
5
+ "</s_NIK>": 57598,
6
+ "</s_Nama>": 57600,
7
+ "</s_RT/RW>": 57592,
8
+ "</s_Tanggal Lahir>": 57602,
9
+ "</s_Tempat Lahir>": 57584,
10
+ "</s_alamat>": 57590,
11
+ "</s_cashprice>": 57549,
12
+ "</s_changeprice>": 57551,
13
+ "</s_cnt>": 57529,
14
+ "</s_creditcardprice>": 57563,
15
+ "</s_discount_price>": 57557,
16
+ "</s_discountprice>": 57567,
17
+ "</s_emoneyprice>": 57569,
18
+ "</s_etc>": 57541,
19
+ "</s_itemsubtotal>": 57577,
20
+ "</s_kecamatan>": 57596,
21
+ "</s_kel_desa>": 57594,
22
+ "</s_menu>": 57525,
23
+ "</s_menuqty_cnt>": 57555,
24
+ "</s_menutype_cnt>": 57553,
25
+ "</s_nama>": 57582,
26
+ "</s_nik>": 57580,
27
+ "</s_nm>": 57527,
28
+ "</s_num>": 57565,
29
+ "</s_othersvc_price>": 57573,
30
+ "</s_price>": 57531,
31
+ "</s_service_price>": 57537,
32
+ "</s_sub>": 57547,
33
+ "</s_sub_total>": 57533,
34
+ "</s_subtotal_price>": 57535,
35
+ "</s_tax_price>": 57539,
36
+ "</s_tgl_lahir>": 57586,
37
+ "</s_total>": 57543,
38
+ "</s_total_etc>": 57561,
39
+ "</s_total_price>": 57545,
40
+ "</s_unitprice>": 57559,
41
+ "</s_vatyn>": 57575,
42
+ "</s_void_menu>": 57571,
43
+ "<s_Alamat>": 57605,
44
+ "<s_Jenis Kelamin>": 57589,
45
+ "<s_Kelurahan>": 57607,
46
+ "<s_NIK>": 57599,
47
+ "<s_Nama>": 57601,
48
+ "<s_RT/RW>": 57593,
49
+ "<s_Tanggal Lahir>": 57603,
50
+ "<s_Tempat Lahir>": 57585,
51
+ "<s_alamat>": 57591,
52
+ "<s_cashprice>": 57550,
53
+ "<s_changeprice>": 57552,
54
+ "<s_cnt>": 57530,
55
+ "<s_cord-v2>": 57579,
56
+ "<s_creditcardprice>": 57564,
57
+ "<s_data>": 57608,
58
+ "<s_discount_price>": 57558,
59
+ "<s_discountprice>": 57568,
60
+ "<s_emoneyprice>": 57570,
61
+ "<s_etc>": 57542,
62
+ "<s_iitcdip>": 57523,
63
+ "<s_itemsubtotal>": 57578,
64
+ "<s_kecamatan>": 57597,
65
+ "<s_kel_desa>": 57595,
66
+ "<s_menu>": 57526,
67
+ "<s_menuqty_cnt>": 57556,
68
+ "<s_menutype_cnt>": 57554,
69
+ "<s_nama>": 57583,
70
+ "<s_nik>": 57581,
71
+ "<s_nm>": 57528,
72
+ "<s_num>": 57566,
73
+ "<s_othersvc_price>": 57574,
74
+ "<s_price>": 57532,
75
+ "<s_service_price>": 57538,
76
+ "<s_sub>": 57548,
77
+ "<s_sub_total>": 57534,
78
+ "<s_subtotal_price>": 57536,
79
+ "<s_synthdog>": 57524,
80
+ "<s_tax_price>": 57540,
81
+ "<s_tgl_lahir>": 57587,
82
+ "<s_total>": 57544,
83
+ "<s_total_etc>": 57562,
84
+ "<s_total_price>": 57546,
85
+ "<s_unitprice>": 57560,
86
+ "<s_vatyn>": 57576,
87
+ "<s_void_menu>": 57572,
88
+ "<sep/>": 57522
89
+ }
20240902_131908/artifacts.ckpt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:64774e03675aeb9f12581810c6289f5e81a6e9bc984ff5f3fecab4bd48604152
3
+ size 1609536155
20240902_131908/config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "naver-clova-ix/donut-base-finetuned-cord-v2",
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
+ }
20240902_131908/config.yaml ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ resume_from_checkpoint_path: None
2
+ result_path: '/home/rizkynindra/result'
3
+ pretrained_model_name_or_path: 'naver-clova-ix/donut-base-finetuned-cord-v2'
4
+ dataset_name_or_paths:
5
+ - '/home/rizkynindra/data'
6
+ sort_json_key: False
7
+ train_batch_sizes:
8
+ - 2
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: 222
20
+ num_training_samples_per_epoch: 148
21
+ max_epochs: 30
22
+ max_steps: -1
23
+ num_workers: 4
24
+ val_check_interval: 1.0
25
+ check_val_every_n_epoch: 3
26
+ gradient_clip_val: 1.0
27
+ verbose: True
28
+ exp_name: 'train_id'
29
+ exp_version: '20240902_131908'
20240902_131908/events.out.tfevents.1725258004.43A-07-APT-5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:464e589745b1fc138beeca8fc2d085aadfb103f9c0a0cf5d87793ab95960be29
3
+ size 3882
20240902_131908/hparams.yaml ADDED
@@ -0,0 +1 @@
 
 
1
+ {}
20240902_131908/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58bb9e6a9f2dcbdbd2f9c8e0349477bb9e134e58c47bcb4869e3dc5936004612
3
+ size 858678217
20240902_131908/sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb9e3dce4c326195d08fc3dd0f7e2eee1da8595c847bf4c1a9c78b7a82d47e2d
3
+ size 1296245
20240902_131908/special_tokens_map.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<s_data>"
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
+ }
20240902_131908/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-finetuned-cord-v2",
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
+ }