Hugging Face's logo

emisilab
/
model-ocr-invoice-v1

Image-Text-to-Text
Transformers
TensorBoard
Safetensors
vision-encoder-decoder
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
emisilab commited on
Commit
972cf2e
•
1 Parent(s): 827c4a5

End of training

Browse files
Files changed (6) hide show
  1. added_tokens.json +30 -28
  2. preprocessor_config.json +4 -4
  3. runs/Jun12_13-37-24_4b62348ce371/events.out.tfevents.1718199475.4b62348ce371.847.1 +2 -2
  4. special_tokens_map.json +224 -30
  5. tokenizer.json +0 -0
  6. tokenizer_config.json +74 -63
added_tokens.json CHANGED
@@ -1,33 +1,35 @@
1
  {
2
- "</s_AlamatLawanTransaksi>": 57552,
3
- "</s_AlamatPenjual>": 57550,
4
- "</s_NPWPlawanTransaksi>": 57548,
5
- "</s_NPWPpenjual>": 57546,
6
- "</s_NamaPenjual>": 57544,
7
- "</s_NoFakturPajak>": 57542,
8
- "</s_NoNPWPlawanTransaksi>": 57540,
9
- "</s_PPnBM>": 57538,
10
- "</s_TanggalFakturPajak>": 57536,
11
- "</s_diskon>": 57534,
12
- "</s_dpp>": 57532,
13
- "</s_hargaTotal>": 57530,
14
- "</s_ppn>": 57528,
15
- "</s_referensi>": 57526,
16
- "<s_AlamatLawanTransaksi>": 57551,
17
- "<s_AlamatPenjual>": 57549,
18
- "<s_NPWPlawanTransaksi>": 57547,
19
- "<s_NPWPpenjual>": 57545,
20
- "<s_NamaPenjual>": 57543,
21
- "<s_NoFakturPajak>": 57541,
22
- "<s_NoNPWPlawanTransaksi>": 57539,
23
- "<s_PPnBM>": 57537,
24
- "<s_TanggalFakturPajak>": 57535,
25
- "<s_diskon>": 57533,
26
- "<s_dpp>": 57531,
27
- "<s_hargaTotal>": 57529,
28
  "<s_iitcdip>": 57523,
29
- "<s_ppn>": 57527,
30
- "<s_referensi>": 57525,
 
31
  "<s_synthdog>": 57524,
 
32
  "<sep/>": 57522
33
  }
 
1
  {
2
+ "</s_DPP>": 57554,
3
+ "</s_Meterai>": 57552,
4
+ "</s_NILAI METERAI>": 57550,
5
+ "</s_Nama_Bank>": 57548,
6
+ "</s_Nama_Rek>": 57546,
7
+ "</s_No_Rek>": 57544,
8
+ "</s_Periode>": 57542,
9
+ "</s_Tax>": 57540,
10
+ "</s_Total>": 57538,
11
+ "</s_Type_Tax>": 57536,
12
+ "</s_Type_Transaksi>": 57534,
13
+ "</s_nama_vendor>": 57532,
14
+ "</s_no_invoice>": 57530,
15
+ "</s_no_kontrak>": 57528,
16
+ "</s_tanggal_invoice>": 57526,
17
+ "<s_DPP>": 57553,
18
+ "<s_Meterai>": 57551,
19
+ "<s_NILAI METERAI>": 57549,
20
+ "<s_Nama_Bank>": 57547,
21
+ "<s_Nama_Rek>": 57545,
22
+ "<s_No_Rek>": 57543,
23
+ "<s_Periode>": 57541,
24
+ "<s_Tax>": 57539,
25
+ "<s_Total>": 57537,
26
+ "<s_Type_Tax>": 57535,
27
+ "<s_Type_Transaksi>": 57533,
28
  "<s_iitcdip>": 57523,
29
+ "<s_nama_vendor>": 57531,
30
+ "<s_no_invoice>": 57529,
31
+ "<s_no_kontrak>": 57527,
32
  "<s_synthdog>": 57524,
33
+ "<s_tanggal_invoice>": 57525,
34
  "<sep/>": 57522
35
  }
preprocessor_config.json CHANGED
@@ -37,8 +37,8 @@
37
  "processor_class": "DonutProcessor",
38
  "resample": 2,
39
  "rescale_factor": 0.00392156862745098,
40
- "size": {
41
- "height": 960,
42
- "width": 720
43
- }
44
  }
 
37
  "processor_class": "DonutProcessor",
38
  "resample": 2,
39
  "rescale_factor": 0.00392156862745098,
40
+ "size": [
41
+ 720,
42
+ 960
43
+ ]
44
  }
runs/Jun12_13-37-24_4b62348ce371/events.out.tfevents.1718199475.4b62348ce371.847.1 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9c44c65a1744dd3e7ff17d9c1f31b7a4a0b7cc279f748c65b5179f5785904b86
3
- size 19649
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4fd6b7d40885fee9001bf792d356392e85be908ea90debb48ed7492e049d15cd
3
+ size 20003
special_tokens_map.json CHANGED
@@ -1,35 +1,229 @@
1
  {
2
  "additional_special_tokens": [
3
- "<s_referensi>",
4
- "</s_referensi>",
5
- "<s_ppn>",
6
- "</s_ppn>",
7
- "<s_hargaTotal>",
8
- "</s_hargaTotal>",
9
- "<s_dpp>",
10
- "</s_dpp>",
11
- "<s_diskon>",
12
- "</s_diskon>",
13
- "<s_TanggalFakturPajak>",
14
- "</s_TanggalFakturPajak>",
15
- "<s_PPnBM>",
16
- "</s_PPnBM>",
17
- "<s_NoNPWPlawanTransaksi>",
18
- "</s_NoNPWPlawanTransaksi>",
19
- "<s_NoFakturPajak>",
20
- "</s_NoFakturPajak>",
21
- "<s_NamaPenjual>",
22
- "</s_NamaPenjual>",
23
- "<s_NPWPpenjual>",
24
- "</s_NPWPpenjual>",
25
- "<s_NPWPlawanTransaksi>",
26
- "</s_NPWPlawanTransaksi>",
27
- "<s_AlamatPenjual>",
28
- "</s_AlamatPenjual>",
29
- "<s_AlamatLawanTransaksi>",
30
- "</s_AlamatLawanTransaksi>",
31
- "<s>",
32
- "</s>"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
33
  ],
34
  "bos_token": {
35
  "content": "<s>",
 
1
  {
2
  "additional_special_tokens": [
3
+ {
4
+ "content": "<s_tanggal_invoice>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ {
11
+ "content": "</s_tanggal_invoice>",
12
+ "lstrip": false,
13
+ "normalized": false,
14
+ "rstrip": false,
15
+ "single_word": false
16
+ },
17
+ {
18
+ "content": "<s_no_kontrak>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ {
25
+ "content": "</s_no_kontrak>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ },
31
+ {
32
+ "content": "<s_no_invoice>",
33
+ "lstrip": false,
34
+ "normalized": false,
35
+ "rstrip": false,
36
+ "single_word": false
37
+ },
38
+ {
39
+ "content": "</s_no_invoice>",
40
+ "lstrip": false,
41
+ "normalized": false,
42
+ "rstrip": false,
43
+ "single_word": false
44
+ },
45
+ {
46
+ "content": "<s_nama_vendor>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false
51
+ },
52
+ {
53
+ "content": "</s_nama_vendor>",
54
+ "lstrip": false,
55
+ "normalized": false,
56
+ "rstrip": false,
57
+ "single_word": false
58
+ },
59
+ {
60
+ "content": "<s_Type_Transaksi>",
61
+ "lstrip": false,
62
+ "normalized": false,
63
+ "rstrip": false,
64
+ "single_word": false
65
+ },
66
+ {
67
+ "content": "</s_Type_Transaksi>",
68
+ "lstrip": false,
69
+ "normalized": false,
70
+ "rstrip": false,
71
+ "single_word": false
72
+ },
73
+ {
74
+ "content": "<s_Type_Tax>",
75
+ "lstrip": false,
76
+ "normalized": false,
77
+ "rstrip": false,
78
+ "single_word": false
79
+ },
80
+ {
81
+ "content": "</s_Type_Tax>",
82
+ "lstrip": false,
83
+ "normalized": false,
84
+ "rstrip": false,
85
+ "single_word": false
86
+ },
87
+ {
88
+ "content": "<s_Total>",
89
+ "lstrip": false,
90
+ "normalized": false,
91
+ "rstrip": false,
92
+ "single_word": false
93
+ },
94
+ {
95
+ "content": "</s_Total>",
96
+ "lstrip": false,
97
+ "normalized": false,
98
+ "rstrip": false,
99
+ "single_word": false
100
+ },
101
+ {
102
+ "content": "<s_Tax>",
103
+ "lstrip": false,
104
+ "normalized": false,
105
+ "rstrip": false,
106
+ "single_word": false
107
+ },
108
+ {
109
+ "content": "</s_Tax>",
110
+ "lstrip": false,
111
+ "normalized": false,
112
+ "rstrip": false,
113
+ "single_word": false
114
+ },
115
+ {
116
+ "content": "<s_Periode>",
117
+ "lstrip": false,
118
+ "normalized": false,
119
+ "rstrip": false,
120
+ "single_word": false
121
+ },
122
+ {
123
+ "content": "</s_Periode>",
124
+ "lstrip": false,
125
+ "normalized": false,
126
+ "rstrip": false,
127
+ "single_word": false
128
+ },
129
+ {
130
+ "content": "<s_No_Rek>",
131
+ "lstrip": false,
132
+ "normalized": false,
133
+ "rstrip": false,
134
+ "single_word": false
135
+ },
136
+ {
137
+ "content": "</s_No_Rek>",
138
+ "lstrip": false,
139
+ "normalized": false,
140
+ "rstrip": false,
141
+ "single_word": false
142
+ },
143
+ {
144
+ "content": "<s_Nama_Rek>",
145
+ "lstrip": false,
146
+ "normalized": false,
147
+ "rstrip": false,
148
+ "single_word": false
149
+ },
150
+ {
151
+ "content": "</s_Nama_Rek>",
152
+ "lstrip": false,
153
+ "normalized": false,
154
+ "rstrip": false,
155
+ "single_word": false
156
+ },
157
+ {
158
+ "content": "<s_Nama_Bank>",
159
+ "lstrip": false,
160
+ "normalized": false,
161
+ "rstrip": false,
162
+ "single_word": false
163
+ },
164
+ {
165
+ "content": "</s_Nama_Bank>",
166
+ "lstrip": false,
167
+ "normalized": false,
168
+ "rstrip": false,
169
+ "single_word": false
170
+ },
171
+ {
172
+ "content": "<s_NILAI METERAI>",
173
+ "lstrip": false,
174
+ "normalized": false,
175
+ "rstrip": false,
176
+ "single_word": false
177
+ },
178
+ {
179
+ "content": "</s_NILAI METERAI>",
180
+ "lstrip": false,
181
+ "normalized": false,
182
+ "rstrip": false,
183
+ "single_word": false
184
+ },
185
+ {
186
+ "content": "<s_Meterai>",
187
+ "lstrip": false,
188
+ "normalized": false,
189
+ "rstrip": false,
190
+ "single_word": false
191
+ },
192
+ {
193
+ "content": "</s_Meterai>",
194
+ "lstrip": false,
195
+ "normalized": false,
196
+ "rstrip": false,
197
+ "single_word": false
198
+ },
199
+ {
200
+ "content": "<s_DPP>",
201
+ "lstrip": false,
202
+ "normalized": false,
203
+ "rstrip": false,
204
+ "single_word": false
205
+ },
206
+ {
207
+ "content": "</s_DPP>",
208
+ "lstrip": false,
209
+ "normalized": false,
210
+ "rstrip": false,
211
+ "single_word": false
212
+ },
213
+ {
214
+ "content": "<s>",
215
+ "lstrip": false,
216
+ "normalized": false,
217
+ "rstrip": false,
218
+ "single_word": false
219
+ },
220
+ {
221
+ "content": "</s>",
222
+ "lstrip": false,
223
+ "normalized": false,
224
+ "rstrip": false,
225
+ "single_word": false
226
+ }
227
  ],
228
  "bos_token": {
229
  "content": "<s>",
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json CHANGED
@@ -65,7 +65,7 @@
65
  "special": true
66
  },
67
  "57525": {
68
- "content": "<s_referensi>",
69
  "lstrip": false,
70
  "normalized": false,
71
  "rstrip": false,
@@ -73,7 +73,7 @@
73
  "special": true
74
  },
75
  "57526": {
76
- "content": "</s_referensi>",
77
  "lstrip": false,
78
  "normalized": false,
79
  "rstrip": false,
@@ -81,7 +81,7 @@
81
  "special": true
82
  },
83
  "57527": {
84
- "content": "<s_ppn>",
85
  "lstrip": false,
86
  "normalized": false,
87
  "rstrip": false,
@@ -89,7 +89,7 @@
89
  "special": true
90
  },
91
  "57528": {
92
- "content": "</s_ppn>",
93
  "lstrip": false,
94
  "normalized": false,
95
  "rstrip": false,
@@ -97,7 +97,7 @@
97
  "special": true
98
  },
99
  "57529": {
100
- "content": "<s_hargaTotal>",
101
  "lstrip": false,
102
  "normalized": false,
103
  "rstrip": false,
@@ -105,7 +105,7 @@
105
  "special": true
106
  },
107
  "57530": {
108
- "content": "</s_hargaTotal>",
109
  "lstrip": false,
110
  "normalized": false,
111
  "rstrip": false,
@@ -113,7 +113,7 @@
113
  "special": true
114
  },
115
  "57531": {
116
- "content": "<s_dpp>",
117
  "lstrip": false,
118
  "normalized": false,
119
  "rstrip": false,
@@ -121,7 +121,7 @@
121
  "special": true
122
  },
123
  "57532": {
124
- "content": "</s_dpp>",
125
  "lstrip": false,
126
  "normalized": false,
127
  "rstrip": false,
@@ -129,7 +129,7 @@
129
  "special": true
130
  },
131
  "57533": {
132
- "content": "<s_diskon>",
133
  "lstrip": false,
134
  "normalized": false,
135
  "rstrip": false,
@@ -137,7 +137,7 @@
137
  "special": true
138
  },
139
  "57534": {
140
- "content": "</s_diskon>",
141
  "lstrip": false,
142
  "normalized": false,
143
  "rstrip": false,
@@ -145,7 +145,7 @@
145
  "special": true
146
  },
147
  "57535": {
148
- "content": "<s_TanggalFakturPajak>",
149
  "lstrip": false,
150
  "normalized": false,
151
  "rstrip": false,
@@ -153,7 +153,7 @@
153
  "special": true
154
  },
155
  "57536": {
156
- "content": "</s_TanggalFakturPajak>",
157
  "lstrip": false,
158
  "normalized": false,
159
  "rstrip": false,
@@ -161,7 +161,7 @@
161
  "special": true
162
  },
163
  "57537": {
164
- "content": "<s_PPnBM>",
165
  "lstrip": false,
166
  "normalized": false,
167
  "rstrip": false,
@@ -169,7 +169,7 @@
169
  "special": true
170
  },
171
  "57538": {
172
- "content": "</s_PPnBM>",
173
  "lstrip": false,
174
  "normalized": false,
175
  "rstrip": false,
@@ -177,7 +177,7 @@
177
  "special": true
178
  },
179
  "57539": {
180
- "content": "<s_NoNPWPlawanTransaksi>",
181
  "lstrip": false,
182
  "normalized": false,
183
  "rstrip": false,
@@ -185,7 +185,7 @@
185
  "special": true
186
  },
187
  "57540": {
188
- "content": "</s_NoNPWPlawanTransaksi>",
189
  "lstrip": false,
190
  "normalized": false,
191
  "rstrip": false,
@@ -193,7 +193,7 @@
193
  "special": true
194
  },
195
  "57541": {
196
- "content": "<s_NoFakturPajak>",
197
  "lstrip": false,
198
  "normalized": false,
199
  "rstrip": false,
@@ -201,7 +201,7 @@
201
  "special": true
202
  },
203
  "57542": {
204
- "content": "</s_NoFakturPajak>",
205
  "lstrip": false,
206
  "normalized": false,
207
  "rstrip": false,
@@ -209,7 +209,7 @@
209
  "special": true
210
  },
211
  "57543": {
212
- "content": "<s_NamaPenjual>",
213
  "lstrip": false,
214
  "normalized": false,
215
  "rstrip": false,
@@ -217,7 +217,7 @@
217
  "special": true
218
  },
219
  "57544": {
220
- "content": "</s_NamaPenjual>",
221
  "lstrip": false,
222
  "normalized": false,
223
  "rstrip": false,
@@ -225,7 +225,7 @@
225
  "special": true
226
  },
227
  "57545": {
228
- "content": "<s_NPWPpenjual>",
229
  "lstrip": false,
230
  "normalized": false,
231
  "rstrip": false,
@@ -233,7 +233,7 @@
233
  "special": true
234
  },
235
  "57546": {
236
- "content": "</s_NPWPpenjual>",
237
  "lstrip": false,
238
  "normalized": false,
239
  "rstrip": false,
@@ -241,7 +241,7 @@
241
  "special": true
242
  },
243
  "57547": {
244
- "content": "<s_NPWPlawanTransaksi>",
245
  "lstrip": false,
246
  "normalized": false,
247
  "rstrip": false,
@@ -249,7 +249,7 @@
249
  "special": true
250
  },
251
  "57548": {
252
- "content": "</s_NPWPlawanTransaksi>",
253
  "lstrip": false,
254
  "normalized": false,
255
  "rstrip": false,
@@ -257,7 +257,7 @@
257
  "special": true
258
  },
259
  "57549": {
260
- "content": "<s_AlamatPenjual>",
261
  "lstrip": false,
262
  "normalized": false,
263
  "rstrip": false,
@@ -265,7 +265,7 @@
265
  "special": true
266
  },
267
  "57550": {
268
- "content": "</s_AlamatPenjual>",
269
  "lstrip": false,
270
  "normalized": false,
271
  "rstrip": false,
@@ -273,7 +273,7 @@
273
  "special": true
274
  },
275
  "57551": {
276
- "content": "<s_AlamatLawanTransaksi>",
277
  "lstrip": false,
278
  "normalized": false,
279
  "rstrip": false,
@@ -281,7 +281,23 @@
281
  "special": true
282
  },
283
  "57552": {
284
- "content": "</s_AlamatLawanTransaksi>",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
285
  "lstrip": false,
286
  "normalized": false,
287
  "rstrip": false,
@@ -290,34 +306,36 @@
290
  }
291
  },
292
  "additional_special_tokens": [
293
- "<s_referensi>",
294
- "</s_referensi>",
295
- "<s_ppn>",
296
- "</s_ppn>",
297
- "<s_hargaTotal>",
298
- "</s_hargaTotal>",
299
- "<s_dpp>",
300
- "</s_dpp>",
301
- "<s_diskon>",
302
- "</s_diskon>",
303
- "<s_TanggalFakturPajak>",
304
- "</s_TanggalFakturPajak>",
305
- "<s_PPnBM>",
306
- "</s_PPnBM>",
307
- "<s_NoNPWPlawanTransaksi>",
308
- "</s_NoNPWPlawanTransaksi>",
309
- "<s_NoFakturPajak>",
310
- "</s_NoFakturPajak>",
311
- "<s_NamaPenjual>",
312
- "</s_NamaPenjual>",
313
- "<s_NPWPpenjual>",
314
- "</s_NPWPpenjual>",
315
- "<s_NPWPlawanTransaksi>",
316
- "</s_NPWPlawanTransaksi>",
317
- "<s_AlamatPenjual>",
318
- "</s_AlamatPenjual>",
319
- "<s_AlamatLawanTransaksi>",
320
- "</s_AlamatLawanTransaksi>",
 
 
321
  "<s>",
322
  "</s>"
323
  ],
@@ -326,18 +344,11 @@
326
  "cls_token": "<s>",
327
  "eos_token": "</s>",
328
  "mask_token": "<mask>",
329
- "max_length": 512,
330
  "model_max_length": 1000000000000000019884624838656,
331
- "pad_to_multiple_of": null,
332
  "pad_token": "<pad>",
333
- "pad_token_type_id": 0,
334
- "padding_side": "right",
335
  "processor_class": "DonutProcessor",
336
  "sep_token": "</s>",
337
  "sp_model_kwargs": {},
338
- "stride": 0,
339
  "tokenizer_class": "XLMRobertaTokenizer",
340
- "truncation_side": "right",
341
- "truncation_strategy": "longest_first",
342
  "unk_token": "<unk>"
343
  }
 
65
  "special": true
66
  },
67
  "57525": {
68
+ "content": "<s_tanggal_invoice>",
69
  "lstrip": false,
70
  "normalized": false,
71
  "rstrip": false,
 
73
  "special": true
74
  },
75
  "57526": {
76
+ "content": "</s_tanggal_invoice>",
77
  "lstrip": false,
78
  "normalized": false,
79
  "rstrip": false,
 
81
  "special": true
82
  },
83
  "57527": {
84
+ "content": "<s_no_kontrak>",
85
  "lstrip": false,
86
  "normalized": false,
87
  "rstrip": false,
 
89
  "special": true
90
  },
91
  "57528": {
92
+ "content": "</s_no_kontrak>",
93
  "lstrip": false,
94
  "normalized": false,
95
  "rstrip": false,
 
97
  "special": true
98
  },
99
  "57529": {
100
+ "content": "<s_no_invoice>",
101
  "lstrip": false,
102
  "normalized": false,
103
  "rstrip": false,
 
105
  "special": true
106
  },
107
  "57530": {
108
+ "content": "</s_no_invoice>",
109
  "lstrip": false,
110
  "normalized": false,
111
  "rstrip": false,
 
113
  "special": true
114
  },
115
  "57531": {
116
+ "content": "<s_nama_vendor>",
117
  "lstrip": false,
118
  "normalized": false,
119
  "rstrip": false,
 
121
  "special": true
122
  },
123
  "57532": {
124
+ "content": "</s_nama_vendor>",
125
  "lstrip": false,
126
  "normalized": false,
127
  "rstrip": false,
 
129
  "special": true
130
  },
131
  "57533": {
132
+ "content": "<s_Type_Transaksi>",
133
  "lstrip": false,
134
  "normalized": false,
135
  "rstrip": false,
 
137
  "special": true
138
  },
139
  "57534": {
140
+ "content": "</s_Type_Transaksi>",
141
  "lstrip": false,
142
  "normalized": false,
143
  "rstrip": false,
 
145
  "special": true
146
  },
147
  "57535": {
148
+ "content": "<s_Type_Tax>",
149
  "lstrip": false,
150
  "normalized": false,
151
  "rstrip": false,
 
153
  "special": true
154
  },
155
  "57536": {
156
+ "content": "</s_Type_Tax>",
157
  "lstrip": false,
158
  "normalized": false,
159
  "rstrip": false,
 
161
  "special": true
162
  },
163
  "57537": {
164
+ "content": "<s_Total>",
165
  "lstrip": false,
166
  "normalized": false,
167
  "rstrip": false,
 
169
  "special": true
170
  },
171
  "57538": {
172
+ "content": "</s_Total>",
173
  "lstrip": false,
174
  "normalized": false,
175
  "rstrip": false,
 
177
  "special": true
178
  },
179
  "57539": {
180
+ "content": "<s_Tax>",
181
  "lstrip": false,
182
  "normalized": false,
183
  "rstrip": false,
 
185
  "special": true
186
  },
187
  "57540": {
188
+ "content": "</s_Tax>",
189
  "lstrip": false,
190
  "normalized": false,
191
  "rstrip": false,
 
193
  "special": true
194
  },
195
  "57541": {
196
+ "content": "<s_Periode>",
197
  "lstrip": false,
198
  "normalized": false,
199
  "rstrip": false,
 
201
  "special": true
202
  },
203
  "57542": {
204
+ "content": "</s_Periode>",
205
  "lstrip": false,
206
  "normalized": false,
207
  "rstrip": false,
 
209
  "special": true
210
  },
211
  "57543": {
212
+ "content": "<s_No_Rek>",
213
  "lstrip": false,
214
  "normalized": false,
215
  "rstrip": false,
 
217
  "special": true
218
  },
219
  "57544": {
220
+ "content": "</s_No_Rek>",
221
  "lstrip": false,
222
  "normalized": false,
223
  "rstrip": false,
 
225
  "special": true
226
  },
227
  "57545": {
228
+ "content": "<s_Nama_Rek>",
229
  "lstrip": false,
230
  "normalized": false,
231
  "rstrip": false,
 
233
  "special": true
234
  },
235
  "57546": {
236
+ "content": "</s_Nama_Rek>",
237
  "lstrip": false,
238
  "normalized": false,
239
  "rstrip": false,
 
241
  "special": true
242
  },
243
  "57547": {
244
+ "content": "<s_Nama_Bank>",
245
  "lstrip": false,
246
  "normalized": false,
247
  "rstrip": false,
 
249
  "special": true
250
  },
251
  "57548": {
252
+ "content": "</s_Nama_Bank>",
253
  "lstrip": false,
254
  "normalized": false,
255
  "rstrip": false,
 
257
  "special": true
258
  },
259
  "57549": {
260
+ "content": "<s_NILAI METERAI>",
261
  "lstrip": false,
262
  "normalized": false,
263
  "rstrip": false,
 
265
  "special": true
266
  },
267
  "57550": {
268
+ "content": "</s_NILAI METERAI>",
269
  "lstrip": false,
270
  "normalized": false,
271
  "rstrip": false,
 
273
  "special": true
274
  },
275
  "57551": {
276
+ "content": "<s_Meterai>",
277
  "lstrip": false,
278
  "normalized": false,
279
  "rstrip": false,
 
281
  "special": true
282
  },
283
  "57552": {
284
+ "content": "</s_Meterai>",
285
+ "lstrip": false,
286
+ "normalized": false,
287
+ "rstrip": false,
288
+ "single_word": false,
289
+ "special": true
290
+ },
291
+ "57553": {
292
+ "content": "<s_DPP>",
293
+ "lstrip": false,
294
+ "normalized": false,
295
+ "rstrip": false,
296
+ "single_word": false,
297
+ "special": true
298
+ },
299
+ "57554": {
300
+ "content": "</s_DPP>",
301
  "lstrip": false,
302
  "normalized": false,
303
  "rstrip": false,
 
306
  }
307
  },
308
  "additional_special_tokens": [
309
+ "<s_tanggal_invoice>",
310
+ "</s_tanggal_invoice>",
311
+ "<s_no_kontrak>",
312
+ "</s_no_kontrak>",
313
+ "<s_no_invoice>",
314
+ "</s_no_invoice>",
315
+ "<s_nama_vendor>",
316
+ "</s_nama_vendor>",
317
+ "<s_Type_Transaksi>",
318
+ "</s_Type_Transaksi>",
319
+ "<s_Type_Tax>",
320
+ "</s_Type_Tax>",
321
+ "<s_Total>",
322
+ "</s_Total>",
323
+ "<s_Tax>",
324
+ "</s_Tax>",
325
+ "<s_Periode>",
326
+ "</s_Periode>",
327
+ "<s_No_Rek>",
328
+ "</s_No_Rek>",
329
+ "<s_Nama_Rek>",
330
+ "</s_Nama_Rek>",
331
+ "<s_Nama_Bank>",
332
+ "</s_Nama_Bank>",
333
+ "<s_NILAI METERAI>",
334
+ "</s_NILAI METERAI>",
335
+ "<s_Meterai>",
336
+ "</s_Meterai>",
337
+ "<s_DPP>",
338
+ "</s_DPP>",
339
  "<s>",
340
  "</s>"
341
  ],
 
344
  "cls_token": "<s>",
345
  "eos_token": "</s>",
346
  "mask_token": "<mask>",
 
347
  "model_max_length": 1000000000000000019884624838656,
 
348
  "pad_token": "<pad>",
 
 
349
  "processor_class": "DonutProcessor",
350
  "sep_token": "</s>",
351
  "sp_model_kwargs": {},
 
352
  "tokenizer_class": "XLMRobertaTokenizer",
 
 
353
  "unk_token": "<unk>"
354
  }