Training in progress, epoch 1
Browse files- config.json +373 -0
- model.safetensors +3 -0
- runs/Jul26_10-53-23_T2P-CPU010/events.out.tfevents.1721966005.T2P-CPU010.3893114.0 +3 -0
- special_tokens_map.json +7 -0
- tokenizer.json +0 -0
- tokenizer_config.json +55 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
config.json
ADDED
@@ -0,0 +1,373 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "bert-base-multilingual-cased",
|
3 |
+
"architectures": [
|
4 |
+
"BertForTokenClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"directionality": "bidi",
|
9 |
+
"hidden_act": "gelu",
|
10 |
+
"hidden_dropout_prob": 0.1,
|
11 |
+
"hidden_size": 768,
|
12 |
+
"id2label": {
|
13 |
+
"0": "O",
|
14 |
+
"1": "B-0",
|
15 |
+
"2": "I-0",
|
16 |
+
"3": "E-0",
|
17 |
+
"4": "S-0",
|
18 |
+
"5": "B-arrivalTrainNo",
|
19 |
+
"6": "I-arrivalTrainNo",
|
20 |
+
"7": "E-arrivalTrainNo",
|
21 |
+
"8": "S-arrivalTrainNo",
|
22 |
+
"9": "B-bookingNumber",
|
23 |
+
"10": "I-bookingNumber",
|
24 |
+
"11": "E-bookingNumber",
|
25 |
+
"12": "S-bookingNumber",
|
26 |
+
"13": "B-cardNumber",
|
27 |
+
"14": "I-cardNumber",
|
28 |
+
"15": "E-cardNumber",
|
29 |
+
"16": "S-cardNumber",
|
30 |
+
"17": "B-cardType",
|
31 |
+
"18": "I-cardType",
|
32 |
+
"19": "E-cardType",
|
33 |
+
"20": "S-cardType",
|
34 |
+
"21": "B-checkIn",
|
35 |
+
"22": "I-checkIn",
|
36 |
+
"23": "E-checkIn",
|
37 |
+
"24": "S-checkIn",
|
38 |
+
"25": "B-checkOut",
|
39 |
+
"26": "I-checkOut",
|
40 |
+
"27": "E-checkOut",
|
41 |
+
"28": "S-checkOut",
|
42 |
+
"29": "B-companyAddress",
|
43 |
+
"30": "I-companyAddress",
|
44 |
+
"31": "E-companyAddress",
|
45 |
+
"32": "S-companyAddress",
|
46 |
+
"33": "B-companyCity",
|
47 |
+
"34": "I-companyCity",
|
48 |
+
"35": "E-companyCity",
|
49 |
+
"36": "S-companyCity",
|
50 |
+
"37": "B-companyCountry",
|
51 |
+
"38": "I-companyCountry",
|
52 |
+
"39": "E-companyCountry",
|
53 |
+
"40": "S-companyCountry",
|
54 |
+
"41": "B-companyName",
|
55 |
+
"42": "I-companyName",
|
56 |
+
"43": "E-companyName",
|
57 |
+
"44": "S-companyName",
|
58 |
+
"45": "B-companyVATNumber",
|
59 |
+
"46": "I-companyVATNumber",
|
60 |
+
"47": "E-companyVATNumber",
|
61 |
+
"48": "S-companyVATNumber",
|
62 |
+
"49": "B-companyZipcode",
|
63 |
+
"50": "I-companyZipcode",
|
64 |
+
"51": "E-companyZipcode",
|
65 |
+
"52": "S-companyZipcode",
|
66 |
+
"53": "B-crewType",
|
67 |
+
"54": "I-crewType",
|
68 |
+
"55": "E-crewType",
|
69 |
+
"56": "S-crewType",
|
70 |
+
"57": "B-currencyCode",
|
71 |
+
"58": "I-currencyCode",
|
72 |
+
"59": "E-currencyCode",
|
73 |
+
"60": "S-currencyCode",
|
74 |
+
"61": "B-documentType",
|
75 |
+
"62": "I-documentType",
|
76 |
+
"63": "E-documentType",
|
77 |
+
"64": "S-documentType",
|
78 |
+
"65": "B-externalReservationNumber",
|
79 |
+
"66": "I-externalReservationNumber",
|
80 |
+
"67": "E-externalReservationNumber",
|
81 |
+
"68": "S-externalReservationNumber",
|
82 |
+
"69": "B-grossAmount",
|
83 |
+
"70": "I-grossAmount",
|
84 |
+
"71": "E-grossAmount",
|
85 |
+
"72": "S-grossAmount",
|
86 |
+
"73": "B-hotelAddress",
|
87 |
+
"74": "I-hotelAddress",
|
88 |
+
"75": "E-hotelAddress",
|
89 |
+
"76": "S-hotelAddress",
|
90 |
+
"77": "B-hotelBankAccount",
|
91 |
+
"78": "I-hotelBankAccount",
|
92 |
+
"79": "E-hotelBankAccount",
|
93 |
+
"80": "S-hotelBankAccount",
|
94 |
+
"81": "B-hotelCity",
|
95 |
+
"82": "I-hotelCity",
|
96 |
+
"83": "E-hotelCity",
|
97 |
+
"84": "S-hotelCity",
|
98 |
+
"85": "B-hotelCountry",
|
99 |
+
"86": "I-hotelCountry",
|
100 |
+
"87": "E-hotelCountry",
|
101 |
+
"88": "S-hotelCountry",
|
102 |
+
"89": "B-hotelEmail",
|
103 |
+
"90": "I-hotelEmail",
|
104 |
+
"91": "E-hotelEmail",
|
105 |
+
"92": "S-hotelEmail",
|
106 |
+
"93": "B-hotelFaxNumber",
|
107 |
+
"94": "I-hotelFaxNumber",
|
108 |
+
"95": "E-hotelFaxNumber",
|
109 |
+
"96": "S-hotelFaxNumber",
|
110 |
+
"97": "B-hotelName",
|
111 |
+
"98": "I-hotelName",
|
112 |
+
"99": "E-hotelName",
|
113 |
+
"100": "S-hotelName",
|
114 |
+
"101": "B-hotelPhoneNumber",
|
115 |
+
"102": "I-hotelPhoneNumber",
|
116 |
+
"103": "E-hotelPhoneNumber",
|
117 |
+
"104": "S-hotelPhoneNumber",
|
118 |
+
"105": "B-hotelTaxCode",
|
119 |
+
"106": "I-hotelTaxCode",
|
120 |
+
"107": "E-hotelTaxCode",
|
121 |
+
"108": "S-hotelTaxCode",
|
122 |
+
"109": "B-hotelVATNumber",
|
123 |
+
"110": "I-hotelVATNumber",
|
124 |
+
"111": "E-hotelVATNumber",
|
125 |
+
"112": "S-hotelVATNumber",
|
126 |
+
"113": "B-hotelZipCode",
|
127 |
+
"114": "I-hotelZipCode",
|
128 |
+
"115": "E-hotelZipCode",
|
129 |
+
"116": "S-hotelZipCode",
|
130 |
+
"117": "B-invoiceDate",
|
131 |
+
"118": "I-invoiceDate",
|
132 |
+
"119": "E-invoiceDate",
|
133 |
+
"120": "S-invoiceDate",
|
134 |
+
"121": "B-invoiceNumber",
|
135 |
+
"122": "I-invoiceNumber",
|
136 |
+
"123": "E-invoiceNumber",
|
137 |
+
"124": "S-invoiceNumber",
|
138 |
+
"125": "B-netAmount",
|
139 |
+
"126": "I-netAmount",
|
140 |
+
"127": "E-netAmount",
|
141 |
+
"128": "S-netAmount",
|
142 |
+
"129": "B-pageNumber",
|
143 |
+
"130": "I-pageNumber",
|
144 |
+
"131": "E-pageNumber",
|
145 |
+
"132": "S-pageNumber",
|
146 |
+
"133": "B-reservationNumber",
|
147 |
+
"134": "I-reservationNumber",
|
148 |
+
"135": "E-reservationNumber",
|
149 |
+
"136": "S-reservationNumber",
|
150 |
+
"137": "B-roomNo",
|
151 |
+
"138": "I-roomNo",
|
152 |
+
"139": "E-roomNo",
|
153 |
+
"140": "S-roomNo",
|
154 |
+
"141": "B-taxAmount",
|
155 |
+
"142": "I-taxAmount",
|
156 |
+
"143": "E-taxAmount",
|
157 |
+
"144": "S-taxAmount",
|
158 |
+
"145": "B-trafficDate",
|
159 |
+
"146": "I-trafficDate",
|
160 |
+
"147": "E-trafficDate",
|
161 |
+
"148": "S-trafficDate",
|
162 |
+
"149": "B-travellerName",
|
163 |
+
"150": "I-travellerName",
|
164 |
+
"151": "E-travellerName",
|
165 |
+
"152": "S-travellerName",
|
166 |
+
"153": "B-Bank_Account_bankInfo",
|
167 |
+
"154": "I-Bank_Account_bankInfo",
|
168 |
+
"155": "E-Bank_Account_bankInfo",
|
169 |
+
"156": "S-Bank_Account_bankInfo",
|
170 |
+
"157": "B-Bank_Account_IBAN",
|
171 |
+
"158": "I-Bank_Account_IBAN",
|
172 |
+
"159": "E-Bank_Account_IBAN",
|
173 |
+
"160": "S-Bank_Account_IBAN",
|
174 |
+
"161": "B-Bank_Account_BIC",
|
175 |
+
"162": "I-Bank_Account_BIC",
|
176 |
+
"163": "E-Bank_Account_BIC",
|
177 |
+
"164": "S-Bank_Account_BIC",
|
178 |
+
"165": "B-Bank_Account_currencyCode",
|
179 |
+
"166": "I-Bank_Account_currencyCode",
|
180 |
+
"167": "E-Bank_Account_currencyCode",
|
181 |
+
"168": "S-Bank_Account_currencyCode"
|
182 |
+
},
|
183 |
+
"initializer_range": 0.02,
|
184 |
+
"intermediate_size": 3072,
|
185 |
+
"label2id": {
|
186 |
+
"B-0": 1,
|
187 |
+
"B-Bank_Account_BIC": 161,
|
188 |
+
"B-Bank_Account_IBAN": 157,
|
189 |
+
"B-Bank_Account_bankInfo": 153,
|
190 |
+
"B-Bank_Account_currencyCode": 165,
|
191 |
+
"B-arrivalTrainNo": 5,
|
192 |
+
"B-bookingNumber": 9,
|
193 |
+
"B-cardNumber": 13,
|
194 |
+
"B-cardType": 17,
|
195 |
+
"B-checkIn": 21,
|
196 |
+
"B-checkOut": 25,
|
197 |
+
"B-companyAddress": 29,
|
198 |
+
"B-companyCity": 33,
|
199 |
+
"B-companyCountry": 37,
|
200 |
+
"B-companyName": 41,
|
201 |
+
"B-companyVATNumber": 45,
|
202 |
+
"B-companyZipcode": 49,
|
203 |
+
"B-crewType": 53,
|
204 |
+
"B-currencyCode": 57,
|
205 |
+
"B-documentType": 61,
|
206 |
+
"B-externalReservationNumber": 65,
|
207 |
+
"B-grossAmount": 69,
|
208 |
+
"B-hotelAddress": 73,
|
209 |
+
"B-hotelBankAccount": 77,
|
210 |
+
"B-hotelCity": 81,
|
211 |
+
"B-hotelCountry": 85,
|
212 |
+
"B-hotelEmail": 89,
|
213 |
+
"B-hotelFaxNumber": 93,
|
214 |
+
"B-hotelName": 97,
|
215 |
+
"B-hotelPhoneNumber": 101,
|
216 |
+
"B-hotelTaxCode": 105,
|
217 |
+
"B-hotelVATNumber": 109,
|
218 |
+
"B-hotelZipCode": 113,
|
219 |
+
"B-invoiceDate": 117,
|
220 |
+
"B-invoiceNumber": 121,
|
221 |
+
"B-netAmount": 125,
|
222 |
+
"B-pageNumber": 129,
|
223 |
+
"B-reservationNumber": 133,
|
224 |
+
"B-roomNo": 137,
|
225 |
+
"B-taxAmount": 141,
|
226 |
+
"B-trafficDate": 145,
|
227 |
+
"B-travellerName": 149,
|
228 |
+
"E-0": 3,
|
229 |
+
"E-Bank_Account_BIC": 163,
|
230 |
+
"E-Bank_Account_IBAN": 159,
|
231 |
+
"E-Bank_Account_bankInfo": 155,
|
232 |
+
"E-Bank_Account_currencyCode": 167,
|
233 |
+
"E-arrivalTrainNo": 7,
|
234 |
+
"E-bookingNumber": 11,
|
235 |
+
"E-cardNumber": 15,
|
236 |
+
"E-cardType": 19,
|
237 |
+
"E-checkIn": 23,
|
238 |
+
"E-checkOut": 27,
|
239 |
+
"E-companyAddress": 31,
|
240 |
+
"E-companyCity": 35,
|
241 |
+
"E-companyCountry": 39,
|
242 |
+
"E-companyName": 43,
|
243 |
+
"E-companyVATNumber": 47,
|
244 |
+
"E-companyZipcode": 51,
|
245 |
+
"E-crewType": 55,
|
246 |
+
"E-currencyCode": 59,
|
247 |
+
"E-documentType": 63,
|
248 |
+
"E-externalReservationNumber": 67,
|
249 |
+
"E-grossAmount": 71,
|
250 |
+
"E-hotelAddress": 75,
|
251 |
+
"E-hotelBankAccount": 79,
|
252 |
+
"E-hotelCity": 83,
|
253 |
+
"E-hotelCountry": 87,
|
254 |
+
"E-hotelEmail": 91,
|
255 |
+
"E-hotelFaxNumber": 95,
|
256 |
+
"E-hotelName": 99,
|
257 |
+
"E-hotelPhoneNumber": 103,
|
258 |
+
"E-hotelTaxCode": 107,
|
259 |
+
"E-hotelVATNumber": 111,
|
260 |
+
"E-hotelZipCode": 115,
|
261 |
+
"E-invoiceDate": 119,
|
262 |
+
"E-invoiceNumber": 123,
|
263 |
+
"E-netAmount": 127,
|
264 |
+
"E-pageNumber": 131,
|
265 |
+
"E-reservationNumber": 135,
|
266 |
+
"E-roomNo": 139,
|
267 |
+
"E-taxAmount": 143,
|
268 |
+
"E-trafficDate": 147,
|
269 |
+
"E-travellerName": 151,
|
270 |
+
"I-0": 2,
|
271 |
+
"I-Bank_Account_BIC": 162,
|
272 |
+
"I-Bank_Account_IBAN": 158,
|
273 |
+
"I-Bank_Account_bankInfo": 154,
|
274 |
+
"I-Bank_Account_currencyCode": 166,
|
275 |
+
"I-arrivalTrainNo": 6,
|
276 |
+
"I-bookingNumber": 10,
|
277 |
+
"I-cardNumber": 14,
|
278 |
+
"I-cardType": 18,
|
279 |
+
"I-checkIn": 22,
|
280 |
+
"I-checkOut": 26,
|
281 |
+
"I-companyAddress": 30,
|
282 |
+
"I-companyCity": 34,
|
283 |
+
"I-companyCountry": 38,
|
284 |
+
"I-companyName": 42,
|
285 |
+
"I-companyVATNumber": 46,
|
286 |
+
"I-companyZipcode": 50,
|
287 |
+
"I-crewType": 54,
|
288 |
+
"I-currencyCode": 58,
|
289 |
+
"I-documentType": 62,
|
290 |
+
"I-externalReservationNumber": 66,
|
291 |
+
"I-grossAmount": 70,
|
292 |
+
"I-hotelAddress": 74,
|
293 |
+
"I-hotelBankAccount": 78,
|
294 |
+
"I-hotelCity": 82,
|
295 |
+
"I-hotelCountry": 86,
|
296 |
+
"I-hotelEmail": 90,
|
297 |
+
"I-hotelFaxNumber": 94,
|
298 |
+
"I-hotelName": 98,
|
299 |
+
"I-hotelPhoneNumber": 102,
|
300 |
+
"I-hotelTaxCode": 106,
|
301 |
+
"I-hotelVATNumber": 110,
|
302 |
+
"I-hotelZipCode": 114,
|
303 |
+
"I-invoiceDate": 118,
|
304 |
+
"I-invoiceNumber": 122,
|
305 |
+
"I-netAmount": 126,
|
306 |
+
"I-pageNumber": 130,
|
307 |
+
"I-reservationNumber": 134,
|
308 |
+
"I-roomNo": 138,
|
309 |
+
"I-taxAmount": 142,
|
310 |
+
"I-trafficDate": 146,
|
311 |
+
"I-travellerName": 150,
|
312 |
+
"O": 0,
|
313 |
+
"S-0": 4,
|
314 |
+
"S-Bank_Account_BIC": 164,
|
315 |
+
"S-Bank_Account_IBAN": 160,
|
316 |
+
"S-Bank_Account_bankInfo": 156,
|
317 |
+
"S-Bank_Account_currencyCode": 168,
|
318 |
+
"S-arrivalTrainNo": 8,
|
319 |
+
"S-bookingNumber": 12,
|
320 |
+
"S-cardNumber": 16,
|
321 |
+
"S-cardType": 20,
|
322 |
+
"S-checkIn": 24,
|
323 |
+
"S-checkOut": 28,
|
324 |
+
"S-companyAddress": 32,
|
325 |
+
"S-companyCity": 36,
|
326 |
+
"S-companyCountry": 40,
|
327 |
+
"S-companyName": 44,
|
328 |
+
"S-companyVATNumber": 48,
|
329 |
+
"S-companyZipcode": 52,
|
330 |
+
"S-crewType": 56,
|
331 |
+
"S-currencyCode": 60,
|
332 |
+
"S-documentType": 64,
|
333 |
+
"S-externalReservationNumber": 68,
|
334 |
+
"S-grossAmount": 72,
|
335 |
+
"S-hotelAddress": 76,
|
336 |
+
"S-hotelBankAccount": 80,
|
337 |
+
"S-hotelCity": 84,
|
338 |
+
"S-hotelCountry": 88,
|
339 |
+
"S-hotelEmail": 92,
|
340 |
+
"S-hotelFaxNumber": 96,
|
341 |
+
"S-hotelName": 100,
|
342 |
+
"S-hotelPhoneNumber": 104,
|
343 |
+
"S-hotelTaxCode": 108,
|
344 |
+
"S-hotelVATNumber": 112,
|
345 |
+
"S-hotelZipCode": 116,
|
346 |
+
"S-invoiceDate": 120,
|
347 |
+
"S-invoiceNumber": 124,
|
348 |
+
"S-netAmount": 128,
|
349 |
+
"S-pageNumber": 132,
|
350 |
+
"S-reservationNumber": 136,
|
351 |
+
"S-roomNo": 140,
|
352 |
+
"S-taxAmount": 144,
|
353 |
+
"S-trafficDate": 148,
|
354 |
+
"S-travellerName": 152
|
355 |
+
},
|
356 |
+
"layer_norm_eps": 1e-12,
|
357 |
+
"max_position_embeddings": 512,
|
358 |
+
"model_type": "bert",
|
359 |
+
"num_attention_heads": 12,
|
360 |
+
"num_hidden_layers": 12,
|
361 |
+
"pad_token_id": 0,
|
362 |
+
"pooler_fc_size": 768,
|
363 |
+
"pooler_num_attention_heads": 12,
|
364 |
+
"pooler_num_fc_layers": 3,
|
365 |
+
"pooler_size_per_head": 128,
|
366 |
+
"pooler_type": "first_token_transform",
|
367 |
+
"position_embedding_type": "absolute",
|
368 |
+
"torch_dtype": "float32",
|
369 |
+
"transformers_version": "4.40.2",
|
370 |
+
"type_vocab_size": 2,
|
371 |
+
"use_cache": true,
|
372 |
+
"vocab_size": 119547
|
373 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1e4a0edb6efafa27db8cf23827c446674d1b82b142f51783e15525d2c56aa2a0
|
3 |
+
size 709594596
|
runs/Jul26_10-53-23_T2P-CPU010/events.out.tfevents.1721966005.T2P-CPU010.3893114.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b117b6bf8f6e95746b0b51c7ec730898d2b3ca5497731a2aeb7b5960782da3a8
|
3 |
+
size 46764
|
special_tokens_map.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"mask_token": "[MASK]",
|
4 |
+
"pad_token": "[PAD]",
|
5 |
+
"sep_token": "[SEP]",
|
6 |
+
"unk_token": "[UNK]"
|
7 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"100": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"101": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"102": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"103": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_lower_case": false,
|
47 |
+
"mask_token": "[MASK]",
|
48 |
+
"model_max_length": 512,
|
49 |
+
"pad_token": "[PAD]",
|
50 |
+
"sep_token": "[SEP]",
|
51 |
+
"strip_accents": null,
|
52 |
+
"tokenize_chinese_chars": true,
|
53 |
+
"tokenizer_class": "BertTokenizer",
|
54 |
+
"unk_token": "[UNK]"
|
55 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:60fa943e286cb992987e7531ffe762f674cf039af88f7d1ea2b6c362f5260811
|
3 |
+
size 5048
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|