Isotonic commited on
Commit
b30df2b
1 Parent(s): 5f3be0f

Training in progress, step 500

Browse files
.gitignore ADDED
@@ -0,0 +1 @@
 
 
1
+ checkpoint-*/
config.json ADDED
@@ -0,0 +1,260 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "distilbert-base-uncased",
3
+ "activation": "gelu",
4
+ "architectures": [
5
+ "DistilBertForTokenClassification"
6
+ ],
7
+ "attention_dropout": 0.1,
8
+ "dim": 768,
9
+ "dropout": 0.1,
10
+ "hidden_dim": 3072,
11
+ "id2label": {
12
+ "0": "B-PREFIX",
13
+ "1": "I-PREFIX",
14
+ "2": "B-FIRSTNAME",
15
+ "3": "I-FIRSTNAME",
16
+ "4": "B-MIDDLENAME",
17
+ "5": "B-LASTNAME",
18
+ "6": "I-LASTNAME",
19
+ "7": "O",
20
+ "8": "B-JOBDESCRIPTOR",
21
+ "9": "B-JOBTITLE",
22
+ "10": "I-JOBTITLE",
23
+ "11": "B-COMPANY_NAME",
24
+ "12": "I-COMPANY_NAME",
25
+ "13": "B-JOBAREA",
26
+ "14": "B-EMAIL",
27
+ "15": "I-EMAIL",
28
+ "16": "B-TIME",
29
+ "17": "I-TIME",
30
+ "18": "B-DATE",
31
+ "19": "I-DATE",
32
+ "20": "B-URL",
33
+ "21": "I-URL",
34
+ "22": "B-BITCOINADDRESS",
35
+ "23": "I-BITCOINADDRESS",
36
+ "24": "B-ETHEREUMADDRESS",
37
+ "25": "I-ETHEREUMADDRESS",
38
+ "26": "B-ACCOUNTNAME",
39
+ "27": "I-ACCOUNTNAME",
40
+ "28": "B-IBAN",
41
+ "29": "I-IBAN",
42
+ "30": "B-ACCOUNTNUMBER",
43
+ "31": "I-ACCOUNTNUMBER",
44
+ "32": "B-BIC",
45
+ "33": "I-BIC",
46
+ "34": "B-IPV4",
47
+ "35": "I-IPV4",
48
+ "36": "B-STREETADDRESS",
49
+ "37": "I-STREETADDRESS",
50
+ "38": "B-CITY",
51
+ "39": "I-CITY",
52
+ "40": "B-ZIPCODE",
53
+ "41": "I-ZIPCODE",
54
+ "42": "B-USERNAME",
55
+ "43": "I-USERNAME",
56
+ "44": "B-IPV6",
57
+ "45": "I-IPV6",
58
+ "46": "B-CREDITCARDNUMBER",
59
+ "47": "I-CREDITCARDNUMBER",
60
+ "48": "B-VEHICLEVIN",
61
+ "49": "I-VEHICLEVIN",
62
+ "50": "B-SUFFIX",
63
+ "51": "I-SUFFIX",
64
+ "52": "B-AMOUNT",
65
+ "53": "I-AMOUNT",
66
+ "54": "B-CURRENCY",
67
+ "55": "I-CURRENCY",
68
+ "56": "B-PASSWORD",
69
+ "57": "I-PASSWORD",
70
+ "58": "B-JOBTYPE",
71
+ "59": "B-STATE",
72
+ "60": "B-BUILDINGNUMBER",
73
+ "61": "I-BUILDINGNUMBER",
74
+ "62": "B-VEHICLEVRM",
75
+ "63": "I-VEHICLEVRM",
76
+ "64": "B-PHONEIMEI",
77
+ "65": "I-PHONEIMEI",
78
+ "66": "I-JOBAREA",
79
+ "67": "I-STATE",
80
+ "68": "B-COUNTY",
81
+ "69": "B-CURRENCYNAME",
82
+ "70": "I-CURRENCYNAME",
83
+ "71": "B-CURRENCYSYMBOL",
84
+ "72": "B-MASKEDNUMBER",
85
+ "73": "I-MASKEDNUMBER",
86
+ "74": "B-PHONE_NUMBER",
87
+ "75": "I-PHONE_NUMBER",
88
+ "76": "B-SECONDARYADDRESS",
89
+ "77": "I-SECONDARYADDRESS",
90
+ "78": "B-SSN",
91
+ "79": "I-SSN",
92
+ "80": "B-CURRENCYCODE",
93
+ "81": "B-LITECOINADDRESS",
94
+ "82": "I-LITECOINADDRESS",
95
+ "83": "B-MAC",
96
+ "84": "I-MAC",
97
+ "85": "B-CREDITCARDISSUER",
98
+ "86": "I-CREDITCARDISSUER",
99
+ "87": "B-CREDITCARDCVV",
100
+ "88": "I-CREDITCARDCVV",
101
+ "89": "B-USERAGENT",
102
+ "90": "I-USERAGENT",
103
+ "91": "B-IP",
104
+ "92": "I-IP",
105
+ "93": "B-SEX",
106
+ "94": "B-STREET",
107
+ "95": "I-STREET",
108
+ "96": "B-PIN",
109
+ "97": "I-PIN",
110
+ "98": "I-JOBTYPE",
111
+ "99": "I-MIDDLENAME",
112
+ "100": "I-CURRENCYCODE",
113
+ "101": "I-CURRENCYSYMBOL",
114
+ "102": "B-FULLNAME",
115
+ "103": "I-FULLNAME",
116
+ "104": "B-NAME",
117
+ "105": "I-NAME",
118
+ "106": "B-GENDER",
119
+ "107": "B-NUMBER",
120
+ "108": "I-NUMBER",
121
+ "109": "I-GENDER",
122
+ "110": "B-NEARBYGPSCOORDINATE",
123
+ "111": "I-NEARBYGPSCOORDINATE",
124
+ "112": "B-DISPLAYNAME",
125
+ "113": "I-DISPLAYNAME",
126
+ "114": "B-SEXTYPE",
127
+ "115": "B-ORDINALDIRECTION"
128
+ },
129
+ "initializer_range": 0.02,
130
+ "label2id": {
131
+ "B-ACCOUNTNAME": 26,
132
+ "B-ACCOUNTNUMBER": 30,
133
+ "B-AMOUNT": 52,
134
+ "B-BIC": 32,
135
+ "B-BITCOINADDRESS": 22,
136
+ "B-BUILDINGNUMBER": 60,
137
+ "B-CITY": 38,
138
+ "B-COMPANY_NAME": 11,
139
+ "B-COUNTY": 68,
140
+ "B-CREDITCARDCVV": 87,
141
+ "B-CREDITCARDISSUER": 85,
142
+ "B-CREDITCARDNUMBER": 46,
143
+ "B-CURRENCY": 54,
144
+ "B-CURRENCYCODE": 80,
145
+ "B-CURRENCYNAME": 69,
146
+ "B-CURRENCYSYMBOL": 71,
147
+ "B-DATE": 18,
148
+ "B-DISPLAYNAME": 112,
149
+ "B-EMAIL": 14,
150
+ "B-ETHEREUMADDRESS": 24,
151
+ "B-FIRSTNAME": 2,
152
+ "B-FULLNAME": 102,
153
+ "B-GENDER": 106,
154
+ "B-IBAN": 28,
155
+ "B-IP": 91,
156
+ "B-IPV4": 34,
157
+ "B-IPV6": 44,
158
+ "B-JOBAREA": 13,
159
+ "B-JOBDESCRIPTOR": 8,
160
+ "B-JOBTITLE": 9,
161
+ "B-JOBTYPE": 58,
162
+ "B-LASTNAME": 5,
163
+ "B-LITECOINADDRESS": 81,
164
+ "B-MAC": 83,
165
+ "B-MASKEDNUMBER": 72,
166
+ "B-MIDDLENAME": 4,
167
+ "B-NAME": 104,
168
+ "B-NEARBYGPSCOORDINATE": 110,
169
+ "B-NUMBER": 107,
170
+ "B-ORDINALDIRECTION": 115,
171
+ "B-PASSWORD": 56,
172
+ "B-PHONEIMEI": 64,
173
+ "B-PHONE_NUMBER": 74,
174
+ "B-PIN": 96,
175
+ "B-PREFIX": 0,
176
+ "B-SECONDARYADDRESS": 76,
177
+ "B-SEX": 93,
178
+ "B-SEXTYPE": 114,
179
+ "B-SSN": 78,
180
+ "B-STATE": 59,
181
+ "B-STREET": 94,
182
+ "B-STREETADDRESS": 36,
183
+ "B-SUFFIX": 50,
184
+ "B-TIME": 16,
185
+ "B-URL": 20,
186
+ "B-USERAGENT": 89,
187
+ "B-USERNAME": 42,
188
+ "B-VEHICLEVIN": 48,
189
+ "B-VEHICLEVRM": 62,
190
+ "B-ZIPCODE": 40,
191
+ "I-ACCOUNTNAME": 27,
192
+ "I-ACCOUNTNUMBER": 31,
193
+ "I-AMOUNT": 53,
194
+ "I-BIC": 33,
195
+ "I-BITCOINADDRESS": 23,
196
+ "I-BUILDINGNUMBER": 61,
197
+ "I-CITY": 39,
198
+ "I-COMPANY_NAME": 12,
199
+ "I-CREDITCARDCVV": 88,
200
+ "I-CREDITCARDISSUER": 86,
201
+ "I-CREDITCARDNUMBER": 47,
202
+ "I-CURRENCY": 55,
203
+ "I-CURRENCYCODE": 100,
204
+ "I-CURRENCYNAME": 70,
205
+ "I-CURRENCYSYMBOL": 101,
206
+ "I-DATE": 19,
207
+ "I-DISPLAYNAME": 113,
208
+ "I-EMAIL": 15,
209
+ "I-ETHEREUMADDRESS": 25,
210
+ "I-FIRSTNAME": 3,
211
+ "I-FULLNAME": 103,
212
+ "I-GENDER": 109,
213
+ "I-IBAN": 29,
214
+ "I-IP": 92,
215
+ "I-IPV4": 35,
216
+ "I-IPV6": 45,
217
+ "I-JOBAREA": 66,
218
+ "I-JOBTITLE": 10,
219
+ "I-JOBTYPE": 98,
220
+ "I-LASTNAME": 6,
221
+ "I-LITECOINADDRESS": 82,
222
+ "I-MAC": 84,
223
+ "I-MASKEDNUMBER": 73,
224
+ "I-MIDDLENAME": 99,
225
+ "I-NAME": 105,
226
+ "I-NEARBYGPSCOORDINATE": 111,
227
+ "I-NUMBER": 108,
228
+ "I-PASSWORD": 57,
229
+ "I-PHONEIMEI": 65,
230
+ "I-PHONE_NUMBER": 75,
231
+ "I-PIN": 97,
232
+ "I-PREFIX": 1,
233
+ "I-SECONDARYADDRESS": 77,
234
+ "I-SSN": 79,
235
+ "I-STATE": 67,
236
+ "I-STREET": 95,
237
+ "I-STREETADDRESS": 37,
238
+ "I-SUFFIX": 51,
239
+ "I-TIME": 17,
240
+ "I-URL": 21,
241
+ "I-USERAGENT": 90,
242
+ "I-USERNAME": 43,
243
+ "I-VEHICLEVIN": 49,
244
+ "I-VEHICLEVRM": 63,
245
+ "I-ZIPCODE": 41,
246
+ "O": 7
247
+ },
248
+ "max_position_embeddings": 512,
249
+ "model_type": "distilbert",
250
+ "n_heads": 12,
251
+ "n_layers": 6,
252
+ "pad_token_id": 0,
253
+ "qa_dropout": 0.1,
254
+ "seq_classif_dropout": 0.2,
255
+ "sinusoidal_pos_embds": false,
256
+ "tie_weights_": true,
257
+ "torch_dtype": "float32",
258
+ "transformers_version": "4.31.0",
259
+ "vocab_size": 30522
260
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d693e8ad7a3acf6d30f2029728bec6a7fad75e6c93204217b840e8bb86619a5f
3
+ size 265843173
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 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "clean_up_tokenization_spaces": true,
3
+ "cls_token": "[CLS]",
4
+ "do_lower_case": true,
5
+ "mask_token": "[MASK]",
6
+ "model_max_length": 512,
7
+ "pad_token": "[PAD]",
8
+ "sep_token": "[SEP]",
9
+ "strip_accents": null,
10
+ "tokenize_chinese_chars": true,
11
+ "tokenizer_class": "DistilBertTokenizer",
12
+ "unk_token": "[UNK]"
13
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c78f5eefe873f13228c3a9de122ce380b1553b0728c7c5301c6770dc9bbbf282
3
+ size 4027
vocab.txt ADDED
The diff for this file is too large to render. See raw diff