williamkgao commited on
Commit
5009022
·
verified ·
1 Parent(s): e9c1472

Upload 6 files

Browse files
config.json ADDED
@@ -0,0 +1,263 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/mobilebert-uncased",
3
+ "architectures": [
4
+ "MobileBertForTokenClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_activation": false,
8
+ "classifier_dropout": null,
9
+ "embedding_size": 128,
10
+ "hidden_act": "relu",
11
+ "hidden_dropout_prob": 0.0,
12
+ "hidden_size": 512,
13
+ "id2label": {
14
+ "0": "I-EMAIL",
15
+ "1": "B-PHONEIMEI",
16
+ "2": "I-COUNTY",
17
+ "3": "I-URL",
18
+ "4": "I-ZIPCODE",
19
+ "5": "I-JOBTITLE",
20
+ "6": "I-PIN",
21
+ "7": "B-AGE",
22
+ "8": "I-BUILDINGNUMBER",
23
+ "9": "I-DATE",
24
+ "10": "B-GENDER",
25
+ "11": "B-PHONENUMBER",
26
+ "12": "B-PIN",
27
+ "13": "I-CURRENCYSYMBOL",
28
+ "14": "I-CREDITCARDISSUER",
29
+ "15": "I-CITY",
30
+ "16": "B-EYECOLOR",
31
+ "17": "B-DATE",
32
+ "18": "B-CREDITCARDISSUER",
33
+ "19": "I-ORDINALDIRECTION",
34
+ "20": "I-ETHEREUMADDRESS",
35
+ "21": "O",
36
+ "22": "B-VEHICLEVRM",
37
+ "23": "I-COMPANYNAME",
38
+ "24": "B-CITY",
39
+ "25": "B-BITCOINADDRESS",
40
+ "26": "I-SECONDARYADDRESS",
41
+ "27": "B-FIRSTNAME",
42
+ "28": "B-JOBTITLE",
43
+ "29": "I-MAC",
44
+ "30": "B-COUNTY",
45
+ "31": "I-BIC",
46
+ "32": "B-HEIGHT",
47
+ "33": "B-ETHEREUMADDRESS",
48
+ "34": "B-NEARBYGPSCOORDINATE",
49
+ "35": "I-LITECOINADDRESS",
50
+ "36": "I-MIDDLENAME",
51
+ "37": "I-PHONENUMBER",
52
+ "38": "B-CURRENCYSYMBOL",
53
+ "39": "B-LASTNAME",
54
+ "40": "I-STATE",
55
+ "41": "B-STATE",
56
+ "42": "B-JOBTYPE",
57
+ "43": "B-MAC",
58
+ "44": "I-LASTNAME",
59
+ "45": "B-USERAGENT",
60
+ "46": "B-PREFIX",
61
+ "47": "I-ACCOUNTNUMBER",
62
+ "48": "B-USERNAME",
63
+ "49": "B-EMAIL",
64
+ "50": "I-USERNAME",
65
+ "51": "I-ACCOUNTNAME",
66
+ "52": "I-CURRENCYNAME",
67
+ "53": "B-VEHICLEVIN",
68
+ "54": "B-IP",
69
+ "55": "B-CREDITCARDNUMBER",
70
+ "56": "B-CREDITCARDCVV",
71
+ "57": "I-CREDITCARDCVV",
72
+ "58": "B-CURRENCYCODE",
73
+ "59": "B-URL",
74
+ "60": "I-IPV6",
75
+ "61": "I-VEHICLEVRM",
76
+ "62": "B-TIME",
77
+ "63": "B-ACCOUNTNUMBER",
78
+ "64": "I-EYECOLOR",
79
+ "65": "B-JOBAREA",
80
+ "66": "I-JOBTYPE",
81
+ "67": "B-MASKEDNUMBER",
82
+ "68": "B-LITECOINADDRESS",
83
+ "69": "B-COMPANYNAME",
84
+ "70": "I-PREFIX",
85
+ "71": "B-BUILDINGNUMBER",
86
+ "72": "B-IPV4",
87
+ "73": "I-SSN",
88
+ "74": "I-TIME",
89
+ "75": "I-HEIGHT",
90
+ "76": "B-ACCOUNTNAME",
91
+ "77": "B-IBAN",
92
+ "78": "B-IPV6",
93
+ "79": "B-AMOUNT",
94
+ "80": "I-CREDITCARDNUMBER",
95
+ "81": "I-NEARBYGPSCOORDINATE",
96
+ "82": "B-ORDINALDIRECTION",
97
+ "83": "I-PASSWORD",
98
+ "84": "I-VEHICLEVIN",
99
+ "85": "I-AGE",
100
+ "86": "I-JOBAREA",
101
+ "87": "I-IP",
102
+ "88": "I-PHONEIMEI",
103
+ "89": "I-CURRENCY",
104
+ "90": "I-GENDER",
105
+ "91": "B-DOB",
106
+ "92": "I-USERAGENT",
107
+ "93": "I-CURRENCYCODE",
108
+ "94": "I-DOB",
109
+ "95": "B-ZIPCODE",
110
+ "96": "B-SECONDARYADDRESS",
111
+ "97": "B-STREET",
112
+ "98": "B-CURRENCY",
113
+ "99": "B-SSN",
114
+ "100": "I-STREET",
115
+ "101": "I-IPV4",
116
+ "102": "I-IBAN",
117
+ "103": "I-AMOUNT",
118
+ "104": "B-CURRENCYNAME",
119
+ "105": "I-SEX",
120
+ "106": "B-PASSWORD",
121
+ "107": "I-FIRSTNAME",
122
+ "108": "I-BITCOINADDRESS",
123
+ "109": "I-MASKEDNUMBER",
124
+ "110": "B-MIDDLENAME",
125
+ "111": "B-SEX",
126
+ "112": "B-BIC"
127
+ },
128
+ "initializer_range": 0.02,
129
+ "intermediate_size": 512,
130
+ "intra_bottleneck_size": 128,
131
+ "key_query_shared_bottleneck": true,
132
+ "label2id": {
133
+ "B-ACCOUNTNAME": 76,
134
+ "B-ACCOUNTNUMBER": 63,
135
+ "B-AGE": 7,
136
+ "B-AMOUNT": 79,
137
+ "B-BIC": 112,
138
+ "B-BITCOINADDRESS": 25,
139
+ "B-BUILDINGNUMBER": 71,
140
+ "B-CITY": 24,
141
+ "B-COMPANYNAME": 69,
142
+ "B-COUNTY": 30,
143
+ "B-CREDITCARDCVV": 56,
144
+ "B-CREDITCARDISSUER": 18,
145
+ "B-CREDITCARDNUMBER": 55,
146
+ "B-CURRENCY": 98,
147
+ "B-CURRENCYCODE": 58,
148
+ "B-CURRENCYNAME": 104,
149
+ "B-CURRENCYSYMBOL": 38,
150
+ "B-DATE": 17,
151
+ "B-DOB": 91,
152
+ "B-EMAIL": 49,
153
+ "B-ETHEREUMADDRESS": 33,
154
+ "B-EYECOLOR": 16,
155
+ "B-FIRSTNAME": 27,
156
+ "B-GENDER": 10,
157
+ "B-HEIGHT": 32,
158
+ "B-IBAN": 77,
159
+ "B-IP": 54,
160
+ "B-IPV4": 72,
161
+ "B-IPV6": 78,
162
+ "B-JOBAREA": 65,
163
+ "B-JOBTITLE": 28,
164
+ "B-JOBTYPE": 42,
165
+ "B-LASTNAME": 39,
166
+ "B-LITECOINADDRESS": 68,
167
+ "B-MAC": 43,
168
+ "B-MASKEDNUMBER": 67,
169
+ "B-MIDDLENAME": 110,
170
+ "B-NEARBYGPSCOORDINATE": 34,
171
+ "B-ORDINALDIRECTION": 82,
172
+ "B-PASSWORD": 106,
173
+ "B-PHONEIMEI": 1,
174
+ "B-PHONENUMBER": 11,
175
+ "B-PIN": 12,
176
+ "B-PREFIX": 46,
177
+ "B-SECONDARYADDRESS": 96,
178
+ "B-SEX": 111,
179
+ "B-SSN": 99,
180
+ "B-STATE": 41,
181
+ "B-STREET": 97,
182
+ "B-TIME": 62,
183
+ "B-URL": 59,
184
+ "B-USERAGENT": 45,
185
+ "B-USERNAME": 48,
186
+ "B-VEHICLEVIN": 53,
187
+ "B-VEHICLEVRM": 22,
188
+ "B-ZIPCODE": 95,
189
+ "I-ACCOUNTNAME": 51,
190
+ "I-ACCOUNTNUMBER": 47,
191
+ "I-AGE": 85,
192
+ "I-AMOUNT": 103,
193
+ "I-BIC": 31,
194
+ "I-BITCOINADDRESS": 108,
195
+ "I-BUILDINGNUMBER": 8,
196
+ "I-CITY": 15,
197
+ "I-COMPANYNAME": 23,
198
+ "I-COUNTY": 2,
199
+ "I-CREDITCARDCVV": 57,
200
+ "I-CREDITCARDISSUER": 14,
201
+ "I-CREDITCARDNUMBER": 80,
202
+ "I-CURRENCY": 89,
203
+ "I-CURRENCYCODE": 93,
204
+ "I-CURRENCYNAME": 52,
205
+ "I-CURRENCYSYMBOL": 13,
206
+ "I-DATE": 9,
207
+ "I-DOB": 94,
208
+ "I-EMAIL": 0,
209
+ "I-ETHEREUMADDRESS": 20,
210
+ "I-EYECOLOR": 64,
211
+ "I-FIRSTNAME": 107,
212
+ "I-GENDER": 90,
213
+ "I-HEIGHT": 75,
214
+ "I-IBAN": 102,
215
+ "I-IP": 87,
216
+ "I-IPV4": 101,
217
+ "I-IPV6": 60,
218
+ "I-JOBAREA": 86,
219
+ "I-JOBTITLE": 5,
220
+ "I-JOBTYPE": 66,
221
+ "I-LASTNAME": 44,
222
+ "I-LITECOINADDRESS": 35,
223
+ "I-MAC": 29,
224
+ "I-MASKEDNUMBER": 109,
225
+ "I-MIDDLENAME": 36,
226
+ "I-NEARBYGPSCOORDINATE": 81,
227
+ "I-ORDINALDIRECTION": 19,
228
+ "I-PASSWORD": 83,
229
+ "I-PHONEIMEI": 88,
230
+ "I-PHONENUMBER": 37,
231
+ "I-PIN": 6,
232
+ "I-PREFIX": 70,
233
+ "I-SECONDARYADDRESS": 26,
234
+ "I-SEX": 105,
235
+ "I-SSN": 73,
236
+ "I-STATE": 40,
237
+ "I-STREET": 100,
238
+ "I-TIME": 74,
239
+ "I-URL": 3,
240
+ "I-USERAGENT": 92,
241
+ "I-USERNAME": 50,
242
+ "I-VEHICLEVIN": 84,
243
+ "I-VEHICLEVRM": 61,
244
+ "I-ZIPCODE": 4,
245
+ "O": 21
246
+ },
247
+ "layer_norm_eps": 1e-12,
248
+ "max_position_embeddings": 512,
249
+ "model_type": "mobilebert",
250
+ "normalization_type": "no_norm",
251
+ "num_attention_heads": 4,
252
+ "num_feedforward_networks": 4,
253
+ "num_hidden_layers": 24,
254
+ "pad_token_id": 0,
255
+ "torch_dtype": "float32",
256
+ "transformers_version": "4.42.3",
257
+ "trigram_input": true,
258
+ "true_hidden_size": 128,
259
+ "type_vocab_size": 2,
260
+ "use_bottleneck": true,
261
+ "use_bottleneck_attention": false,
262
+ "vocab_size": 30522
263
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f175006fa482bc7371f3a22e7e9c242ebb0ad0e4d2f6e31dad0f75a2a6031418
3
+ size 98697908
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": true,
47
+ "mask_token": "[MASK]",
48
+ "model_max_length": 1000000000000000019884624838656,
49
+ "pad_token": "[PAD]",
50
+ "sep_token": "[SEP]",
51
+ "strip_accents": null,
52
+ "tokenize_chinese_chars": true,
53
+ "tokenizer_class": "MobileBertTokenizer",
54
+ "unk_token": "[UNK]"
55
+ }
vocab.txt ADDED
The diff for this file is too large to render. See raw diff