xXiaobuding commited on
Commit
a6dd1a0
1 Parent(s): b4cf952

Training in progress, epoch 1

Browse files
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
.gitignore ADDED
@@ -0,0 +1 @@
 
 
1
+ checkpoint-*/
config.json ADDED
@@ -0,0 +1,254 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "FacebookAI/xlm-roberta-base",
3
+ "architectures": [
4
+ "XLMRobertaForTokenClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "id2label": {
14
+ "0": "O",
15
+ "1": "B-PHONEIMEI",
16
+ "2": "I-PHONEIMEI",
17
+ "3": "B-JOBAREA",
18
+ "4": "B-FIRSTNAME",
19
+ "5": "I-FIRSTNAME",
20
+ "6": "B-VEHICLEVIN",
21
+ "7": "I-VEHICLEVIN",
22
+ "8": "B-AGE",
23
+ "9": "B-GENDER",
24
+ "10": "I-GENDER",
25
+ "11": "B-HEIGHT",
26
+ "12": "I-HEIGHT",
27
+ "13": "B-BUILDINGNUMBER",
28
+ "14": "I-BUILDINGNUMBER",
29
+ "15": "B-MASKEDNUMBER",
30
+ "16": "I-MASKEDNUMBER",
31
+ "17": "B-PASSWORD",
32
+ "18": "I-PASSWORD",
33
+ "19": "B-DOB",
34
+ "20": "I-DOB",
35
+ "21": "B-IPV6",
36
+ "22": "I-IPV6",
37
+ "23": "B-NEARBYGPSCOORDINATE",
38
+ "24": "I-NEARBYGPSCOORDINATE",
39
+ "25": "B-USERAGENT",
40
+ "26": "I-USERAGENT",
41
+ "27": "B-TIME",
42
+ "28": "I-TIME",
43
+ "29": "B-JOBTITLE",
44
+ "30": "I-JOBTITLE",
45
+ "31": "B-COUNTY",
46
+ "32": "B-EMAIL",
47
+ "33": "I-EMAIL",
48
+ "34": "B-ACCOUNTNUMBER",
49
+ "35": "I-ACCOUNTNUMBER",
50
+ "36": "B-PIN",
51
+ "37": "I-PIN",
52
+ "38": "B-EYECOLOR",
53
+ "39": "I-EYECOLOR",
54
+ "40": "B-LASTNAME",
55
+ "41": "I-LASTNAME",
56
+ "42": "I-JOBAREA",
57
+ "43": "B-IPV4",
58
+ "44": "I-IPV4",
59
+ "45": "B-DATE",
60
+ "46": "I-DATE",
61
+ "47": "B-STREET",
62
+ "48": "I-STREET",
63
+ "49": "B-CITY",
64
+ "50": "I-CITY",
65
+ "51": "B-PREFIX",
66
+ "52": "I-PREFIX",
67
+ "53": "B-CREDITCARDISSUER",
68
+ "54": "B-CREDITCARDNUMBER",
69
+ "55": "I-CREDITCARDNUMBER",
70
+ "56": "I-CREDITCARDISSUER",
71
+ "57": "B-MIDDLENAME",
72
+ "58": "B-STATE",
73
+ "59": "I-STATE",
74
+ "60": "B-VEHICLEVRM",
75
+ "61": "I-VEHICLEVRM",
76
+ "62": "B-ORDINALDIRECTION",
77
+ "63": "B-SEX",
78
+ "64": "B-JOBTYPE",
79
+ "65": "I-JOBTYPE",
80
+ "66": "B-CURRENCYCODE",
81
+ "67": "I-CURRENCYCODE",
82
+ "68": "B-CURRENCYSYMBOL",
83
+ "69": "I-AMOUNT",
84
+ "70": "B-ACCOUNTNAME",
85
+ "71": "I-ACCOUNTNAME",
86
+ "72": "B-BITCOINADDRESS",
87
+ "73": "I-BITCOINADDRESS",
88
+ "74": "B-LITECOINADDRESS",
89
+ "75": "I-LITECOINADDRESS",
90
+ "76": "B-PHONENUMBER",
91
+ "77": "I-PHONENUMBER",
92
+ "78": "B-MAC",
93
+ "79": "I-MAC",
94
+ "80": "B-CURRENCY",
95
+ "81": "B-IBAN",
96
+ "82": "I-IBAN",
97
+ "83": "B-COMPANYNAME",
98
+ "84": "I-COMPANYNAME",
99
+ "85": "B-CURRENCYNAME",
100
+ "86": "I-CURRENCYNAME",
101
+ "87": "I-CURRENCYSYMBOL",
102
+ "88": "B-ZIPCODE",
103
+ "89": "I-ZIPCODE",
104
+ "90": "B-SSN",
105
+ "91": "I-SSN",
106
+ "92": "B-AMOUNT",
107
+ "93": "I-CURRENCY",
108
+ "94": "B-URL",
109
+ "95": "I-URL",
110
+ "96": "B-IP",
111
+ "97": "I-IP",
112
+ "98": "B-SECONDARYADDRESS",
113
+ "99": "I-SECONDARYADDRESS",
114
+ "100": "B-USERNAME",
115
+ "101": "I-USERNAME",
116
+ "102": "B-ETHEREUMADDRESS",
117
+ "103": "I-ETHEREUMADDRESS",
118
+ "104": "B-CREDITCARDCVV",
119
+ "105": "I-CREDITCARDCVV",
120
+ "106": "I-COUNTY",
121
+ "107": "I-AGE",
122
+ "108": "I-MIDDLENAME",
123
+ "109": "B-BIC",
124
+ "110": "I-BIC"
125
+ },
126
+ "initializer_range": 0.02,
127
+ "intermediate_size": 3072,
128
+ "label2id": {
129
+ "B-ACCOUNTNAME": 70,
130
+ "B-ACCOUNTNUMBER": 34,
131
+ "B-AGE": 8,
132
+ "B-AMOUNT": 92,
133
+ "B-BIC": 109,
134
+ "B-BITCOINADDRESS": 72,
135
+ "B-BUILDINGNUMBER": 13,
136
+ "B-CITY": 49,
137
+ "B-COMPANYNAME": 83,
138
+ "B-COUNTY": 31,
139
+ "B-CREDITCARDCVV": 104,
140
+ "B-CREDITCARDISSUER": 53,
141
+ "B-CREDITCARDNUMBER": 54,
142
+ "B-CURRENCY": 80,
143
+ "B-CURRENCYCODE": 66,
144
+ "B-CURRENCYNAME": 85,
145
+ "B-CURRENCYSYMBOL": 68,
146
+ "B-DATE": 45,
147
+ "B-DOB": 19,
148
+ "B-EMAIL": 32,
149
+ "B-ETHEREUMADDRESS": 102,
150
+ "B-EYECOLOR": 38,
151
+ "B-FIRSTNAME": 4,
152
+ "B-GENDER": 9,
153
+ "B-HEIGHT": 11,
154
+ "B-IBAN": 81,
155
+ "B-IP": 96,
156
+ "B-IPV4": 43,
157
+ "B-IPV6": 21,
158
+ "B-JOBAREA": 3,
159
+ "B-JOBTITLE": 29,
160
+ "B-JOBTYPE": 64,
161
+ "B-LASTNAME": 40,
162
+ "B-LITECOINADDRESS": 74,
163
+ "B-MAC": 78,
164
+ "B-MASKEDNUMBER": 15,
165
+ "B-MIDDLENAME": 57,
166
+ "B-NEARBYGPSCOORDINATE": 23,
167
+ "B-ORDINALDIRECTION": 62,
168
+ "B-PASSWORD": 17,
169
+ "B-PHONEIMEI": 1,
170
+ "B-PHONENUMBER": 76,
171
+ "B-PIN": 36,
172
+ "B-PREFIX": 51,
173
+ "B-SECONDARYADDRESS": 98,
174
+ "B-SEX": 63,
175
+ "B-SSN": 90,
176
+ "B-STATE": 58,
177
+ "B-STREET": 47,
178
+ "B-TIME": 27,
179
+ "B-URL": 94,
180
+ "B-USERAGENT": 25,
181
+ "B-USERNAME": 100,
182
+ "B-VEHICLEVIN": 6,
183
+ "B-VEHICLEVRM": 60,
184
+ "B-ZIPCODE": 88,
185
+ "I-ACCOUNTNAME": 71,
186
+ "I-ACCOUNTNUMBER": 35,
187
+ "I-AGE": 107,
188
+ "I-AMOUNT": 69,
189
+ "I-BIC": 110,
190
+ "I-BITCOINADDRESS": 73,
191
+ "I-BUILDINGNUMBER": 14,
192
+ "I-CITY": 50,
193
+ "I-COMPANYNAME": 84,
194
+ "I-COUNTY": 106,
195
+ "I-CREDITCARDCVV": 105,
196
+ "I-CREDITCARDISSUER": 56,
197
+ "I-CREDITCARDNUMBER": 55,
198
+ "I-CURRENCY": 93,
199
+ "I-CURRENCYCODE": 67,
200
+ "I-CURRENCYNAME": 86,
201
+ "I-CURRENCYSYMBOL": 87,
202
+ "I-DATE": 46,
203
+ "I-DOB": 20,
204
+ "I-EMAIL": 33,
205
+ "I-ETHEREUMADDRESS": 103,
206
+ "I-EYECOLOR": 39,
207
+ "I-FIRSTNAME": 5,
208
+ "I-GENDER": 10,
209
+ "I-HEIGHT": 12,
210
+ "I-IBAN": 82,
211
+ "I-IP": 97,
212
+ "I-IPV4": 44,
213
+ "I-IPV6": 22,
214
+ "I-JOBAREA": 42,
215
+ "I-JOBTITLE": 30,
216
+ "I-JOBTYPE": 65,
217
+ "I-LASTNAME": 41,
218
+ "I-LITECOINADDRESS": 75,
219
+ "I-MAC": 79,
220
+ "I-MASKEDNUMBER": 16,
221
+ "I-MIDDLENAME": 108,
222
+ "I-NEARBYGPSCOORDINATE": 24,
223
+ "I-PASSWORD": 18,
224
+ "I-PHONEIMEI": 2,
225
+ "I-PHONENUMBER": 77,
226
+ "I-PIN": 37,
227
+ "I-PREFIX": 52,
228
+ "I-SECONDARYADDRESS": 99,
229
+ "I-SSN": 91,
230
+ "I-STATE": 59,
231
+ "I-STREET": 48,
232
+ "I-TIME": 28,
233
+ "I-URL": 95,
234
+ "I-USERAGENT": 26,
235
+ "I-USERNAME": 101,
236
+ "I-VEHICLEVIN": 7,
237
+ "I-VEHICLEVRM": 61,
238
+ "I-ZIPCODE": 89,
239
+ "O": 0
240
+ },
241
+ "layer_norm_eps": 1e-05,
242
+ "max_position_embeddings": 514,
243
+ "model_type": "xlm-roberta",
244
+ "num_attention_heads": 12,
245
+ "num_hidden_layers": 12,
246
+ "output_past": true,
247
+ "pad_token_id": 1,
248
+ "position_embedding_type": "absolute",
249
+ "torch_dtype": "float32",
250
+ "transformers_version": "4.26.1",
251
+ "type_vocab_size": 1,
252
+ "use_cache": true,
253
+ "vocab_size": 250002
254
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b0a76ec2026bbcc19dc0b0bcb95019c6fbd09b7d2abc8c5aba2d7e5b537b6050
3
+ size 1110223661
sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
3
+ size 5069051
special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62c24cdc13d4c9952d63718d6c9fa4c287974249e16b7ade6d5a85e7bbb75626
3
+ size 17082660
tokenizer_config.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "__type": "AddedToken",
7
+ "content": "<mask>",
8
+ "lstrip": true,
9
+ "normalized": true,
10
+ "rstrip": false,
11
+ "single_word": false
12
+ },
13
+ "model_max_length": 512,
14
+ "name_or_path": "FacebookAI/xlm-roberta-base",
15
+ "pad_token": "<pad>",
16
+ "sep_token": "</s>",
17
+ "special_tokens_map_file": null,
18
+ "tokenizer_class": "XLMRobertaTokenizer",
19
+ "unk_token": "<unk>"
20
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ced730cc5c5efd3865bd4e24cb15c3305d398a3d2e21f902d9074befae35f0c1
3
+ size 3515