RedCaesar commited on
Commit
304c489
·
verified ·
1 Parent(s): 47117a5

Upload 11 files

Browse files
config.json ADDED
@@ -0,0 +1,257 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "dslim/bert-base-NER",
3
+ "_num_labels": 9,
4
+ "architectures": [
5
+ "BertForTokenClassification"
6
+ ],
7
+ "attention_probs_dropout_prob": 0.1,
8
+ "classifier_dropout": null,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "id2label": {
13
+ "0": "B-ACCOUNTNAME",
14
+ "1": "B-ACCOUNTNUMBER",
15
+ "2": "B-AGE",
16
+ "3": "B-AMOUNT",
17
+ "4": "B-BIC",
18
+ "5": "B-BITCOINADDRESS",
19
+ "6": "B-BUILDINGNUMBER",
20
+ "7": "B-CITY",
21
+ "8": "B-COMPANYNAME",
22
+ "9": "B-COUNTY",
23
+ "10": "B-CREDITCARDCVV",
24
+ "11": "B-CREDITCARDISSUER",
25
+ "12": "B-CREDITCARDNUMBER",
26
+ "13": "B-CURRENCY",
27
+ "14": "B-CURRENCYCODE",
28
+ "15": "B-CURRENCYNAME",
29
+ "16": "B-CURRENCYSYMBOL",
30
+ "17": "B-DATE",
31
+ "18": "B-DOB",
32
+ "19": "B-EMAIL",
33
+ "20": "B-ETHEREUMADDRESS",
34
+ "21": "B-EYECOLOR",
35
+ "22": "B-FIRSTNAME",
36
+ "23": "B-GENDER",
37
+ "24": "B-HEIGHT",
38
+ "25": "B-IBAN",
39
+ "26": "B-IP",
40
+ "27": "B-IPV4",
41
+ "28": "B-IPV6",
42
+ "29": "B-JOBAREA",
43
+ "30": "B-JOBTITLE",
44
+ "31": "B-JOBTYPE",
45
+ "32": "B-LASTNAME",
46
+ "33": "B-LITECOINADDRESS",
47
+ "34": "B-MAC",
48
+ "35": "B-MASKEDNUMBER",
49
+ "36": "B-MIDDLENAME",
50
+ "37": "B-NEARBYGPSCOORDINATE",
51
+ "38": "B-ORDINALDIRECTION",
52
+ "39": "B-PASSWORD",
53
+ "40": "B-PHONEIMEI",
54
+ "41": "B-PHONENUMBER",
55
+ "42": "B-PIN",
56
+ "43": "B-PREFIX",
57
+ "44": "B-SECONDARYADDRESS",
58
+ "45": "B-SEX",
59
+ "46": "B-SSN",
60
+ "47": "B-STATE",
61
+ "48": "B-STREET",
62
+ "49": "B-TIME",
63
+ "50": "B-URL",
64
+ "51": "B-USERAGENT",
65
+ "52": "B-USERNAME",
66
+ "53": "B-VEHICLEVIN",
67
+ "54": "B-VEHICLEVRM",
68
+ "55": "B-ZIPCODE",
69
+ "56": "I-ACCOUNTNAME",
70
+ "57": "I-ACCOUNTNUMBER",
71
+ "58": "I-AGE",
72
+ "59": "I-AMOUNT",
73
+ "60": "I-BIC",
74
+ "61": "I-BITCOINADDRESS",
75
+ "62": "I-BUILDINGNUMBER",
76
+ "63": "I-CITY",
77
+ "64": "I-COMPANYNAME",
78
+ "65": "I-COUNTY",
79
+ "66": "I-CREDITCARDCVV",
80
+ "67": "I-CREDITCARDISSUER",
81
+ "68": "I-CREDITCARDNUMBER",
82
+ "69": "I-CURRENCY",
83
+ "70": "I-CURRENCYCODE",
84
+ "71": "I-CURRENCYNAME",
85
+ "72": "I-CURRENCYSYMBOL",
86
+ "73": "I-DATE",
87
+ "74": "I-DOB",
88
+ "75": "I-EMAIL",
89
+ "76": "I-ETHEREUMADDRESS",
90
+ "77": "I-EYECOLOR",
91
+ "78": "I-FIRSTNAME",
92
+ "79": "I-GENDER",
93
+ "80": "I-HEIGHT",
94
+ "81": "I-IBAN",
95
+ "82": "I-IP",
96
+ "83": "I-IPV4",
97
+ "84": "I-IPV6",
98
+ "85": "I-JOBAREA",
99
+ "86": "I-JOBTITLE",
100
+ "87": "I-JOBTYPE",
101
+ "88": "I-LASTNAME",
102
+ "89": "I-LITECOINADDRESS",
103
+ "90": "I-MAC",
104
+ "91": "I-MASKEDNUMBER",
105
+ "92": "I-MIDDLENAME",
106
+ "93": "I-NEARBYGPSCOORDINATE",
107
+ "94": "I-ORDINALDIRECTION",
108
+ "95": "I-PASSWORD",
109
+ "96": "I-PHONEIMEI",
110
+ "97": "I-PHONENUMBER",
111
+ "98": "I-PIN",
112
+ "99": "I-PREFIX",
113
+ "100": "I-SECONDARYADDRESS",
114
+ "101": "I-SEX",
115
+ "102": "I-SSN",
116
+ "103": "I-STATE",
117
+ "104": "I-STREET",
118
+ "105": "I-TIME",
119
+ "106": "I-URL",
120
+ "107": "I-USERAGENT",
121
+ "108": "I-USERNAME",
122
+ "109": "I-VEHICLEVIN",
123
+ "110": "I-VEHICLEVRM",
124
+ "111": "I-ZIPCODE",
125
+ "112": "O"
126
+ },
127
+ "initializer_range": 0.02,
128
+ "intermediate_size": 3072,
129
+ "label2id": {
130
+ "B-ACCOUNTNAME": 0,
131
+ "B-ACCOUNTNUMBER": 1,
132
+ "B-AGE": 2,
133
+ "B-AMOUNT": 3,
134
+ "B-BIC": 4,
135
+ "B-BITCOINADDRESS": 5,
136
+ "B-BUILDINGNUMBER": 6,
137
+ "B-CITY": 7,
138
+ "B-COMPANYNAME": 8,
139
+ "B-COUNTY": 9,
140
+ "B-CREDITCARDCVV": 10,
141
+ "B-CREDITCARDISSUER": 11,
142
+ "B-CREDITCARDNUMBER": 12,
143
+ "B-CURRENCY": 13,
144
+ "B-CURRENCYCODE": 14,
145
+ "B-CURRENCYNAME": 15,
146
+ "B-CURRENCYSYMBOL": 16,
147
+ "B-DATE": 17,
148
+ "B-DOB": 18,
149
+ "B-EMAIL": 19,
150
+ "B-ETHEREUMADDRESS": 20,
151
+ "B-EYECOLOR": 21,
152
+ "B-FIRSTNAME": 22,
153
+ "B-GENDER": 23,
154
+ "B-HEIGHT": 24,
155
+ "B-IBAN": 25,
156
+ "B-IP": 26,
157
+ "B-IPV4": 27,
158
+ "B-IPV6": 28,
159
+ "B-JOBAREA": 29,
160
+ "B-JOBTITLE": 30,
161
+ "B-JOBTYPE": 31,
162
+ "B-LASTNAME": 32,
163
+ "B-LITECOINADDRESS": 33,
164
+ "B-MAC": 34,
165
+ "B-MASKEDNUMBER": 35,
166
+ "B-MIDDLENAME": 36,
167
+ "B-NEARBYGPSCOORDINATE": 37,
168
+ "B-ORDINALDIRECTION": 38,
169
+ "B-PASSWORD": 39,
170
+ "B-PHONEIMEI": 40,
171
+ "B-PHONENUMBER": 41,
172
+ "B-PIN": 42,
173
+ "B-PREFIX": 43,
174
+ "B-SECONDARYADDRESS": 44,
175
+ "B-SEX": 45,
176
+ "B-SSN": 46,
177
+ "B-STATE": 47,
178
+ "B-STREET": 48,
179
+ "B-TIME": 49,
180
+ "B-URL": 50,
181
+ "B-USERAGENT": 51,
182
+ "B-USERNAME": 52,
183
+ "B-VEHICLEVIN": 53,
184
+ "B-VEHICLEVRM": 54,
185
+ "B-ZIPCODE": 55,
186
+ "I-ACCOUNTNAME": 56,
187
+ "I-ACCOUNTNUMBER": 57,
188
+ "I-AGE": 58,
189
+ "I-AMOUNT": 59,
190
+ "I-BIC": 60,
191
+ "I-BITCOINADDRESS": 61,
192
+ "I-BUILDINGNUMBER": 62,
193
+ "I-CITY": 63,
194
+ "I-COMPANYNAME": 64,
195
+ "I-COUNTY": 65,
196
+ "I-CREDITCARDCVV": 66,
197
+ "I-CREDITCARDISSUER": 67,
198
+ "I-CREDITCARDNUMBER": 68,
199
+ "I-CURRENCY": 69,
200
+ "I-CURRENCYCODE": 70,
201
+ "I-CURRENCYNAME": 71,
202
+ "I-CURRENCYSYMBOL": 72,
203
+ "I-DATE": 73,
204
+ "I-DOB": 74,
205
+ "I-EMAIL": 75,
206
+ "I-ETHEREUMADDRESS": 76,
207
+ "I-EYECOLOR": 77,
208
+ "I-FIRSTNAME": 78,
209
+ "I-GENDER": 79,
210
+ "I-HEIGHT": 80,
211
+ "I-IBAN": 81,
212
+ "I-IP": 82,
213
+ "I-IPV4": 83,
214
+ "I-IPV6": 84,
215
+ "I-JOBAREA": 85,
216
+ "I-JOBTITLE": 86,
217
+ "I-JOBTYPE": 87,
218
+ "I-LASTNAME": 88,
219
+ "I-LITECOINADDRESS": 89,
220
+ "I-MAC": 90,
221
+ "I-MASKEDNUMBER": 91,
222
+ "I-MIDDLENAME": 92,
223
+ "I-NEARBYGPSCOORDINATE": 93,
224
+ "I-ORDINALDIRECTION": 94,
225
+ "I-PASSWORD": 95,
226
+ "I-PHONEIMEI": 96,
227
+ "I-PHONENUMBER": 97,
228
+ "I-PIN": 98,
229
+ "I-PREFIX": 99,
230
+ "I-SECONDARYADDRESS": 100,
231
+ "I-SEX": 101,
232
+ "I-SSN": 102,
233
+ "I-STATE": 103,
234
+ "I-STREET": 104,
235
+ "I-TIME": 105,
236
+ "I-URL": 106,
237
+ "I-USERAGENT": 107,
238
+ "I-USERNAME": 108,
239
+ "I-VEHICLEVIN": 109,
240
+ "I-VEHICLEVRM": 110,
241
+ "I-ZIPCODE": 111,
242
+ "O": 112
243
+ },
244
+ "layer_norm_eps": 1e-12,
245
+ "max_position_embeddings": 512,
246
+ "model_type": "bert",
247
+ "num_attention_heads": 12,
248
+ "num_hidden_layers": 12,
249
+ "output_past": true,
250
+ "pad_token_id": 0,
251
+ "position_embedding_type": "absolute",
252
+ "torch_dtype": "float32",
253
+ "transformers_version": "4.47.0",
254
+ "type_vocab_size": 2,
255
+ "use_cache": true,
256
+ "vocab_size": 28996
257
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d9e0b33f39622d125b79d681f808c4e1a911903ed9c2cb9675d42e3bcc3ba71
3
+ size 431249644
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc4659ba2ab936b0cfa63c12e83eb5f8020b9eebf3a53d3fde2675deed06ff4f
3
+ size 862619066
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5a2b676a495d6f55553fd30b54d8204f34df7ffb180dce1c673bfce55f1bbe1
3
+ size 14244
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9db35652316cd18818079609bdbd22b09de59e7ca3cd85099fc0bf1dbb6d1001
3
+ size 1064
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 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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_basic_tokenize": true,
47
+ "do_lower_case": false,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "max_len": 512,
51
+ "model_max_length": 512,
52
+ "never_split": null,
53
+ "pad_token": "[PAD]",
54
+ "sep_token": "[SEP]",
55
+ "strip_accents": null,
56
+ "tokenize_chinese_chars": true,
57
+ "tokenizer_class": "BertTokenizer",
58
+ "unk_token": "[UNK]"
59
+ }
trainer_state.json ADDED
@@ -0,0 +1,113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.09755547344684601,
3
+ "best_model_checkpoint": "dslim/bert-base-NER-model/checkpoint-4390",
4
+ "epoch": 2.0,
5
+ "eval_steps": 500,
6
+ "global_step": 4390,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.22779043280182232,
13
+ "grad_norm": 1.933945894241333,
14
+ "learning_rate": 1.7722095671981778e-05,
15
+ "loss": 0.6767,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.45558086560364464,
20
+ "grad_norm": 2.780022144317627,
21
+ "learning_rate": 1.5444191343963555e-05,
22
+ "loss": 0.2153,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.683371298405467,
27
+ "grad_norm": 1.4999059438705444,
28
+ "learning_rate": 1.3166287015945332e-05,
29
+ "loss": 0.1472,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 0.9111617312072893,
34
+ "grad_norm": 0.7693917155265808,
35
+ "learning_rate": 1.0888382687927108e-05,
36
+ "loss": 0.1308,
37
+ "step": 2000
38
+ },
39
+ {
40
+ "epoch": 1.0,
41
+ "eval_accuracy": 0.9589862807506537,
42
+ "eval_f1": 0.8543397844850693,
43
+ "eval_loss": 0.1143401563167572,
44
+ "eval_precision": 0.8349575605155611,
45
+ "eval_recall": 0.8746432491767289,
46
+ "eval_runtime": 105.1757,
47
+ "eval_samples_per_second": 83.451,
48
+ "eval_steps_per_second": 5.22,
49
+ "step": 2195
50
+ },
51
+ {
52
+ "epoch": 1.1389521640091116,
53
+ "grad_norm": 0.9619026184082031,
54
+ "learning_rate": 8.610478359908885e-06,
55
+ "loss": 0.1068,
56
+ "step": 2500
57
+ },
58
+ {
59
+ "epoch": 1.366742596810934,
60
+ "grad_norm": 0.6502276659011841,
61
+ "learning_rate": 6.3325740318906616e-06,
62
+ "loss": 0.1,
63
+ "step": 3000
64
+ },
65
+ {
66
+ "epoch": 1.5945330296127562,
67
+ "grad_norm": 0.8502880930900574,
68
+ "learning_rate": 4.054669703872437e-06,
69
+ "loss": 0.0963,
70
+ "step": 3500
71
+ },
72
+ {
73
+ "epoch": 1.8223234624145785,
74
+ "grad_norm": 1.196458101272583,
75
+ "learning_rate": 1.7767653758542143e-06,
76
+ "loss": 0.0965,
77
+ "step": 4000
78
+ },
79
+ {
80
+ "epoch": 2.0,
81
+ "eval_accuracy": 0.963877173871149,
82
+ "eval_f1": 0.8823518854209665,
83
+ "eval_loss": 0.09755547344684601,
84
+ "eval_precision": 0.8659245429245781,
85
+ "eval_recall": 0.8994145627515551,
86
+ "eval_runtime": 97.9073,
87
+ "eval_samples_per_second": 89.646,
88
+ "eval_steps_per_second": 5.607,
89
+ "step": 4390
90
+ }
91
+ ],
92
+ "logging_steps": 500,
93
+ "max_steps": 4390,
94
+ "num_input_tokens_seen": 0,
95
+ "num_train_epochs": 2,
96
+ "save_steps": 500,
97
+ "stateful_callbacks": {
98
+ "TrainerControl": {
99
+ "args": {
100
+ "should_epoch_stop": false,
101
+ "should_evaluate": false,
102
+ "should_log": false,
103
+ "should_save": true,
104
+ "should_training_stop": true
105
+ },
106
+ "attributes": {}
107
+ }
108
+ },
109
+ "total_flos": 8407997307385200.0,
110
+ "train_batch_size": 16,
111
+ "trial_name": null,
112
+ "trial_params": null
113
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b451458eba455584c48f7bc3192f60282899f92001d6ae4c582ade217d55817
3
+ size 5368
vocab.txt ADDED
The diff for this file is too large to render. See raw diff