PuxAI commited on
Commit
a7eec8b
·
verified ·
1 Parent(s): 0c331ce

Upload folder using huggingface_hub

Browse files
gretel-pii-ready-augmented/bert-crf-bio-bert-base-multilingual-cased/config.json ADDED
@@ -0,0 +1,204 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "TransformerCrfForTokenClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "directionality": "bidi",
8
+ "dtype": "float32",
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "id2label": {
13
+ "0": "O",
14
+ "1": "B-ACCOUNT_NUMBER",
15
+ "2": "I-ACCOUNT_NUMBER",
16
+ "3": "B-ADDRESS",
17
+ "4": "I-ADDRESS",
18
+ "5": "B-API_KEY",
19
+ "6": "I-API_KEY",
20
+ "7": "B-BANK_ROUTING_NUMBER",
21
+ "8": "I-BANK_ROUTING_NUMBER",
22
+ "9": "B-BIOMETRIC_IDENTIFIER",
23
+ "10": "I-BIOMETRIC_IDENTIFIER",
24
+ "11": "B-CERTIFICATE_LICENSE_NUMBER",
25
+ "12": "I-CERTIFICATE_LICENSE_NUMBER",
26
+ "13": "B-CITY",
27
+ "14": "I-CITY",
28
+ "15": "B-COMPANY_NAME",
29
+ "16": "I-COMPANY_NAME",
30
+ "17": "B-COORDINATE",
31
+ "18": "I-COORDINATE",
32
+ "19": "B-COUNTRY",
33
+ "20": "I-COUNTRY",
34
+ "21": "B-CREDIT_CARD_NUMBER",
35
+ "22": "I-CREDIT_CARD_NUMBER",
36
+ "23": "B-CUSTOMER_ID",
37
+ "24": "I-CUSTOMER_ID",
38
+ "25": "B-CVV",
39
+ "26": "I-CVV",
40
+ "27": "B-DATE",
41
+ "28": "I-DATE",
42
+ "29": "B-DATE_OF_BIRTH",
43
+ "30": "I-DATE_OF_BIRTH",
44
+ "31": "B-DATE_TIME",
45
+ "32": "I-DATE_TIME",
46
+ "33": "B-DEVICE_IDENTIFIER",
47
+ "34": "I-DEVICE_IDENTIFIER",
48
+ "35": "B-EMAIL",
49
+ "36": "I-EMAIL",
50
+ "37": "B-EMPLOYEE_ID",
51
+ "38": "I-EMPLOYEE_ID",
52
+ "39": "B-FIRST_NAME",
53
+ "40": "I-FIRST_NAME",
54
+ "41": "B-HEALTH_PLAN_BENEFICIARY_NUMBER",
55
+ "42": "I-HEALTH_PLAN_BENEFICIARY_NUMBER",
56
+ "43": "B-IPV4",
57
+ "44": "I-IPV4",
58
+ "45": "B-IPV6",
59
+ "46": "I-IPV6",
60
+ "47": "B-LAST_NAME",
61
+ "48": "I-LAST_NAME",
62
+ "49": "B-LICENSE_PLATE",
63
+ "50": "I-LICENSE_PLATE",
64
+ "51": "B-MEDICAL_RECORD_NUMBER",
65
+ "52": "I-MEDICAL_RECORD_NUMBER",
66
+ "53": "B-NAME",
67
+ "54": "I-NAME",
68
+ "55": "B-NATIONAL_ID",
69
+ "56": "I-NATIONAL_ID",
70
+ "57": "B-PASSWORD",
71
+ "58": "I-PASSWORD",
72
+ "59": "B-PHONE_NUMBER",
73
+ "60": "I-PHONE_NUMBER",
74
+ "61": "B-PIN",
75
+ "62": "I-PIN",
76
+ "63": "B-POSTCODE",
77
+ "64": "I-POSTCODE",
78
+ "65": "B-SSN",
79
+ "66": "I-SSN",
80
+ "67": "B-STATE",
81
+ "68": "I-STATE",
82
+ "69": "B-STREET_ADDRESS",
83
+ "70": "I-STREET_ADDRESS",
84
+ "71": "B-SWIFT_BIC",
85
+ "72": "I-SWIFT_BIC",
86
+ "73": "B-TAX_ID",
87
+ "74": "I-TAX_ID",
88
+ "75": "B-TIME",
89
+ "76": "I-TIME",
90
+ "77": "B-UNIQUE_IDENTIFIER",
91
+ "78": "I-UNIQUE_IDENTIFIER",
92
+ "79": "B-URL",
93
+ "80": "I-URL",
94
+ "81": "B-USER_NAME",
95
+ "82": "I-USER_NAME",
96
+ "83": "B-VEHICLE_IDENTIFIER",
97
+ "84": "I-VEHICLE_IDENTIFIER"
98
+ },
99
+ "initializer_range": 0.02,
100
+ "intermediate_size": 3072,
101
+ "label2id": {
102
+ "B-ACCOUNT_NUMBER": 1,
103
+ "B-ADDRESS": 3,
104
+ "B-API_KEY": 5,
105
+ "B-BANK_ROUTING_NUMBER": 7,
106
+ "B-BIOMETRIC_IDENTIFIER": 9,
107
+ "B-CERTIFICATE_LICENSE_NUMBER": 11,
108
+ "B-CITY": 13,
109
+ "B-COMPANY_NAME": 15,
110
+ "B-COORDINATE": 17,
111
+ "B-COUNTRY": 19,
112
+ "B-CREDIT_CARD_NUMBER": 21,
113
+ "B-CUSTOMER_ID": 23,
114
+ "B-CVV": 25,
115
+ "B-DATE": 27,
116
+ "B-DATE_OF_BIRTH": 29,
117
+ "B-DATE_TIME": 31,
118
+ "B-DEVICE_IDENTIFIER": 33,
119
+ "B-EMAIL": 35,
120
+ "B-EMPLOYEE_ID": 37,
121
+ "B-FIRST_NAME": 39,
122
+ "B-HEALTH_PLAN_BENEFICIARY_NUMBER": 41,
123
+ "B-IPV4": 43,
124
+ "B-IPV6": 45,
125
+ "B-LAST_NAME": 47,
126
+ "B-LICENSE_PLATE": 49,
127
+ "B-MEDICAL_RECORD_NUMBER": 51,
128
+ "B-NAME": 53,
129
+ "B-NATIONAL_ID": 55,
130
+ "B-PASSWORD": 57,
131
+ "B-PHONE_NUMBER": 59,
132
+ "B-PIN": 61,
133
+ "B-POSTCODE": 63,
134
+ "B-SSN": 65,
135
+ "B-STATE": 67,
136
+ "B-STREET_ADDRESS": 69,
137
+ "B-SWIFT_BIC": 71,
138
+ "B-TAX_ID": 73,
139
+ "B-TIME": 75,
140
+ "B-UNIQUE_IDENTIFIER": 77,
141
+ "B-URL": 79,
142
+ "B-USER_NAME": 81,
143
+ "B-VEHICLE_IDENTIFIER": 83,
144
+ "I-ACCOUNT_NUMBER": 2,
145
+ "I-ADDRESS": 4,
146
+ "I-API_KEY": 6,
147
+ "I-BANK_ROUTING_NUMBER": 8,
148
+ "I-BIOMETRIC_IDENTIFIER": 10,
149
+ "I-CERTIFICATE_LICENSE_NUMBER": 12,
150
+ "I-CITY": 14,
151
+ "I-COMPANY_NAME": 16,
152
+ "I-COORDINATE": 18,
153
+ "I-COUNTRY": 20,
154
+ "I-CREDIT_CARD_NUMBER": 22,
155
+ "I-CUSTOMER_ID": 24,
156
+ "I-CVV": 26,
157
+ "I-DATE": 28,
158
+ "I-DATE_OF_BIRTH": 30,
159
+ "I-DATE_TIME": 32,
160
+ "I-DEVICE_IDENTIFIER": 34,
161
+ "I-EMAIL": 36,
162
+ "I-EMPLOYEE_ID": 38,
163
+ "I-FIRST_NAME": 40,
164
+ "I-HEALTH_PLAN_BENEFICIARY_NUMBER": 42,
165
+ "I-IPV4": 44,
166
+ "I-IPV6": 46,
167
+ "I-LAST_NAME": 48,
168
+ "I-LICENSE_PLATE": 50,
169
+ "I-MEDICAL_RECORD_NUMBER": 52,
170
+ "I-NAME": 54,
171
+ "I-NATIONAL_ID": 56,
172
+ "I-PASSWORD": 58,
173
+ "I-PHONE_NUMBER": 60,
174
+ "I-PIN": 62,
175
+ "I-POSTCODE": 64,
176
+ "I-SSN": 66,
177
+ "I-STATE": 68,
178
+ "I-STREET_ADDRESS": 70,
179
+ "I-SWIFT_BIC": 72,
180
+ "I-TAX_ID": 74,
181
+ "I-TIME": 76,
182
+ "I-UNIQUE_IDENTIFIER": 78,
183
+ "I-URL": 80,
184
+ "I-USER_NAME": 82,
185
+ "I-VEHICLE_IDENTIFIER": 84,
186
+ "O": 0
187
+ },
188
+ "layer_norm_eps": 1e-12,
189
+ "max_position_embeddings": 512,
190
+ "model_type": "bert",
191
+ "num_attention_heads": 12,
192
+ "num_hidden_layers": 12,
193
+ "pad_token_id": 0,
194
+ "pooler_fc_size": 768,
195
+ "pooler_num_attention_heads": 12,
196
+ "pooler_num_fc_layers": 3,
197
+ "pooler_size_per_head": 128,
198
+ "pooler_type": "first_token_transform",
199
+ "position_embedding_type": "absolute",
200
+ "transformers_version": "4.57.6",
201
+ "type_vocab_size": 2,
202
+ "use_cache": true,
203
+ "vocab_size": 119547
204
+ }
gretel-pii-ready-augmented/bert-crf-bio-bert-base-multilingual-cased/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:50cc4c5c0cf33d46a6d05a119078322aa6d4c5a4d77917686f75cc786d1267d3
3
+ size 711729976
gretel-pii-ready-augmented/bert-crf-bio-bert-base-multilingual-cased/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
+ }
gretel-pii-ready-augmented/bert-crf-bio-bert-base-multilingual-cased/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
gretel-pii-ready-augmented/bert-crf-bio-bert-base-multilingual-cased/tokenizer_config.json ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "[PAD]",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "100": {
13
+ "content": "[UNK]",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "101": {
21
+ "content": "[CLS]",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "102": {
29
+ "content": "[SEP]",
30
+ "lstrip": false,
31
+ "normalized": false,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "103": {
37
+ "content": "[MASK]",
38
+ "lstrip": false,
39
+ "normalized": false,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ }
44
+ },
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "[CLS]",
47
+ "do_lower_case": false,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 512,
51
+ "pad_token": "[PAD]",
52
+ "sep_token": "[SEP]",
53
+ "strip_accents": null,
54
+ "tokenize_chinese_chars": true,
55
+ "tokenizer_class": "BertTokenizer",
56
+ "unk_token": "[UNK]"
57
+ }
gretel-pii-ready-augmented/bert-crf-bio-bert-base-multilingual-cased/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b198f517aa070ce7fbd3e97e2c6f9447f3a92efefd239a219ae6133097ce2bbe
3
+ size 5905
gretel-pii-ready-augmented/bert-crf-bio-bert-base-multilingual-cased/training_meta.json ADDED
@@ -0,0 +1,96 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architecture": "token_based",
3
+ "base_model": "google-bert/bert-base-multilingual-cased",
4
+ "tagging_scheme": "bio",
5
+ "use_crf": true,
6
+ "dataset": "gretel-pii-ready-augmented",
7
+ "label_list": [
8
+ "O",
9
+ "B-ACCOUNT_NUMBER",
10
+ "I-ACCOUNT_NUMBER",
11
+ "B-ADDRESS",
12
+ "I-ADDRESS",
13
+ "B-API_KEY",
14
+ "I-API_KEY",
15
+ "B-BANK_ROUTING_NUMBER",
16
+ "I-BANK_ROUTING_NUMBER",
17
+ "B-BIOMETRIC_IDENTIFIER",
18
+ "I-BIOMETRIC_IDENTIFIER",
19
+ "B-CERTIFICATE_LICENSE_NUMBER",
20
+ "I-CERTIFICATE_LICENSE_NUMBER",
21
+ "B-CITY",
22
+ "I-CITY",
23
+ "B-COMPANY_NAME",
24
+ "I-COMPANY_NAME",
25
+ "B-COORDINATE",
26
+ "I-COORDINATE",
27
+ "B-COUNTRY",
28
+ "I-COUNTRY",
29
+ "B-CREDIT_CARD_NUMBER",
30
+ "I-CREDIT_CARD_NUMBER",
31
+ "B-CUSTOMER_ID",
32
+ "I-CUSTOMER_ID",
33
+ "B-CVV",
34
+ "I-CVV",
35
+ "B-DATE",
36
+ "I-DATE",
37
+ "B-DATE_OF_BIRTH",
38
+ "I-DATE_OF_BIRTH",
39
+ "B-DATE_TIME",
40
+ "I-DATE_TIME",
41
+ "B-DEVICE_IDENTIFIER",
42
+ "I-DEVICE_IDENTIFIER",
43
+ "B-EMAIL",
44
+ "I-EMAIL",
45
+ "B-EMPLOYEE_ID",
46
+ "I-EMPLOYEE_ID",
47
+ "B-FIRST_NAME",
48
+ "I-FIRST_NAME",
49
+ "B-HEALTH_PLAN_BENEFICIARY_NUMBER",
50
+ "I-HEALTH_PLAN_BENEFICIARY_NUMBER",
51
+ "B-IPV4",
52
+ "I-IPV4",
53
+ "B-IPV6",
54
+ "I-IPV6",
55
+ "B-LAST_NAME",
56
+ "I-LAST_NAME",
57
+ "B-LICENSE_PLATE",
58
+ "I-LICENSE_PLATE",
59
+ "B-MEDICAL_RECORD_NUMBER",
60
+ "I-MEDICAL_RECORD_NUMBER",
61
+ "B-NAME",
62
+ "I-NAME",
63
+ "B-NATIONAL_ID",
64
+ "I-NATIONAL_ID",
65
+ "B-PASSWORD",
66
+ "I-PASSWORD",
67
+ "B-PHONE_NUMBER",
68
+ "I-PHONE_NUMBER",
69
+ "B-PIN",
70
+ "I-PIN",
71
+ "B-POSTCODE",
72
+ "I-POSTCODE",
73
+ "B-SSN",
74
+ "I-SSN",
75
+ "B-STATE",
76
+ "I-STATE",
77
+ "B-STREET_ADDRESS",
78
+ "I-STREET_ADDRESS",
79
+ "B-SWIFT_BIC",
80
+ "I-SWIFT_BIC",
81
+ "B-TAX_ID",
82
+ "I-TAX_ID",
83
+ "B-TIME",
84
+ "I-TIME",
85
+ "B-UNIQUE_IDENTIFIER",
86
+ "I-UNIQUE_IDENTIFIER",
87
+ "B-URL",
88
+ "I-URL",
89
+ "B-USER_NAME",
90
+ "I-USER_NAME",
91
+ "B-VEHICLE_IDENTIFIER",
92
+ "I-VEHICLE_IDENTIFIER"
93
+ ],
94
+ "max_seq_length": 512,
95
+ "timestamp": "2026-04-07T14:42:25.458084"
96
+ }
gretel-pii-ready-augmented/bert-crf-bio-bert-base-multilingual-cased/vocab.txt ADDED
The diff for this file is too large to render. See raw diff