Hugging Face's logo

PuxAI
/
GenPII-framework-Data

Safetensors
Model card Files
xet
 Community
PuxAI commited on
Commit
bd09e62
·
verified ·
1 Parent(s): 586ce5a

Upload folder using huggingface_hub

Browse files
Files changed (13) hide show
  1. open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/config.json +120 -0
  2. open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/model.safetensors +3 -0
  3. open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/optimizer.pt +3 -0
  4. open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/rng_state.pth +3 -0
  5. open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/scheduler.pt +3 -0
  6. open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/tokenizer.json +0 -0
  7. open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/tokenizer_config.json +15 -0
  8. open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/trainer_state.json +33 -0
  9. open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/training_args.bin +3 -0
  10. open-pii-masking-500k-ai4privacy/TokenBased-CRF/config.json +120 -0
  11. open-pii-masking-500k-ai4privacy/TokenBased-CRF/pytorch_model.bin +3 -0
  12. open-pii-masking-500k-ai4privacy/TokenBased-CRF/tokenizer.json +0 -0
  13. open-pii-masking-500k-ai4privacy/TokenBased-CRF/tokenizer_config.json +15 -0
open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/config.json ADDED
@@ -0,0 +1,120 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_cross_attention": false,
3
+ "architectures": [
4
+ "TransformerCrfForTokenClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": null,
8
+ "classifier_dropout": null,
9
+ "directionality": "bidi",
10
+ "dtype": "float32",
11
+ "eos_token_id": null,
12
+ "hidden_act": "gelu",
13
+ "hidden_dropout_prob": 0.1,
14
+ "hidden_size": 768,
15
+ "id2label": {
16
+ "0": "O",
17
+ "1": "B-AGE",
18
+ "2": "I-AGE",
19
+ "3": "B-BUILDINGNUM",
20
+ "4": "I-BUILDINGNUM",
21
+ "5": "B-CITY",
22
+ "6": "I-CITY",
23
+ "7": "B-CREDITCARDNUMBER",
24
+ "8": "I-CREDITCARDNUMBER",
25
+ "9": "B-DATE",
26
+ "10": "I-DATE",
27
+ "11": "B-DRIVERLICENSENUM",
28
+ "12": "I-DRIVERLICENSENUM",
29
+ "13": "B-EMAIL",
30
+ "14": "I-EMAIL",
31
+ "15": "B-GENDER",
32
+ "16": "I-GENDER",
33
+ "17": "B-GIVENNAME",
34
+ "18": "I-GIVENNAME",
35
+ "19": "B-IDCARDNUM",
36
+ "20": "I-IDCARDNUM",
37
+ "21": "B-PASSPORTNUM",
38
+ "22": "I-PASSPORTNUM",
39
+ "23": "B-SEX",
40
+ "24": "I-SEX",
41
+ "25": "B-SOCIALNUM",
42
+ "26": "I-SOCIALNUM",
43
+ "27": "B-STREET",
44
+ "28": "I-STREET",
45
+ "29": "B-SURNAME",
46
+ "30": "I-SURNAME",
47
+ "31": "B-TAXNUM",
48
+ "32": "I-TAXNUM",
49
+ "33": "B-TELEPHONENUM",
50
+ "34": "I-TELEPHONENUM",
51
+ "35": "B-TIME",
52
+ "36": "I-TIME",
53
+ "37": "B-TITLE",
54
+ "38": "I-TITLE",
55
+ "39": "B-ZIPCODE",
56
+ "40": "I-ZIPCODE"
57
+ },
58
+ "initializer_range": 0.02,
59
+ "intermediate_size": 3072,
60
+ "is_decoder": false,
61
+ "label2id": {
62
+ "B-AGE": 1,
63
+ "B-BUILDINGNUM": 3,
64
+ "B-CITY": 5,
65
+ "B-CREDITCARDNUMBER": 7,
66
+ "B-DATE": 9,
67
+ "B-DRIVERLICENSENUM": 11,
68
+ "B-EMAIL": 13,
69
+ "B-GENDER": 15,
70
+ "B-GIVENNAME": 17,
71
+ "B-IDCARDNUM": 19,
72
+ "B-PASSPORTNUM": 21,
73
+ "B-SEX": 23,
74
+ "B-SOCIALNUM": 25,
75
+ "B-STREET": 27,
76
+ "B-SURNAME": 29,
77
+ "B-TAXNUM": 31,
78
+ "B-TELEPHONENUM": 33,
79
+ "B-TIME": 35,
80
+ "B-TITLE": 37,
81
+ "B-ZIPCODE": 39,
82
+ "I-AGE": 2,
83
+ "I-BUILDINGNUM": 4,
84
+ "I-CITY": 6,
85
+ "I-CREDITCARDNUMBER": 8,
86
+ "I-DATE": 10,
87
+ "I-DRIVERLICENSENUM": 12,
88
+ "I-EMAIL": 14,
89
+ "I-GENDER": 16,
90
+ "I-GIVENNAME": 18,
91
+ "I-IDCARDNUM": 20,
92
+ "I-PASSPORTNUM": 22,
93
+ "I-SEX": 24,
94
+ "I-SOCIALNUM": 26,
95
+ "I-STREET": 28,
96
+ "I-SURNAME": 30,
97
+ "I-TAXNUM": 32,
98
+ "I-TELEPHONENUM": 34,
99
+ "I-TIME": 36,
100
+ "I-TITLE": 38,
101
+ "I-ZIPCODE": 40,
102
+ "O": 0
103
+ },
104
+ "layer_norm_eps": 1e-12,
105
+ "max_position_embeddings": 512,
106
+ "model_type": "bert",
107
+ "num_attention_heads": 12,
108
+ "num_hidden_layers": 12,
109
+ "pad_token_id": 0,
110
+ "pooler_fc_size": 768,
111
+ "pooler_num_attention_heads": 12,
112
+ "pooler_num_fc_layers": 3,
113
+ "pooler_size_per_head": 128,
114
+ "pooler_type": "first_token_transform",
115
+ "tie_word_embeddings": true,
116
+ "transformers_version": "5.3.0",
117
+ "type_vocab_size": 2,
118
+ "use_cache": false,
119
+ "vocab_size": 119547
120
+ }
open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:988fe44477a3633c01f74398d42a45d67e46e272c89bf183fa404ff371ae3be6
3
+ size 711572104
open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6ce6f82e3963c0b944789c4de89f4bc3affccf18537a788f4d88f11ad232af03
3
+ size 1418541131
open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:77e1bc0cc29528179cf7713b093fffbd0754c58c2875616aa74e1b6cd67bfe7d
3
+ size 14645
open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8baf2b5a46fac85f95c0a15caf2414a92c036d41b6ccf016dc0542ed26e499c2
3
+ size 1465
open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/tokenizer_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "backend": "tokenizers",
4
+ "cls_token": "[CLS]",
5
+ "do_lower_case": false,
6
+ "is_local": false,
7
+ "mask_token": "[MASK]",
8
+ "model_max_length": 512,
9
+ "pad_token": "[PAD]",
10
+ "sep_token": "[SEP]",
11
+ "strip_accents": null,
12
+ "tokenize_chinese_chars": true,
13
+ "tokenizer_class": "BertTokenizer",
14
+ "unk_token": "[UNK]"
15
+ }
open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/trainer_state.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.0,
6
+ "eval_steps": 500,
7
+ "global_step": 63,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [],
12
+ "logging_steps": 500,
13
+ "max_steps": 63,
14
+ "num_input_tokens_seen": 0,
15
+ "num_train_epochs": 1,
16
+ "save_steps": 500,
17
+ "stateful_callbacks": {
18
+ "TrainerControl": {
19
+ "args": {
20
+ "should_epoch_stop": false,
21
+ "should_evaluate": false,
22
+ "should_log": false,
23
+ "should_save": true,
24
+ "should_training_stop": true
25
+ },
26
+ "attributes": {}
27
+ }
28
+ },
29
+ "total_flos": 131604301824000.0,
30
+ "train_batch_size": 8,
31
+ "trial_name": null,
32
+ "trial_params": null
33
+ }
open-pii-masking-500k-ai4privacy/TokenBased-CRF/checkpoint-63/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4e5cf459979c7c595ed5f7c36736caa950a566433ec0f1908e8fc19c60ff5a21
3
+ size 5201
open-pii-masking-500k-ai4privacy/TokenBased-CRF/config.json ADDED
@@ -0,0 +1,120 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_cross_attention": false,
3
+ "architectures": [
4
+ "TransformerCrfForTokenClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": null,
8
+ "classifier_dropout": null,
9
+ "directionality": "bidi",
10
+ "dtype": "float32",
11
+ "eos_token_id": null,
12
+ "hidden_act": "gelu",
13
+ "hidden_dropout_prob": 0.1,
14
+ "hidden_size": 768,
15
+ "id2label": {
16
+ "0": "O",
17
+ "1": "B-AGE",
18
+ "2": "I-AGE",
19
+ "3": "B-BUILDINGNUM",
20
+ "4": "I-BUILDINGNUM",
21
+ "5": "B-CITY",
22
+ "6": "I-CITY",
23
+ "7": "B-CREDITCARDNUMBER",
24
+ "8": "I-CREDITCARDNUMBER",
25
+ "9": "B-DATE",
26
+ "10": "I-DATE",
27
+ "11": "B-DRIVERLICENSENUM",
28
+ "12": "I-DRIVERLICENSENUM",
29
+ "13": "B-EMAIL",
30
+ "14": "I-EMAIL",
31
+ "15": "B-GENDER",
32
+ "16": "I-GENDER",
33
+ "17": "B-GIVENNAME",
34
+ "18": "I-GIVENNAME",
35
+ "19": "B-IDCARDNUM",
36
+ "20": "I-IDCARDNUM",
37
+ "21": "B-PASSPORTNUM",
38
+ "22": "I-PASSPORTNUM",
39
+ "23": "B-SEX",
40
+ "24": "I-SEX",
41
+ "25": "B-SOCIALNUM",
42
+ "26": "I-SOCIALNUM",
43
+ "27": "B-STREET",
44
+ "28": "I-STREET",
45
+ "29": "B-SURNAME",
46
+ "30": "I-SURNAME",
47
+ "31": "B-TAXNUM",
48
+ "32": "I-TAXNUM",
49
+ "33": "B-TELEPHONENUM",
50
+ "34": "I-TELEPHONENUM",
51
+ "35": "B-TIME",
52
+ "36": "I-TIME",
53
+ "37": "B-TITLE",
54
+ "38": "I-TITLE",
55
+ "39": "B-ZIPCODE",
56
+ "40": "I-ZIPCODE"
57
+ },
58
+ "initializer_range": 0.02,
59
+ "intermediate_size": 3072,
60
+ "is_decoder": false,
61
+ "label2id": {
62
+ "B-AGE": 1,
63
+ "B-BUILDINGNUM": 3,
64
+ "B-CITY": 5,
65
+ "B-CREDITCARDNUMBER": 7,
66
+ "B-DATE": 9,
67
+ "B-DRIVERLICENSENUM": 11,
68
+ "B-EMAIL": 13,
69
+ "B-GENDER": 15,
70
+ "B-GIVENNAME": 17,
71
+ "B-IDCARDNUM": 19,
72
+ "B-PASSPORTNUM": 21,
73
+ "B-SEX": 23,
74
+ "B-SOCIALNUM": 25,
75
+ "B-STREET": 27,
76
+ "B-SURNAME": 29,
77
+ "B-TAXNUM": 31,
78
+ "B-TELEPHONENUM": 33,
79
+ "B-TIME": 35,
80
+ "B-TITLE": 37,
81
+ "B-ZIPCODE": 39,
82
+ "I-AGE": 2,
83
+ "I-BUILDINGNUM": 4,
84
+ "I-CITY": 6,
85
+ "I-CREDITCARDNUMBER": 8,
86
+ "I-DATE": 10,
87
+ "I-DRIVERLICENSENUM": 12,
88
+ "I-EMAIL": 14,
89
+ "I-GENDER": 16,
90
+ "I-GIVENNAME": 18,
91
+ "I-IDCARDNUM": 20,
92
+ "I-PASSPORTNUM": 22,
93
+ "I-SEX": 24,
94
+ "I-SOCIALNUM": 26,
95
+ "I-STREET": 28,
96
+ "I-SURNAME": 30,
97
+ "I-TAXNUM": 32,
98
+ "I-TELEPHONENUM": 34,
99
+ "I-TIME": 36,
100
+ "I-TITLE": 38,
101
+ "I-ZIPCODE": 40,
102
+ "O": 0
103
+ },
104
+ "layer_norm_eps": 1e-12,
105
+ "max_position_embeddings": 512,
106
+ "model_type": "bert",
107
+ "num_attention_heads": 12,
108
+ "num_hidden_layers": 12,
109
+ "pad_token_id": 0,
110
+ "pooler_fc_size": 768,
111
+ "pooler_num_attention_heads": 12,
112
+ "pooler_num_fc_layers": 3,
113
+ "pooler_size_per_head": 128,
114
+ "pooler_type": "first_token_transform",
115
+ "tie_word_embeddings": true,
116
+ "transformers_version": "5.3.0",
117
+ "type_vocab_size": 2,
118
+ "use_cache": false,
119
+ "vocab_size": 119547
120
+ }
open-pii-masking-500k-ai4privacy/TokenBased-CRF/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a2291ae713e208e7f762adbcbf6f6927ec7eb8be2dbeef176a6b54423d8c255
3
+ size 711634067
open-pii-masking-500k-ai4privacy/TokenBased-CRF/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
open-pii-masking-500k-ai4privacy/TokenBased-CRF/tokenizer_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "backend": "tokenizers",
4
+ "cls_token": "[CLS]",
5
+ "do_lower_case": false,
6
+ "is_local": false,
7
+ "mask_token": "[MASK]",
8
+ "model_max_length": 512,
9
+ "pad_token": "[PAD]",
10
+ "sep_token": "[SEP]",
11
+ "strip_accents": null,
12
+ "tokenize_chinese_chars": true,
13
+ "tokenizer_class": "BertTokenizer",
14
+ "unk_token": "[UNK]"
15
+ }