PuxAI commited on
Commit
2f41b62
·
verified ·
1 Parent(s): cfcd9ef

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/config.json +155 -0
  2. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/model.safetensors +3 -0
  3. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/optimizer.pt +3 -0
  4. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/rng_state.pth +3 -0
  5. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/scheduler.pt +3 -0
  6. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer.json +0 -0
  7. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer_config.json +23 -0
  8. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/trainer_state.json +48 -0
  9. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/training_args.bin +3 -0
  10. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/config.json +155 -0
  11. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/model.safetensors +3 -0
  12. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/optimizer.pt +3 -0
  13. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/rng_state.pth +3 -0
  14. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/scheduler.pt +3 -0
  15. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/tokenizer.json +0 -0
  16. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/tokenizer_config.json +23 -0
  17. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/trainer_state.json +174 -0
  18. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/training_args.bin +3 -0
  19. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/config.json +155 -0
  20. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/model.safetensors +3 -0
  21. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/optimizer.pt +3 -0
  22. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/rng_state.pth +3 -0
  23. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/scheduler.pt +3 -0
  24. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/tokenizer.json +0 -0
  25. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/tokenizer_config.json +23 -0
  26. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/trainer_state.json +181 -0
  27. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/training_args.bin +3 -0
  28. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/config.json +155 -0
  29. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/model.safetensors +3 -0
  30. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/optimizer.pt +3 -0
  31. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/rng_state.pth +3 -0
  32. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/scheduler.pt +3 -0
  33. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/tokenizer.json +0 -0
  34. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/tokenizer_config.json +23 -0
  35. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/trainer_state.json +188 -0
  36. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/training_args.bin +3 -0
  37. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/config.json +155 -0
  38. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/model.safetensors +3 -0
  39. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/optimizer.pt +3 -0
  40. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/rng_state.pth +3 -0
  41. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/scheduler.pt +3 -0
  42. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/tokenizer.json +0 -0
  43. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/tokenizer_config.json +23 -0
  44. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/trainer_state.json +188 -0
  45. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/training_args.bin +3 -0
  46. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/config.json +155 -0
  47. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/model.safetensors +3 -0
  48. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/optimizer.pt +3 -0
  49. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/rng_state.pth +3 -0
  50. nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/scheduler.pt +3 -0
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/config.json ADDED
@@ -0,0 +1,155 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaAdvancedSpanClassifier"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": null,
7
+ "dtype": "float32",
8
+ "eos_token_id": null,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "id2label": {
13
+ "0": "O",
14
+ "1": "ACCOUNT_NUMBER",
15
+ "2": "AGE",
16
+ "3": "API_KEY",
17
+ "4": "BANK_ROUTING_NUMBER",
18
+ "5": "BIOMETRIC_IDENTIFIER",
19
+ "6": "BLOOD_TYPE",
20
+ "7": "CERTIFICATE_LICENSE_NUMBER",
21
+ "8": "CITY",
22
+ "9": "COMPANY_NAME",
23
+ "10": "COORDINATE",
24
+ "11": "COUNTRY",
25
+ "12": "COUNTY",
26
+ "13": "CREDIT_DEBIT_CARD",
27
+ "14": "CUSTOMER_ID",
28
+ "15": "CVV",
29
+ "16": "DATE",
30
+ "17": "DATE_OF_BIRTH",
31
+ "18": "DATE_TIME",
32
+ "19": "DEVICE_IDENTIFIER",
33
+ "20": "EDUCATION_LEVEL",
34
+ "21": "EMAIL",
35
+ "22": "EMPLOYEE_ID",
36
+ "23": "EMPLOYMENT_STATUS",
37
+ "24": "FAX_NUMBER",
38
+ "25": "FIRST_NAME",
39
+ "26": "GENDER",
40
+ "27": "HEALTH_PLAN_BENEFICIARY_NUMBER",
41
+ "28": "HTTP_COOKIE",
42
+ "29": "IPV4",
43
+ "30": "IPV6",
44
+ "31": "LANGUAGE",
45
+ "32": "LAST_NAME",
46
+ "33": "LICENSE_PLATE",
47
+ "34": "MAC_ADDRESS",
48
+ "35": "MEDICAL_RECORD_NUMBER",
49
+ "36": "NATIONAL_ID",
50
+ "37": "OCCUPATION",
51
+ "38": "PASSWORD",
52
+ "39": "PHONE_NUMBER",
53
+ "40": "PIN",
54
+ "41": "POLITICAL_VIEW",
55
+ "42": "POSTCODE",
56
+ "43": "RACE_ETHNICITY",
57
+ "44": "RELIGIOUS_BELIEF",
58
+ "45": "SEXUALITY",
59
+ "46": "SSN",
60
+ "47": "STATE",
61
+ "48": "STREET_ADDRESS",
62
+ "49": "SWIFT_BIC",
63
+ "50": "TAX_ID",
64
+ "51": "TIME",
65
+ "52": "UNIQUE_ID",
66
+ "53": "URL",
67
+ "54": "USER_NAME",
68
+ "55": "VEHICLE_IDENTIFIER"
69
+ },
70
+ "initializer_range": 0.02,
71
+ "intermediate_size": 3072,
72
+ "label2id": {
73
+ "ACCOUNT_NUMBER": 1,
74
+ "AGE": 2,
75
+ "API_KEY": 3,
76
+ "BANK_ROUTING_NUMBER": 4,
77
+ "BIOMETRIC_IDENTIFIER": 5,
78
+ "BLOOD_TYPE": 6,
79
+ "CERTIFICATE_LICENSE_NUMBER": 7,
80
+ "CITY": 8,
81
+ "COMPANY_NAME": 9,
82
+ "COORDINATE": 10,
83
+ "COUNTRY": 11,
84
+ "COUNTY": 12,
85
+ "CREDIT_DEBIT_CARD": 13,
86
+ "CUSTOMER_ID": 14,
87
+ "CVV": 15,
88
+ "DATE": 16,
89
+ "DATE_OF_BIRTH": 17,
90
+ "DATE_TIME": 18,
91
+ "DEVICE_IDENTIFIER": 19,
92
+ "EDUCATION_LEVEL": 20,
93
+ "EMAIL": 21,
94
+ "EMPLOYEE_ID": 22,
95
+ "EMPLOYMENT_STATUS": 23,
96
+ "FAX_NUMBER": 24,
97
+ "FIRST_NAME": 25,
98
+ "GENDER": 26,
99
+ "HEALTH_PLAN_BENEFICIARY_NUMBER": 27,
100
+ "HTTP_COOKIE": 28,
101
+ "IPV4": 29,
102
+ "IPV6": 30,
103
+ "LANGUAGE": 31,
104
+ "LAST_NAME": 32,
105
+ "LICENSE_PLATE": 33,
106
+ "MAC_ADDRESS": 34,
107
+ "MEDICAL_RECORD_NUMBER": 35,
108
+ "NATIONAL_ID": 36,
109
+ "O": 0,
110
+ "OCCUPATION": 37,
111
+ "PASSWORD": 38,
112
+ "PHONE_NUMBER": 39,
113
+ "PIN": 40,
114
+ "POLITICAL_VIEW": 41,
115
+ "POSTCODE": 42,
116
+ "RACE_ETHNICITY": 43,
117
+ "RELIGIOUS_BELIEF": 44,
118
+ "SEXUALITY": 45,
119
+ "SSN": 46,
120
+ "STATE": 47,
121
+ "STREET_ADDRESS": 48,
122
+ "SWIFT_BIC": 49,
123
+ "TAX_ID": 50,
124
+ "TIME": 51,
125
+ "UNIQUE_ID": 52,
126
+ "URL": 53,
127
+ "USER_NAME": 54,
128
+ "VEHICLE_IDENTIFIER": 55
129
+ },
130
+ "layer_norm_eps": 1e-07,
131
+ "legacy": true,
132
+ "max_position_embeddings": 512,
133
+ "max_relative_positions": -1,
134
+ "model_type": "deberta-v2",
135
+ "norm_rel_ebd": "layer_norm",
136
+ "num_attention_heads": 12,
137
+ "num_hidden_layers": 12,
138
+ "pad_token_id": 0,
139
+ "pooler_dropout": 0,
140
+ "pooler_hidden_act": "gelu",
141
+ "pooler_hidden_size": 768,
142
+ "pos_att_type": [
143
+ "p2c",
144
+ "c2p"
145
+ ],
146
+ "position_biased_input": false,
147
+ "position_buckets": 256,
148
+ "relative_attention": true,
149
+ "share_att_key": true,
150
+ "tie_word_embeddings": true,
151
+ "transformers_version": "5.3.0",
152
+ "type_vocab_size": 0,
153
+ "use_cache": false,
154
+ "vocab_size": 128100
155
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76f75c8534f981e23f7fccc1ab0f7f9084396c3c5f3a36078c03b2bc09c365da
3
+ size 741501240
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e0a4d7e0ae933fc53e94d922bc8794b03c3003e800ff8c676aab57b230b9502
3
+ size 1483126155
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d04c045bc2cee166dddc907e324bf95386c7c6f6c5fb0e2ffedf11267f26c246
3
+ size 14645
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fbd98a0874b454541cfb27cce18d68c0c5c3b0b13712840db50ce463bdbb579e
3
+ size 1465
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "backend": "tokenizers",
4
+ "bos_token": "[CLS]",
5
+ "cls_token": "[CLS]",
6
+ "do_lower_case": false,
7
+ "eos_token": "[SEP]",
8
+ "extra_special_tokens": [
9
+ "[PAD]",
10
+ "[CLS]",
11
+ "[SEP]"
12
+ ],
13
+ "is_local": false,
14
+ "mask_token": "[MASK]",
15
+ "model_max_length": 1000000000000000019884624838656,
16
+ "pad_token": "[PAD]",
17
+ "sep_token": "[SEP]",
18
+ "split_by_punct": false,
19
+ "tokenizer_class": "DebertaV2Tokenizer",
20
+ "unk_id": 3,
21
+ "unk_token": "[UNK]",
22
+ "vocab_type": "spm"
23
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/trainer_state.json ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.17777777777777778,
6
+ "eval_steps": 500,
7
+ "global_step": 1000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.08888888888888889,
14
+ "grad_norm": 0.003407392418012023,
15
+ "learning_rate": 1.911288888888889e-05,
16
+ "loss": 0.004852625846862793,
17
+ "step": 500
18
+ },
19
+ {
20
+ "epoch": 0.17777777777777778,
21
+ "grad_norm": 0.0030484353192150593,
22
+ "learning_rate": 1.8224e-05,
23
+ "loss": 0.0013473896980285645,
24
+ "step": 1000
25
+ }
26
+ ],
27
+ "logging_steps": 500,
28
+ "max_steps": 11250,
29
+ "num_input_tokens_seen": 0,
30
+ "num_train_epochs": 2,
31
+ "save_steps": 500,
32
+ "stateful_callbacks": {
33
+ "TrainerControl": {
34
+ "args": {
35
+ "should_epoch_stop": false,
36
+ "should_evaluate": false,
37
+ "should_log": false,
38
+ "should_save": true,
39
+ "should_training_stop": false
40
+ },
41
+ "attributes": {}
42
+ }
43
+ },
44
+ "total_flos": 8450120860925952.0,
45
+ "train_batch_size": 32,
46
+ "trial_name": null,
47
+ "trial_params": null
48
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ed75abcfe811d82fa424040c0ba3f7056bfdd8999b142c56688e699c57934f8
3
+ size 5201
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/config.json ADDED
@@ -0,0 +1,155 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaAdvancedSpanClassifier"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": null,
7
+ "dtype": "float32",
8
+ "eos_token_id": null,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "id2label": {
13
+ "0": "O",
14
+ "1": "ACCOUNT_NUMBER",
15
+ "2": "AGE",
16
+ "3": "API_KEY",
17
+ "4": "BANK_ROUTING_NUMBER",
18
+ "5": "BIOMETRIC_IDENTIFIER",
19
+ "6": "BLOOD_TYPE",
20
+ "7": "CERTIFICATE_LICENSE_NUMBER",
21
+ "8": "CITY",
22
+ "9": "COMPANY_NAME",
23
+ "10": "COORDINATE",
24
+ "11": "COUNTRY",
25
+ "12": "COUNTY",
26
+ "13": "CREDIT_DEBIT_CARD",
27
+ "14": "CUSTOMER_ID",
28
+ "15": "CVV",
29
+ "16": "DATE",
30
+ "17": "DATE_OF_BIRTH",
31
+ "18": "DATE_TIME",
32
+ "19": "DEVICE_IDENTIFIER",
33
+ "20": "EDUCATION_LEVEL",
34
+ "21": "EMAIL",
35
+ "22": "EMPLOYEE_ID",
36
+ "23": "EMPLOYMENT_STATUS",
37
+ "24": "FAX_NUMBER",
38
+ "25": "FIRST_NAME",
39
+ "26": "GENDER",
40
+ "27": "HEALTH_PLAN_BENEFICIARY_NUMBER",
41
+ "28": "HTTP_COOKIE",
42
+ "29": "IPV4",
43
+ "30": "IPV6",
44
+ "31": "LANGUAGE",
45
+ "32": "LAST_NAME",
46
+ "33": "LICENSE_PLATE",
47
+ "34": "MAC_ADDRESS",
48
+ "35": "MEDICAL_RECORD_NUMBER",
49
+ "36": "NATIONAL_ID",
50
+ "37": "OCCUPATION",
51
+ "38": "PASSWORD",
52
+ "39": "PHONE_NUMBER",
53
+ "40": "PIN",
54
+ "41": "POLITICAL_VIEW",
55
+ "42": "POSTCODE",
56
+ "43": "RACE_ETHNICITY",
57
+ "44": "RELIGIOUS_BELIEF",
58
+ "45": "SEXUALITY",
59
+ "46": "SSN",
60
+ "47": "STATE",
61
+ "48": "STREET_ADDRESS",
62
+ "49": "SWIFT_BIC",
63
+ "50": "TAX_ID",
64
+ "51": "TIME",
65
+ "52": "UNIQUE_ID",
66
+ "53": "URL",
67
+ "54": "USER_NAME",
68
+ "55": "VEHICLE_IDENTIFIER"
69
+ },
70
+ "initializer_range": 0.02,
71
+ "intermediate_size": 3072,
72
+ "label2id": {
73
+ "ACCOUNT_NUMBER": 1,
74
+ "AGE": 2,
75
+ "API_KEY": 3,
76
+ "BANK_ROUTING_NUMBER": 4,
77
+ "BIOMETRIC_IDENTIFIER": 5,
78
+ "BLOOD_TYPE": 6,
79
+ "CERTIFICATE_LICENSE_NUMBER": 7,
80
+ "CITY": 8,
81
+ "COMPANY_NAME": 9,
82
+ "COORDINATE": 10,
83
+ "COUNTRY": 11,
84
+ "COUNTY": 12,
85
+ "CREDIT_DEBIT_CARD": 13,
86
+ "CUSTOMER_ID": 14,
87
+ "CVV": 15,
88
+ "DATE": 16,
89
+ "DATE_OF_BIRTH": 17,
90
+ "DATE_TIME": 18,
91
+ "DEVICE_IDENTIFIER": 19,
92
+ "EDUCATION_LEVEL": 20,
93
+ "EMAIL": 21,
94
+ "EMPLOYEE_ID": 22,
95
+ "EMPLOYMENT_STATUS": 23,
96
+ "FAX_NUMBER": 24,
97
+ "FIRST_NAME": 25,
98
+ "GENDER": 26,
99
+ "HEALTH_PLAN_BENEFICIARY_NUMBER": 27,
100
+ "HTTP_COOKIE": 28,
101
+ "IPV4": 29,
102
+ "IPV6": 30,
103
+ "LANGUAGE": 31,
104
+ "LAST_NAME": 32,
105
+ "LICENSE_PLATE": 33,
106
+ "MAC_ADDRESS": 34,
107
+ "MEDICAL_RECORD_NUMBER": 35,
108
+ "NATIONAL_ID": 36,
109
+ "O": 0,
110
+ "OCCUPATION": 37,
111
+ "PASSWORD": 38,
112
+ "PHONE_NUMBER": 39,
113
+ "PIN": 40,
114
+ "POLITICAL_VIEW": 41,
115
+ "POSTCODE": 42,
116
+ "RACE_ETHNICITY": 43,
117
+ "RELIGIOUS_BELIEF": 44,
118
+ "SEXUALITY": 45,
119
+ "SSN": 46,
120
+ "STATE": 47,
121
+ "STREET_ADDRESS": 48,
122
+ "SWIFT_BIC": 49,
123
+ "TAX_ID": 50,
124
+ "TIME": 51,
125
+ "UNIQUE_ID": 52,
126
+ "URL": 53,
127
+ "USER_NAME": 54,
128
+ "VEHICLE_IDENTIFIER": 55
129
+ },
130
+ "layer_norm_eps": 1e-07,
131
+ "legacy": true,
132
+ "max_position_embeddings": 512,
133
+ "max_relative_positions": -1,
134
+ "model_type": "deberta-v2",
135
+ "norm_rel_ebd": "layer_norm",
136
+ "num_attention_heads": 12,
137
+ "num_hidden_layers": 12,
138
+ "pad_token_id": 0,
139
+ "pooler_dropout": 0,
140
+ "pooler_hidden_act": "gelu",
141
+ "pooler_hidden_size": 768,
142
+ "pos_att_type": [
143
+ "p2c",
144
+ "c2p"
145
+ ],
146
+ "position_biased_input": false,
147
+ "position_buckets": 256,
148
+ "relative_attention": true,
149
+ "share_att_key": true,
150
+ "tie_word_embeddings": true,
151
+ "transformers_version": "5.3.0",
152
+ "type_vocab_size": 0,
153
+ "use_cache": false,
154
+ "vocab_size": 128100
155
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ec09039d4ae8b1555b6ecd016d67a1043f39fae12235664621a2ff6b16a8fef8
3
+ size 741501240
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d8d3b52a582c9e8c3dd2da496e60e7caae3122248c8bcaddad40995b92d7125
3
+ size 1483126155
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7cb31df8160280960dc05170b250a06a0a3727a8d8fadeb093665dfed6cfe336
3
+ size 14645
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5feec1def9cb568d079c80b60d2a59cc4dbeadb9d634921c1b730407492f1792
3
+ size 1465
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/tokenizer_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "backend": "tokenizers",
4
+ "bos_token": "[CLS]",
5
+ "cls_token": "[CLS]",
6
+ "do_lower_case": false,
7
+ "eos_token": "[SEP]",
8
+ "extra_special_tokens": [
9
+ "[PAD]",
10
+ "[CLS]",
11
+ "[SEP]"
12
+ ],
13
+ "is_local": false,
14
+ "mask_token": "[MASK]",
15
+ "model_max_length": 1000000000000000019884624838656,
16
+ "pad_token": "[PAD]",
17
+ "sep_token": "[SEP]",
18
+ "split_by_punct": false,
19
+ "tokenizer_class": "DebertaV2Tokenizer",
20
+ "unk_id": 3,
21
+ "unk_token": "[UNK]",
22
+ "vocab_type": "spm"
23
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/trainer_state.json ADDED
@@ -0,0 +1,174 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.7777777777777777,
6
+ "eval_steps": 500,
7
+ "global_step": 10000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.08888888888888889,
14
+ "grad_norm": 0.003407392418012023,
15
+ "learning_rate": 1.911288888888889e-05,
16
+ "loss": 0.004852625846862793,
17
+ "step": 500
18
+ },
19
+ {
20
+ "epoch": 0.17777777777777778,
21
+ "grad_norm": 0.0030484353192150593,
22
+ "learning_rate": 1.8224e-05,
23
+ "loss": 0.0013473896980285645,
24
+ "step": 1000
25
+ },
26
+ {
27
+ "epoch": 0.26666666666666666,
28
+ "grad_norm": 0.0035167669411748648,
29
+ "learning_rate": 1.733511111111111e-05,
30
+ "loss": 0.0006659740805625916,
31
+ "step": 1500
32
+ },
33
+ {
34
+ "epoch": 0.35555555555555557,
35
+ "grad_norm": 0.0029440049547702074,
36
+ "learning_rate": 1.6446222222222223e-05,
37
+ "loss": 0.00036768344044685363,
38
+ "step": 2000
39
+ },
40
+ {
41
+ "epoch": 0.4444444444444444,
42
+ "grad_norm": 0.0034908612724393606,
43
+ "learning_rate": 1.5557333333333334e-05,
44
+ "loss": 0.00025459024310112,
45
+ "step": 2500
46
+ },
47
+ {
48
+ "epoch": 0.5333333333333333,
49
+ "grad_norm": 0.0030976070556789637,
50
+ "learning_rate": 1.4668444444444446e-05,
51
+ "loss": 0.00019871483743190765,
52
+ "step": 3000
53
+ },
54
+ {
55
+ "epoch": 0.6222222222222222,
56
+ "grad_norm": 0.0017464511329308152,
57
+ "learning_rate": 1.3779555555555557e-05,
58
+ "loss": 0.000175091952085495,
59
+ "step": 3500
60
+ },
61
+ {
62
+ "epoch": 0.7111111111111111,
63
+ "grad_norm": 0.003212156007066369,
64
+ "learning_rate": 1.2890666666666667e-05,
65
+ "loss": 0.00014939682185649873,
66
+ "step": 4000
67
+ },
68
+ {
69
+ "epoch": 0.8,
70
+ "grad_norm": 0.0026216867845505476,
71
+ "learning_rate": 1.2001777777777778e-05,
72
+ "loss": 0.00013630734384059907,
73
+ "step": 4500
74
+ },
75
+ {
76
+ "epoch": 0.8888888888888888,
77
+ "grad_norm": 0.0014595076208934188,
78
+ "learning_rate": 1.1112888888888891e-05,
79
+ "loss": 0.00013093747198581696,
80
+ "step": 5000
81
+ },
82
+ {
83
+ "epoch": 0.9777777777777777,
84
+ "grad_norm": 0.0013789460062980652,
85
+ "learning_rate": 1.0224e-05,
86
+ "loss": 0.0001222904920578003,
87
+ "step": 5500
88
+ },
89
+ {
90
+ "epoch": 1.0666666666666667,
91
+ "grad_norm": 0.0016316684195771813,
92
+ "learning_rate": 9.335111111111112e-06,
93
+ "loss": 0.00011386586725711822,
94
+ "step": 6000
95
+ },
96
+ {
97
+ "epoch": 1.1555555555555554,
98
+ "grad_norm": 0.0016652554040774703,
99
+ "learning_rate": 8.446222222222223e-06,
100
+ "loss": 0.00010670755803585053,
101
+ "step": 6500
102
+ },
103
+ {
104
+ "epoch": 1.2444444444444445,
105
+ "grad_norm": 0.0012231464497745037,
106
+ "learning_rate": 7.557333333333334e-06,
107
+ "loss": 0.00010442142933607102,
108
+ "step": 7000
109
+ },
110
+ {
111
+ "epoch": 1.3333333333333333,
112
+ "grad_norm": 0.002133500762283802,
113
+ "learning_rate": 6.668444444444445e-06,
114
+ "loss": 0.00010099935531616211,
115
+ "step": 7500
116
+ },
117
+ {
118
+ "epoch": 1.4222222222222223,
119
+ "grad_norm": 0.0011294716969132423,
120
+ "learning_rate": 5.779555555555556e-06,
121
+ "loss": 9.695450961589813e-05,
122
+ "step": 8000
123
+ },
124
+ {
125
+ "epoch": 1.511111111111111,
126
+ "grad_norm": 0.0015969782834872603,
127
+ "learning_rate": 4.890666666666667e-06,
128
+ "loss": 9.648169577121735e-05,
129
+ "step": 8500
130
+ },
131
+ {
132
+ "epoch": 1.6,
133
+ "grad_norm": 0.0016485441010445356,
134
+ "learning_rate": 4.001777777777778e-06,
135
+ "loss": 9.064872562885284e-05,
136
+ "step": 9000
137
+ },
138
+ {
139
+ "epoch": 1.6888888888888889,
140
+ "grad_norm": 0.002174828900024295,
141
+ "learning_rate": 3.1128888888888893e-06,
142
+ "loss": 8.987343311309815e-05,
143
+ "step": 9500
144
+ },
145
+ {
146
+ "epoch": 1.7777777777777777,
147
+ "grad_norm": 0.002475686138495803,
148
+ "learning_rate": 2.2240000000000002e-06,
149
+ "loss": 8.707723021507264e-05,
150
+ "step": 10000
151
+ }
152
+ ],
153
+ "logging_steps": 500,
154
+ "max_steps": 11250,
155
+ "num_input_tokens_seen": 0,
156
+ "num_train_epochs": 2,
157
+ "save_steps": 500,
158
+ "stateful_callbacks": {
159
+ "TrainerControl": {
160
+ "args": {
161
+ "should_epoch_stop": false,
162
+ "should_evaluate": false,
163
+ "should_log": false,
164
+ "should_save": true,
165
+ "should_training_stop": false
166
+ },
167
+ "attributes": {}
168
+ }
169
+ },
170
+ "total_flos": 8.450353577020416e+16,
171
+ "train_batch_size": 32,
172
+ "trial_name": null,
173
+ "trial_params": null
174
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ed75abcfe811d82fa424040c0ba3f7056bfdd8999b142c56688e699c57934f8
3
+ size 5201
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/config.json ADDED
@@ -0,0 +1,155 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaAdvancedSpanClassifier"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": null,
7
+ "dtype": "float32",
8
+ "eos_token_id": null,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "id2label": {
13
+ "0": "O",
14
+ "1": "ACCOUNT_NUMBER",
15
+ "2": "AGE",
16
+ "3": "API_KEY",
17
+ "4": "BANK_ROUTING_NUMBER",
18
+ "5": "BIOMETRIC_IDENTIFIER",
19
+ "6": "BLOOD_TYPE",
20
+ "7": "CERTIFICATE_LICENSE_NUMBER",
21
+ "8": "CITY",
22
+ "9": "COMPANY_NAME",
23
+ "10": "COORDINATE",
24
+ "11": "COUNTRY",
25
+ "12": "COUNTY",
26
+ "13": "CREDIT_DEBIT_CARD",
27
+ "14": "CUSTOMER_ID",
28
+ "15": "CVV",
29
+ "16": "DATE",
30
+ "17": "DATE_OF_BIRTH",
31
+ "18": "DATE_TIME",
32
+ "19": "DEVICE_IDENTIFIER",
33
+ "20": "EDUCATION_LEVEL",
34
+ "21": "EMAIL",
35
+ "22": "EMPLOYEE_ID",
36
+ "23": "EMPLOYMENT_STATUS",
37
+ "24": "FAX_NUMBER",
38
+ "25": "FIRST_NAME",
39
+ "26": "GENDER",
40
+ "27": "HEALTH_PLAN_BENEFICIARY_NUMBER",
41
+ "28": "HTTP_COOKIE",
42
+ "29": "IPV4",
43
+ "30": "IPV6",
44
+ "31": "LANGUAGE",
45
+ "32": "LAST_NAME",
46
+ "33": "LICENSE_PLATE",
47
+ "34": "MAC_ADDRESS",
48
+ "35": "MEDICAL_RECORD_NUMBER",
49
+ "36": "NATIONAL_ID",
50
+ "37": "OCCUPATION",
51
+ "38": "PASSWORD",
52
+ "39": "PHONE_NUMBER",
53
+ "40": "PIN",
54
+ "41": "POLITICAL_VIEW",
55
+ "42": "POSTCODE",
56
+ "43": "RACE_ETHNICITY",
57
+ "44": "RELIGIOUS_BELIEF",
58
+ "45": "SEXUALITY",
59
+ "46": "SSN",
60
+ "47": "STATE",
61
+ "48": "STREET_ADDRESS",
62
+ "49": "SWIFT_BIC",
63
+ "50": "TAX_ID",
64
+ "51": "TIME",
65
+ "52": "UNIQUE_ID",
66
+ "53": "URL",
67
+ "54": "USER_NAME",
68
+ "55": "VEHICLE_IDENTIFIER"
69
+ },
70
+ "initializer_range": 0.02,
71
+ "intermediate_size": 3072,
72
+ "label2id": {
73
+ "ACCOUNT_NUMBER": 1,
74
+ "AGE": 2,
75
+ "API_KEY": 3,
76
+ "BANK_ROUTING_NUMBER": 4,
77
+ "BIOMETRIC_IDENTIFIER": 5,
78
+ "BLOOD_TYPE": 6,
79
+ "CERTIFICATE_LICENSE_NUMBER": 7,
80
+ "CITY": 8,
81
+ "COMPANY_NAME": 9,
82
+ "COORDINATE": 10,
83
+ "COUNTRY": 11,
84
+ "COUNTY": 12,
85
+ "CREDIT_DEBIT_CARD": 13,
86
+ "CUSTOMER_ID": 14,
87
+ "CVV": 15,
88
+ "DATE": 16,
89
+ "DATE_OF_BIRTH": 17,
90
+ "DATE_TIME": 18,
91
+ "DEVICE_IDENTIFIER": 19,
92
+ "EDUCATION_LEVEL": 20,
93
+ "EMAIL": 21,
94
+ "EMPLOYEE_ID": 22,
95
+ "EMPLOYMENT_STATUS": 23,
96
+ "FAX_NUMBER": 24,
97
+ "FIRST_NAME": 25,
98
+ "GENDER": 26,
99
+ "HEALTH_PLAN_BENEFICIARY_NUMBER": 27,
100
+ "HTTP_COOKIE": 28,
101
+ "IPV4": 29,
102
+ "IPV6": 30,
103
+ "LANGUAGE": 31,
104
+ "LAST_NAME": 32,
105
+ "LICENSE_PLATE": 33,
106
+ "MAC_ADDRESS": 34,
107
+ "MEDICAL_RECORD_NUMBER": 35,
108
+ "NATIONAL_ID": 36,
109
+ "O": 0,
110
+ "OCCUPATION": 37,
111
+ "PASSWORD": 38,
112
+ "PHONE_NUMBER": 39,
113
+ "PIN": 40,
114
+ "POLITICAL_VIEW": 41,
115
+ "POSTCODE": 42,
116
+ "RACE_ETHNICITY": 43,
117
+ "RELIGIOUS_BELIEF": 44,
118
+ "SEXUALITY": 45,
119
+ "SSN": 46,
120
+ "STATE": 47,
121
+ "STREET_ADDRESS": 48,
122
+ "SWIFT_BIC": 49,
123
+ "TAX_ID": 50,
124
+ "TIME": 51,
125
+ "UNIQUE_ID": 52,
126
+ "URL": 53,
127
+ "USER_NAME": 54,
128
+ "VEHICLE_IDENTIFIER": 55
129
+ },
130
+ "layer_norm_eps": 1e-07,
131
+ "legacy": true,
132
+ "max_position_embeddings": 512,
133
+ "max_relative_positions": -1,
134
+ "model_type": "deberta-v2",
135
+ "norm_rel_ebd": "layer_norm",
136
+ "num_attention_heads": 12,
137
+ "num_hidden_layers": 12,
138
+ "pad_token_id": 0,
139
+ "pooler_dropout": 0,
140
+ "pooler_hidden_act": "gelu",
141
+ "pooler_hidden_size": 768,
142
+ "pos_att_type": [
143
+ "p2c",
144
+ "c2p"
145
+ ],
146
+ "position_biased_input": false,
147
+ "position_buckets": 256,
148
+ "relative_attention": true,
149
+ "share_att_key": true,
150
+ "tie_word_embeddings": true,
151
+ "transformers_version": "5.3.0",
152
+ "type_vocab_size": 0,
153
+ "use_cache": false,
154
+ "vocab_size": 128100
155
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f7f607062e4f4872884b52ee41cf579f49f6bd860c039c046beb9c88e1142a9
3
+ size 741501240
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f79c7b8f8c84f83eac969aed46cef75783bb41d463faaa0b8bf796eb4d342a83
3
+ size 1483126155
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:207e2c29fad9a6c38df605f97d0ef5b6272b7e2734633c40a3a087ad075f38ff
3
+ size 14645
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a5c45bd7c8e043887f27e164946b4a5d4ce0a57f2621887939b57adb28d3273
3
+ size 1465
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/tokenizer_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "backend": "tokenizers",
4
+ "bos_token": "[CLS]",
5
+ "cls_token": "[CLS]",
6
+ "do_lower_case": false,
7
+ "eos_token": "[SEP]",
8
+ "extra_special_tokens": [
9
+ "[PAD]",
10
+ "[CLS]",
11
+ "[SEP]"
12
+ ],
13
+ "is_local": false,
14
+ "mask_token": "[MASK]",
15
+ "model_max_length": 1000000000000000019884624838656,
16
+ "pad_token": "[PAD]",
17
+ "sep_token": "[SEP]",
18
+ "split_by_punct": false,
19
+ "tokenizer_class": "DebertaV2Tokenizer",
20
+ "unk_id": 3,
21
+ "unk_token": "[UNK]",
22
+ "vocab_type": "spm"
23
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/trainer_state.json ADDED
@@ -0,0 +1,181 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.8666666666666667,
6
+ "eval_steps": 500,
7
+ "global_step": 10500,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.08888888888888889,
14
+ "grad_norm": 0.003407392418012023,
15
+ "learning_rate": 1.911288888888889e-05,
16
+ "loss": 0.004852625846862793,
17
+ "step": 500
18
+ },
19
+ {
20
+ "epoch": 0.17777777777777778,
21
+ "grad_norm": 0.0030484353192150593,
22
+ "learning_rate": 1.8224e-05,
23
+ "loss": 0.0013473896980285645,
24
+ "step": 1000
25
+ },
26
+ {
27
+ "epoch": 0.26666666666666666,
28
+ "grad_norm": 0.0035167669411748648,
29
+ "learning_rate": 1.733511111111111e-05,
30
+ "loss": 0.0006659740805625916,
31
+ "step": 1500
32
+ },
33
+ {
34
+ "epoch": 0.35555555555555557,
35
+ "grad_norm": 0.0029440049547702074,
36
+ "learning_rate": 1.6446222222222223e-05,
37
+ "loss": 0.00036768344044685363,
38
+ "step": 2000
39
+ },
40
+ {
41
+ "epoch": 0.4444444444444444,
42
+ "grad_norm": 0.0034908612724393606,
43
+ "learning_rate": 1.5557333333333334e-05,
44
+ "loss": 0.00025459024310112,
45
+ "step": 2500
46
+ },
47
+ {
48
+ "epoch": 0.5333333333333333,
49
+ "grad_norm": 0.0030976070556789637,
50
+ "learning_rate": 1.4668444444444446e-05,
51
+ "loss": 0.00019871483743190765,
52
+ "step": 3000
53
+ },
54
+ {
55
+ "epoch": 0.6222222222222222,
56
+ "grad_norm": 0.0017464511329308152,
57
+ "learning_rate": 1.3779555555555557e-05,
58
+ "loss": 0.000175091952085495,
59
+ "step": 3500
60
+ },
61
+ {
62
+ "epoch": 0.7111111111111111,
63
+ "grad_norm": 0.003212156007066369,
64
+ "learning_rate": 1.2890666666666667e-05,
65
+ "loss": 0.00014939682185649873,
66
+ "step": 4000
67
+ },
68
+ {
69
+ "epoch": 0.8,
70
+ "grad_norm": 0.0026216867845505476,
71
+ "learning_rate": 1.2001777777777778e-05,
72
+ "loss": 0.00013630734384059907,
73
+ "step": 4500
74
+ },
75
+ {
76
+ "epoch": 0.8888888888888888,
77
+ "grad_norm": 0.0014595076208934188,
78
+ "learning_rate": 1.1112888888888891e-05,
79
+ "loss": 0.00013093747198581696,
80
+ "step": 5000
81
+ },
82
+ {
83
+ "epoch": 0.9777777777777777,
84
+ "grad_norm": 0.0013789460062980652,
85
+ "learning_rate": 1.0224e-05,
86
+ "loss": 0.0001222904920578003,
87
+ "step": 5500
88
+ },
89
+ {
90
+ "epoch": 1.0666666666666667,
91
+ "grad_norm": 0.0016316684195771813,
92
+ "learning_rate": 9.335111111111112e-06,
93
+ "loss": 0.00011386586725711822,
94
+ "step": 6000
95
+ },
96
+ {
97
+ "epoch": 1.1555555555555554,
98
+ "grad_norm": 0.0016652554040774703,
99
+ "learning_rate": 8.446222222222223e-06,
100
+ "loss": 0.00010670755803585053,
101
+ "step": 6500
102
+ },
103
+ {
104
+ "epoch": 1.2444444444444445,
105
+ "grad_norm": 0.0012231464497745037,
106
+ "learning_rate": 7.557333333333334e-06,
107
+ "loss": 0.00010442142933607102,
108
+ "step": 7000
109
+ },
110
+ {
111
+ "epoch": 1.3333333333333333,
112
+ "grad_norm": 0.002133500762283802,
113
+ "learning_rate": 6.668444444444445e-06,
114
+ "loss": 0.00010099935531616211,
115
+ "step": 7500
116
+ },
117
+ {
118
+ "epoch": 1.4222222222222223,
119
+ "grad_norm": 0.0011294716969132423,
120
+ "learning_rate": 5.779555555555556e-06,
121
+ "loss": 9.695450961589813e-05,
122
+ "step": 8000
123
+ },
124
+ {
125
+ "epoch": 1.511111111111111,
126
+ "grad_norm": 0.0015969782834872603,
127
+ "learning_rate": 4.890666666666667e-06,
128
+ "loss": 9.648169577121735e-05,
129
+ "step": 8500
130
+ },
131
+ {
132
+ "epoch": 1.6,
133
+ "grad_norm": 0.0016485441010445356,
134
+ "learning_rate": 4.001777777777778e-06,
135
+ "loss": 9.064872562885284e-05,
136
+ "step": 9000
137
+ },
138
+ {
139
+ "epoch": 1.6888888888888889,
140
+ "grad_norm": 0.002174828900024295,
141
+ "learning_rate": 3.1128888888888893e-06,
142
+ "loss": 8.987343311309815e-05,
143
+ "step": 9500
144
+ },
145
+ {
146
+ "epoch": 1.7777777777777777,
147
+ "grad_norm": 0.002475686138495803,
148
+ "learning_rate": 2.2240000000000002e-06,
149
+ "loss": 8.707723021507264e-05,
150
+ "step": 10000
151
+ },
152
+ {
153
+ "epoch": 1.8666666666666667,
154
+ "grad_norm": 0.0028533614240586758,
155
+ "learning_rate": 1.3351111111111111e-06,
156
+ "loss": 8.645709604024888e-05,
157
+ "step": 10500
158
+ }
159
+ ],
160
+ "logging_steps": 500,
161
+ "max_steps": 11250,
162
+ "num_input_tokens_seen": 0,
163
+ "num_train_epochs": 2,
164
+ "save_steps": 500,
165
+ "stateful_callbacks": {
166
+ "TrainerControl": {
167
+ "args": {
168
+ "should_epoch_stop": false,
169
+ "should_evaluate": false,
170
+ "should_log": false,
171
+ "should_save": true,
172
+ "should_training_stop": false
173
+ },
174
+ "attributes": {}
175
+ }
176
+ },
177
+ "total_flos": 8.873474655459226e+16,
178
+ "train_batch_size": 32,
179
+ "trial_name": null,
180
+ "trial_params": null
181
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ed75abcfe811d82fa424040c0ba3f7056bfdd8999b142c56688e699c57934f8
3
+ size 5201
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/config.json ADDED
@@ -0,0 +1,155 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaAdvancedSpanClassifier"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": null,
7
+ "dtype": "float32",
8
+ "eos_token_id": null,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "id2label": {
13
+ "0": "O",
14
+ "1": "ACCOUNT_NUMBER",
15
+ "2": "AGE",
16
+ "3": "API_KEY",
17
+ "4": "BANK_ROUTING_NUMBER",
18
+ "5": "BIOMETRIC_IDENTIFIER",
19
+ "6": "BLOOD_TYPE",
20
+ "7": "CERTIFICATE_LICENSE_NUMBER",
21
+ "8": "CITY",
22
+ "9": "COMPANY_NAME",
23
+ "10": "COORDINATE",
24
+ "11": "COUNTRY",
25
+ "12": "COUNTY",
26
+ "13": "CREDIT_DEBIT_CARD",
27
+ "14": "CUSTOMER_ID",
28
+ "15": "CVV",
29
+ "16": "DATE",
30
+ "17": "DATE_OF_BIRTH",
31
+ "18": "DATE_TIME",
32
+ "19": "DEVICE_IDENTIFIER",
33
+ "20": "EDUCATION_LEVEL",
34
+ "21": "EMAIL",
35
+ "22": "EMPLOYEE_ID",
36
+ "23": "EMPLOYMENT_STATUS",
37
+ "24": "FAX_NUMBER",
38
+ "25": "FIRST_NAME",
39
+ "26": "GENDER",
40
+ "27": "HEALTH_PLAN_BENEFICIARY_NUMBER",
41
+ "28": "HTTP_COOKIE",
42
+ "29": "IPV4",
43
+ "30": "IPV6",
44
+ "31": "LANGUAGE",
45
+ "32": "LAST_NAME",
46
+ "33": "LICENSE_PLATE",
47
+ "34": "MAC_ADDRESS",
48
+ "35": "MEDICAL_RECORD_NUMBER",
49
+ "36": "NATIONAL_ID",
50
+ "37": "OCCUPATION",
51
+ "38": "PASSWORD",
52
+ "39": "PHONE_NUMBER",
53
+ "40": "PIN",
54
+ "41": "POLITICAL_VIEW",
55
+ "42": "POSTCODE",
56
+ "43": "RACE_ETHNICITY",
57
+ "44": "RELIGIOUS_BELIEF",
58
+ "45": "SEXUALITY",
59
+ "46": "SSN",
60
+ "47": "STATE",
61
+ "48": "STREET_ADDRESS",
62
+ "49": "SWIFT_BIC",
63
+ "50": "TAX_ID",
64
+ "51": "TIME",
65
+ "52": "UNIQUE_ID",
66
+ "53": "URL",
67
+ "54": "USER_NAME",
68
+ "55": "VEHICLE_IDENTIFIER"
69
+ },
70
+ "initializer_range": 0.02,
71
+ "intermediate_size": 3072,
72
+ "label2id": {
73
+ "ACCOUNT_NUMBER": 1,
74
+ "AGE": 2,
75
+ "API_KEY": 3,
76
+ "BANK_ROUTING_NUMBER": 4,
77
+ "BIOMETRIC_IDENTIFIER": 5,
78
+ "BLOOD_TYPE": 6,
79
+ "CERTIFICATE_LICENSE_NUMBER": 7,
80
+ "CITY": 8,
81
+ "COMPANY_NAME": 9,
82
+ "COORDINATE": 10,
83
+ "COUNTRY": 11,
84
+ "COUNTY": 12,
85
+ "CREDIT_DEBIT_CARD": 13,
86
+ "CUSTOMER_ID": 14,
87
+ "CVV": 15,
88
+ "DATE": 16,
89
+ "DATE_OF_BIRTH": 17,
90
+ "DATE_TIME": 18,
91
+ "DEVICE_IDENTIFIER": 19,
92
+ "EDUCATION_LEVEL": 20,
93
+ "EMAIL": 21,
94
+ "EMPLOYEE_ID": 22,
95
+ "EMPLOYMENT_STATUS": 23,
96
+ "FAX_NUMBER": 24,
97
+ "FIRST_NAME": 25,
98
+ "GENDER": 26,
99
+ "HEALTH_PLAN_BENEFICIARY_NUMBER": 27,
100
+ "HTTP_COOKIE": 28,
101
+ "IPV4": 29,
102
+ "IPV6": 30,
103
+ "LANGUAGE": 31,
104
+ "LAST_NAME": 32,
105
+ "LICENSE_PLATE": 33,
106
+ "MAC_ADDRESS": 34,
107
+ "MEDICAL_RECORD_NUMBER": 35,
108
+ "NATIONAL_ID": 36,
109
+ "O": 0,
110
+ "OCCUPATION": 37,
111
+ "PASSWORD": 38,
112
+ "PHONE_NUMBER": 39,
113
+ "PIN": 40,
114
+ "POLITICAL_VIEW": 41,
115
+ "POSTCODE": 42,
116
+ "RACE_ETHNICITY": 43,
117
+ "RELIGIOUS_BELIEF": 44,
118
+ "SEXUALITY": 45,
119
+ "SSN": 46,
120
+ "STATE": 47,
121
+ "STREET_ADDRESS": 48,
122
+ "SWIFT_BIC": 49,
123
+ "TAX_ID": 50,
124
+ "TIME": 51,
125
+ "UNIQUE_ID": 52,
126
+ "URL": 53,
127
+ "USER_NAME": 54,
128
+ "VEHICLE_IDENTIFIER": 55
129
+ },
130
+ "layer_norm_eps": 1e-07,
131
+ "legacy": true,
132
+ "max_position_embeddings": 512,
133
+ "max_relative_positions": -1,
134
+ "model_type": "deberta-v2",
135
+ "norm_rel_ebd": "layer_norm",
136
+ "num_attention_heads": 12,
137
+ "num_hidden_layers": 12,
138
+ "pad_token_id": 0,
139
+ "pooler_dropout": 0,
140
+ "pooler_hidden_act": "gelu",
141
+ "pooler_hidden_size": 768,
142
+ "pos_att_type": [
143
+ "p2c",
144
+ "c2p"
145
+ ],
146
+ "position_biased_input": false,
147
+ "position_buckets": 256,
148
+ "relative_attention": true,
149
+ "share_att_key": true,
150
+ "tie_word_embeddings": true,
151
+ "transformers_version": "5.3.0",
152
+ "type_vocab_size": 0,
153
+ "use_cache": false,
154
+ "vocab_size": 128100
155
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1fe3fa4017cbeab13d20d912a5008047e8809537797ce219c9060836f85b89f5
3
+ size 741501240
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a7d0715bfafa14520176d2c3b8b274db7400dcca12fe37bb6f02f66d47bfd2e
3
+ size 1483126155
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b0868e303112da708c2e8bac1a4bdb22b06e348f848011f035901c300cf3dcc1
3
+ size 14645
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1e450f95d37a4e1d6eb7501dcc91be394dcc5ad198718236d56653726e2ad9a
3
+ size 1465
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/tokenizer_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "backend": "tokenizers",
4
+ "bos_token": "[CLS]",
5
+ "cls_token": "[CLS]",
6
+ "do_lower_case": false,
7
+ "eos_token": "[SEP]",
8
+ "extra_special_tokens": [
9
+ "[PAD]",
10
+ "[CLS]",
11
+ "[SEP]"
12
+ ],
13
+ "is_local": false,
14
+ "mask_token": "[MASK]",
15
+ "model_max_length": 1000000000000000019884624838656,
16
+ "pad_token": "[PAD]",
17
+ "sep_token": "[SEP]",
18
+ "split_by_punct": false,
19
+ "tokenizer_class": "DebertaV2Tokenizer",
20
+ "unk_id": 3,
21
+ "unk_token": "[UNK]",
22
+ "vocab_type": "spm"
23
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/trainer_state.json ADDED
@@ -0,0 +1,188 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.9555555555555557,
6
+ "eval_steps": 500,
7
+ "global_step": 11000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.08888888888888889,
14
+ "grad_norm": 0.003407392418012023,
15
+ "learning_rate": 1.911288888888889e-05,
16
+ "loss": 0.004852625846862793,
17
+ "step": 500
18
+ },
19
+ {
20
+ "epoch": 0.17777777777777778,
21
+ "grad_norm": 0.0030484353192150593,
22
+ "learning_rate": 1.8224e-05,
23
+ "loss": 0.0013473896980285645,
24
+ "step": 1000
25
+ },
26
+ {
27
+ "epoch": 0.26666666666666666,
28
+ "grad_norm": 0.0035167669411748648,
29
+ "learning_rate": 1.733511111111111e-05,
30
+ "loss": 0.0006659740805625916,
31
+ "step": 1500
32
+ },
33
+ {
34
+ "epoch": 0.35555555555555557,
35
+ "grad_norm": 0.0029440049547702074,
36
+ "learning_rate": 1.6446222222222223e-05,
37
+ "loss": 0.00036768344044685363,
38
+ "step": 2000
39
+ },
40
+ {
41
+ "epoch": 0.4444444444444444,
42
+ "grad_norm": 0.0034908612724393606,
43
+ "learning_rate": 1.5557333333333334e-05,
44
+ "loss": 0.00025459024310112,
45
+ "step": 2500
46
+ },
47
+ {
48
+ "epoch": 0.5333333333333333,
49
+ "grad_norm": 0.0030976070556789637,
50
+ "learning_rate": 1.4668444444444446e-05,
51
+ "loss": 0.00019871483743190765,
52
+ "step": 3000
53
+ },
54
+ {
55
+ "epoch": 0.6222222222222222,
56
+ "grad_norm": 0.0017464511329308152,
57
+ "learning_rate": 1.3779555555555557e-05,
58
+ "loss": 0.000175091952085495,
59
+ "step": 3500
60
+ },
61
+ {
62
+ "epoch": 0.7111111111111111,
63
+ "grad_norm": 0.003212156007066369,
64
+ "learning_rate": 1.2890666666666667e-05,
65
+ "loss": 0.00014939682185649873,
66
+ "step": 4000
67
+ },
68
+ {
69
+ "epoch": 0.8,
70
+ "grad_norm": 0.0026216867845505476,
71
+ "learning_rate": 1.2001777777777778e-05,
72
+ "loss": 0.00013630734384059907,
73
+ "step": 4500
74
+ },
75
+ {
76
+ "epoch": 0.8888888888888888,
77
+ "grad_norm": 0.0014595076208934188,
78
+ "learning_rate": 1.1112888888888891e-05,
79
+ "loss": 0.00013093747198581696,
80
+ "step": 5000
81
+ },
82
+ {
83
+ "epoch": 0.9777777777777777,
84
+ "grad_norm": 0.0013789460062980652,
85
+ "learning_rate": 1.0224e-05,
86
+ "loss": 0.0001222904920578003,
87
+ "step": 5500
88
+ },
89
+ {
90
+ "epoch": 1.0666666666666667,
91
+ "grad_norm": 0.0016316684195771813,
92
+ "learning_rate": 9.335111111111112e-06,
93
+ "loss": 0.00011386586725711822,
94
+ "step": 6000
95
+ },
96
+ {
97
+ "epoch": 1.1555555555555554,
98
+ "grad_norm": 0.0016652554040774703,
99
+ "learning_rate": 8.446222222222223e-06,
100
+ "loss": 0.00010670755803585053,
101
+ "step": 6500
102
+ },
103
+ {
104
+ "epoch": 1.2444444444444445,
105
+ "grad_norm": 0.0012231464497745037,
106
+ "learning_rate": 7.557333333333334e-06,
107
+ "loss": 0.00010442142933607102,
108
+ "step": 7000
109
+ },
110
+ {
111
+ "epoch": 1.3333333333333333,
112
+ "grad_norm": 0.002133500762283802,
113
+ "learning_rate": 6.668444444444445e-06,
114
+ "loss": 0.00010099935531616211,
115
+ "step": 7500
116
+ },
117
+ {
118
+ "epoch": 1.4222222222222223,
119
+ "grad_norm": 0.0011294716969132423,
120
+ "learning_rate": 5.779555555555556e-06,
121
+ "loss": 9.695450961589813e-05,
122
+ "step": 8000
123
+ },
124
+ {
125
+ "epoch": 1.511111111111111,
126
+ "grad_norm": 0.0015969782834872603,
127
+ "learning_rate": 4.890666666666667e-06,
128
+ "loss": 9.648169577121735e-05,
129
+ "step": 8500
130
+ },
131
+ {
132
+ "epoch": 1.6,
133
+ "grad_norm": 0.0016485441010445356,
134
+ "learning_rate": 4.001777777777778e-06,
135
+ "loss": 9.064872562885284e-05,
136
+ "step": 9000
137
+ },
138
+ {
139
+ "epoch": 1.6888888888888889,
140
+ "grad_norm": 0.002174828900024295,
141
+ "learning_rate": 3.1128888888888893e-06,
142
+ "loss": 8.987343311309815e-05,
143
+ "step": 9500
144
+ },
145
+ {
146
+ "epoch": 1.7777777777777777,
147
+ "grad_norm": 0.002475686138495803,
148
+ "learning_rate": 2.2240000000000002e-06,
149
+ "loss": 8.707723021507264e-05,
150
+ "step": 10000
151
+ },
152
+ {
153
+ "epoch": 1.8666666666666667,
154
+ "grad_norm": 0.0028533614240586758,
155
+ "learning_rate": 1.3351111111111111e-06,
156
+ "loss": 8.645709604024888e-05,
157
+ "step": 10500
158
+ },
159
+ {
160
+ "epoch": 1.9555555555555557,
161
+ "grad_norm": 0.001382834860123694,
162
+ "learning_rate": 4.4622222222222227e-07,
163
+ "loss": 8.54133814573288e-05,
164
+ "step": 11000
165
+ }
166
+ ],
167
+ "logging_steps": 500,
168
+ "max_steps": 11250,
169
+ "num_input_tokens_seen": 0,
170
+ "num_train_epochs": 2,
171
+ "save_steps": 500,
172
+ "stateful_callbacks": {
173
+ "TrainerControl": {
174
+ "args": {
175
+ "should_epoch_stop": false,
176
+ "should_evaluate": false,
177
+ "should_log": false,
178
+ "should_save": true,
179
+ "should_training_stop": false
180
+ },
181
+ "attributes": {}
182
+ }
183
+ },
184
+ "total_flos": 9.296524256811878e+16,
185
+ "train_batch_size": 32,
186
+ "trial_name": null,
187
+ "trial_params": null
188
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ed75abcfe811d82fa424040c0ba3f7056bfdd8999b142c56688e699c57934f8
3
+ size 5201
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/config.json ADDED
@@ -0,0 +1,155 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaAdvancedSpanClassifier"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": null,
7
+ "dtype": "float32",
8
+ "eos_token_id": null,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "id2label": {
13
+ "0": "O",
14
+ "1": "ACCOUNT_NUMBER",
15
+ "2": "AGE",
16
+ "3": "API_KEY",
17
+ "4": "BANK_ROUTING_NUMBER",
18
+ "5": "BIOMETRIC_IDENTIFIER",
19
+ "6": "BLOOD_TYPE",
20
+ "7": "CERTIFICATE_LICENSE_NUMBER",
21
+ "8": "CITY",
22
+ "9": "COMPANY_NAME",
23
+ "10": "COORDINATE",
24
+ "11": "COUNTRY",
25
+ "12": "COUNTY",
26
+ "13": "CREDIT_DEBIT_CARD",
27
+ "14": "CUSTOMER_ID",
28
+ "15": "CVV",
29
+ "16": "DATE",
30
+ "17": "DATE_OF_BIRTH",
31
+ "18": "DATE_TIME",
32
+ "19": "DEVICE_IDENTIFIER",
33
+ "20": "EDUCATION_LEVEL",
34
+ "21": "EMAIL",
35
+ "22": "EMPLOYEE_ID",
36
+ "23": "EMPLOYMENT_STATUS",
37
+ "24": "FAX_NUMBER",
38
+ "25": "FIRST_NAME",
39
+ "26": "GENDER",
40
+ "27": "HEALTH_PLAN_BENEFICIARY_NUMBER",
41
+ "28": "HTTP_COOKIE",
42
+ "29": "IPV4",
43
+ "30": "IPV6",
44
+ "31": "LANGUAGE",
45
+ "32": "LAST_NAME",
46
+ "33": "LICENSE_PLATE",
47
+ "34": "MAC_ADDRESS",
48
+ "35": "MEDICAL_RECORD_NUMBER",
49
+ "36": "NATIONAL_ID",
50
+ "37": "OCCUPATION",
51
+ "38": "PASSWORD",
52
+ "39": "PHONE_NUMBER",
53
+ "40": "PIN",
54
+ "41": "POLITICAL_VIEW",
55
+ "42": "POSTCODE",
56
+ "43": "RACE_ETHNICITY",
57
+ "44": "RELIGIOUS_BELIEF",
58
+ "45": "SEXUALITY",
59
+ "46": "SSN",
60
+ "47": "STATE",
61
+ "48": "STREET_ADDRESS",
62
+ "49": "SWIFT_BIC",
63
+ "50": "TAX_ID",
64
+ "51": "TIME",
65
+ "52": "UNIQUE_ID",
66
+ "53": "URL",
67
+ "54": "USER_NAME",
68
+ "55": "VEHICLE_IDENTIFIER"
69
+ },
70
+ "initializer_range": 0.02,
71
+ "intermediate_size": 3072,
72
+ "label2id": {
73
+ "ACCOUNT_NUMBER": 1,
74
+ "AGE": 2,
75
+ "API_KEY": 3,
76
+ "BANK_ROUTING_NUMBER": 4,
77
+ "BIOMETRIC_IDENTIFIER": 5,
78
+ "BLOOD_TYPE": 6,
79
+ "CERTIFICATE_LICENSE_NUMBER": 7,
80
+ "CITY": 8,
81
+ "COMPANY_NAME": 9,
82
+ "COORDINATE": 10,
83
+ "COUNTRY": 11,
84
+ "COUNTY": 12,
85
+ "CREDIT_DEBIT_CARD": 13,
86
+ "CUSTOMER_ID": 14,
87
+ "CVV": 15,
88
+ "DATE": 16,
89
+ "DATE_OF_BIRTH": 17,
90
+ "DATE_TIME": 18,
91
+ "DEVICE_IDENTIFIER": 19,
92
+ "EDUCATION_LEVEL": 20,
93
+ "EMAIL": 21,
94
+ "EMPLOYEE_ID": 22,
95
+ "EMPLOYMENT_STATUS": 23,
96
+ "FAX_NUMBER": 24,
97
+ "FIRST_NAME": 25,
98
+ "GENDER": 26,
99
+ "HEALTH_PLAN_BENEFICIARY_NUMBER": 27,
100
+ "HTTP_COOKIE": 28,
101
+ "IPV4": 29,
102
+ "IPV6": 30,
103
+ "LANGUAGE": 31,
104
+ "LAST_NAME": 32,
105
+ "LICENSE_PLATE": 33,
106
+ "MAC_ADDRESS": 34,
107
+ "MEDICAL_RECORD_NUMBER": 35,
108
+ "NATIONAL_ID": 36,
109
+ "O": 0,
110
+ "OCCUPATION": 37,
111
+ "PASSWORD": 38,
112
+ "PHONE_NUMBER": 39,
113
+ "PIN": 40,
114
+ "POLITICAL_VIEW": 41,
115
+ "POSTCODE": 42,
116
+ "RACE_ETHNICITY": 43,
117
+ "RELIGIOUS_BELIEF": 44,
118
+ "SEXUALITY": 45,
119
+ "SSN": 46,
120
+ "STATE": 47,
121
+ "STREET_ADDRESS": 48,
122
+ "SWIFT_BIC": 49,
123
+ "TAX_ID": 50,
124
+ "TIME": 51,
125
+ "UNIQUE_ID": 52,
126
+ "URL": 53,
127
+ "USER_NAME": 54,
128
+ "VEHICLE_IDENTIFIER": 55
129
+ },
130
+ "layer_norm_eps": 1e-07,
131
+ "legacy": true,
132
+ "max_position_embeddings": 512,
133
+ "max_relative_positions": -1,
134
+ "model_type": "deberta-v2",
135
+ "norm_rel_ebd": "layer_norm",
136
+ "num_attention_heads": 12,
137
+ "num_hidden_layers": 12,
138
+ "pad_token_id": 0,
139
+ "pooler_dropout": 0,
140
+ "pooler_hidden_act": "gelu",
141
+ "pooler_hidden_size": 768,
142
+ "pos_att_type": [
143
+ "p2c",
144
+ "c2p"
145
+ ],
146
+ "position_biased_input": false,
147
+ "position_buckets": 256,
148
+ "relative_attention": true,
149
+ "share_att_key": true,
150
+ "tie_word_embeddings": true,
151
+ "transformers_version": "5.3.0",
152
+ "type_vocab_size": 0,
153
+ "use_cache": false,
154
+ "vocab_size": 128100
155
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cdfcf4a278f75a1a79b1c8e607634e77844f908549f0f28e8ffa9c5d2c937ba0
3
+ size 741501240
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:47af5c7bc12c1cc9b5cda3723b2fc507804a2535b987559ae841f1c57a7ded3b
3
+ size 1483126155
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2314055b5798cffc71125c3891e7725d2df8a06c7170d4905e311cd1360fc7d4
3
+ size 14645
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:656f5bb14aef18ab8de277c99a2bd4b73386ca1c93f560a5a2f465573a80245d
3
+ size 1465
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/tokenizer_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "backend": "tokenizers",
4
+ "bos_token": "[CLS]",
5
+ "cls_token": "[CLS]",
6
+ "do_lower_case": false,
7
+ "eos_token": "[SEP]",
8
+ "extra_special_tokens": [
9
+ "[PAD]",
10
+ "[CLS]",
11
+ "[SEP]"
12
+ ],
13
+ "is_local": false,
14
+ "mask_token": "[MASK]",
15
+ "model_max_length": 1000000000000000019884624838656,
16
+ "pad_token": "[PAD]",
17
+ "sep_token": "[SEP]",
18
+ "split_by_punct": false,
19
+ "tokenizer_class": "DebertaV2Tokenizer",
20
+ "unk_id": 3,
21
+ "unk_token": "[UNK]",
22
+ "vocab_type": "spm"
23
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/trainer_state.json ADDED
@@ -0,0 +1,188 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 2.0,
6
+ "eval_steps": 500,
7
+ "global_step": 11250,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.08888888888888889,
14
+ "grad_norm": 0.003407392418012023,
15
+ "learning_rate": 1.911288888888889e-05,
16
+ "loss": 0.004852625846862793,
17
+ "step": 500
18
+ },
19
+ {
20
+ "epoch": 0.17777777777777778,
21
+ "grad_norm": 0.0030484353192150593,
22
+ "learning_rate": 1.8224e-05,
23
+ "loss": 0.0013473896980285645,
24
+ "step": 1000
25
+ },
26
+ {
27
+ "epoch": 0.26666666666666666,
28
+ "grad_norm": 0.0035167669411748648,
29
+ "learning_rate": 1.733511111111111e-05,
30
+ "loss": 0.0006659740805625916,
31
+ "step": 1500
32
+ },
33
+ {
34
+ "epoch": 0.35555555555555557,
35
+ "grad_norm": 0.0029440049547702074,
36
+ "learning_rate": 1.6446222222222223e-05,
37
+ "loss": 0.00036768344044685363,
38
+ "step": 2000
39
+ },
40
+ {
41
+ "epoch": 0.4444444444444444,
42
+ "grad_norm": 0.0034908612724393606,
43
+ "learning_rate": 1.5557333333333334e-05,
44
+ "loss": 0.00025459024310112,
45
+ "step": 2500
46
+ },
47
+ {
48
+ "epoch": 0.5333333333333333,
49
+ "grad_norm": 0.0030976070556789637,
50
+ "learning_rate": 1.4668444444444446e-05,
51
+ "loss": 0.00019871483743190765,
52
+ "step": 3000
53
+ },
54
+ {
55
+ "epoch": 0.6222222222222222,
56
+ "grad_norm": 0.0017464511329308152,
57
+ "learning_rate": 1.3779555555555557e-05,
58
+ "loss": 0.000175091952085495,
59
+ "step": 3500
60
+ },
61
+ {
62
+ "epoch": 0.7111111111111111,
63
+ "grad_norm": 0.003212156007066369,
64
+ "learning_rate": 1.2890666666666667e-05,
65
+ "loss": 0.00014939682185649873,
66
+ "step": 4000
67
+ },
68
+ {
69
+ "epoch": 0.8,
70
+ "grad_norm": 0.0026216867845505476,
71
+ "learning_rate": 1.2001777777777778e-05,
72
+ "loss": 0.00013630734384059907,
73
+ "step": 4500
74
+ },
75
+ {
76
+ "epoch": 0.8888888888888888,
77
+ "grad_norm": 0.0014595076208934188,
78
+ "learning_rate": 1.1112888888888891e-05,
79
+ "loss": 0.00013093747198581696,
80
+ "step": 5000
81
+ },
82
+ {
83
+ "epoch": 0.9777777777777777,
84
+ "grad_norm": 0.0013789460062980652,
85
+ "learning_rate": 1.0224e-05,
86
+ "loss": 0.0001222904920578003,
87
+ "step": 5500
88
+ },
89
+ {
90
+ "epoch": 1.0666666666666667,
91
+ "grad_norm": 0.0016316684195771813,
92
+ "learning_rate": 9.335111111111112e-06,
93
+ "loss": 0.00011386586725711822,
94
+ "step": 6000
95
+ },
96
+ {
97
+ "epoch": 1.1555555555555554,
98
+ "grad_norm": 0.0016652554040774703,
99
+ "learning_rate": 8.446222222222223e-06,
100
+ "loss": 0.00010670755803585053,
101
+ "step": 6500
102
+ },
103
+ {
104
+ "epoch": 1.2444444444444445,
105
+ "grad_norm": 0.0012231464497745037,
106
+ "learning_rate": 7.557333333333334e-06,
107
+ "loss": 0.00010442142933607102,
108
+ "step": 7000
109
+ },
110
+ {
111
+ "epoch": 1.3333333333333333,
112
+ "grad_norm": 0.002133500762283802,
113
+ "learning_rate": 6.668444444444445e-06,
114
+ "loss": 0.00010099935531616211,
115
+ "step": 7500
116
+ },
117
+ {
118
+ "epoch": 1.4222222222222223,
119
+ "grad_norm": 0.0011294716969132423,
120
+ "learning_rate": 5.779555555555556e-06,
121
+ "loss": 9.695450961589813e-05,
122
+ "step": 8000
123
+ },
124
+ {
125
+ "epoch": 1.511111111111111,
126
+ "grad_norm": 0.0015969782834872603,
127
+ "learning_rate": 4.890666666666667e-06,
128
+ "loss": 9.648169577121735e-05,
129
+ "step": 8500
130
+ },
131
+ {
132
+ "epoch": 1.6,
133
+ "grad_norm": 0.0016485441010445356,
134
+ "learning_rate": 4.001777777777778e-06,
135
+ "loss": 9.064872562885284e-05,
136
+ "step": 9000
137
+ },
138
+ {
139
+ "epoch": 1.6888888888888889,
140
+ "grad_norm": 0.002174828900024295,
141
+ "learning_rate": 3.1128888888888893e-06,
142
+ "loss": 8.987343311309815e-05,
143
+ "step": 9500
144
+ },
145
+ {
146
+ "epoch": 1.7777777777777777,
147
+ "grad_norm": 0.002475686138495803,
148
+ "learning_rate": 2.2240000000000002e-06,
149
+ "loss": 8.707723021507264e-05,
150
+ "step": 10000
151
+ },
152
+ {
153
+ "epoch": 1.8666666666666667,
154
+ "grad_norm": 0.0028533614240586758,
155
+ "learning_rate": 1.3351111111111111e-06,
156
+ "loss": 8.645709604024888e-05,
157
+ "step": 10500
158
+ },
159
+ {
160
+ "epoch": 1.9555555555555557,
161
+ "grad_norm": 0.001382834860123694,
162
+ "learning_rate": 4.4622222222222227e-07,
163
+ "loss": 8.54133814573288e-05,
164
+ "step": 11000
165
+ }
166
+ ],
167
+ "logging_steps": 500,
168
+ "max_steps": 11250,
169
+ "num_input_tokens_seen": 0,
170
+ "num_train_epochs": 2,
171
+ "save_steps": 500,
172
+ "stateful_callbacks": {
173
+ "TrainerControl": {
174
+ "args": {
175
+ "should_epoch_stop": false,
176
+ "should_evaluate": false,
177
+ "should_log": false,
178
+ "should_save": true,
179
+ "should_training_stop": true
180
+ },
181
+ "attributes": {}
182
+ }
183
+ },
184
+ "total_flos": 9.508566850798387e+16,
185
+ "train_batch_size": 32,
186
+ "trial_name": null,
187
+ "trial_params": null
188
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-11250/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ed75abcfe811d82fa424040c0ba3f7056bfdd8999b142c56688e699c57934f8
3
+ size 5201
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/config.json ADDED
@@ -0,0 +1,155 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaAdvancedSpanClassifier"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": null,
7
+ "dtype": "float32",
8
+ "eos_token_id": null,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "id2label": {
13
+ "0": "O",
14
+ "1": "ACCOUNT_NUMBER",
15
+ "2": "AGE",
16
+ "3": "API_KEY",
17
+ "4": "BANK_ROUTING_NUMBER",
18
+ "5": "BIOMETRIC_IDENTIFIER",
19
+ "6": "BLOOD_TYPE",
20
+ "7": "CERTIFICATE_LICENSE_NUMBER",
21
+ "8": "CITY",
22
+ "9": "COMPANY_NAME",
23
+ "10": "COORDINATE",
24
+ "11": "COUNTRY",
25
+ "12": "COUNTY",
26
+ "13": "CREDIT_DEBIT_CARD",
27
+ "14": "CUSTOMER_ID",
28
+ "15": "CVV",
29
+ "16": "DATE",
30
+ "17": "DATE_OF_BIRTH",
31
+ "18": "DATE_TIME",
32
+ "19": "DEVICE_IDENTIFIER",
33
+ "20": "EDUCATION_LEVEL",
34
+ "21": "EMAIL",
35
+ "22": "EMPLOYEE_ID",
36
+ "23": "EMPLOYMENT_STATUS",
37
+ "24": "FAX_NUMBER",
38
+ "25": "FIRST_NAME",
39
+ "26": "GENDER",
40
+ "27": "HEALTH_PLAN_BENEFICIARY_NUMBER",
41
+ "28": "HTTP_COOKIE",
42
+ "29": "IPV4",
43
+ "30": "IPV6",
44
+ "31": "LANGUAGE",
45
+ "32": "LAST_NAME",
46
+ "33": "LICENSE_PLATE",
47
+ "34": "MAC_ADDRESS",
48
+ "35": "MEDICAL_RECORD_NUMBER",
49
+ "36": "NATIONAL_ID",
50
+ "37": "OCCUPATION",
51
+ "38": "PASSWORD",
52
+ "39": "PHONE_NUMBER",
53
+ "40": "PIN",
54
+ "41": "POLITICAL_VIEW",
55
+ "42": "POSTCODE",
56
+ "43": "RACE_ETHNICITY",
57
+ "44": "RELIGIOUS_BELIEF",
58
+ "45": "SEXUALITY",
59
+ "46": "SSN",
60
+ "47": "STATE",
61
+ "48": "STREET_ADDRESS",
62
+ "49": "SWIFT_BIC",
63
+ "50": "TAX_ID",
64
+ "51": "TIME",
65
+ "52": "UNIQUE_ID",
66
+ "53": "URL",
67
+ "54": "USER_NAME",
68
+ "55": "VEHICLE_IDENTIFIER"
69
+ },
70
+ "initializer_range": 0.02,
71
+ "intermediate_size": 3072,
72
+ "label2id": {
73
+ "ACCOUNT_NUMBER": 1,
74
+ "AGE": 2,
75
+ "API_KEY": 3,
76
+ "BANK_ROUTING_NUMBER": 4,
77
+ "BIOMETRIC_IDENTIFIER": 5,
78
+ "BLOOD_TYPE": 6,
79
+ "CERTIFICATE_LICENSE_NUMBER": 7,
80
+ "CITY": 8,
81
+ "COMPANY_NAME": 9,
82
+ "COORDINATE": 10,
83
+ "COUNTRY": 11,
84
+ "COUNTY": 12,
85
+ "CREDIT_DEBIT_CARD": 13,
86
+ "CUSTOMER_ID": 14,
87
+ "CVV": 15,
88
+ "DATE": 16,
89
+ "DATE_OF_BIRTH": 17,
90
+ "DATE_TIME": 18,
91
+ "DEVICE_IDENTIFIER": 19,
92
+ "EDUCATION_LEVEL": 20,
93
+ "EMAIL": 21,
94
+ "EMPLOYEE_ID": 22,
95
+ "EMPLOYMENT_STATUS": 23,
96
+ "FAX_NUMBER": 24,
97
+ "FIRST_NAME": 25,
98
+ "GENDER": 26,
99
+ "HEALTH_PLAN_BENEFICIARY_NUMBER": 27,
100
+ "HTTP_COOKIE": 28,
101
+ "IPV4": 29,
102
+ "IPV6": 30,
103
+ "LANGUAGE": 31,
104
+ "LAST_NAME": 32,
105
+ "LICENSE_PLATE": 33,
106
+ "MAC_ADDRESS": 34,
107
+ "MEDICAL_RECORD_NUMBER": 35,
108
+ "NATIONAL_ID": 36,
109
+ "O": 0,
110
+ "OCCUPATION": 37,
111
+ "PASSWORD": 38,
112
+ "PHONE_NUMBER": 39,
113
+ "PIN": 40,
114
+ "POLITICAL_VIEW": 41,
115
+ "POSTCODE": 42,
116
+ "RACE_ETHNICITY": 43,
117
+ "RELIGIOUS_BELIEF": 44,
118
+ "SEXUALITY": 45,
119
+ "SSN": 46,
120
+ "STATE": 47,
121
+ "STREET_ADDRESS": 48,
122
+ "SWIFT_BIC": 49,
123
+ "TAX_ID": 50,
124
+ "TIME": 51,
125
+ "UNIQUE_ID": 52,
126
+ "URL": 53,
127
+ "USER_NAME": 54,
128
+ "VEHICLE_IDENTIFIER": 55
129
+ },
130
+ "layer_norm_eps": 1e-07,
131
+ "legacy": true,
132
+ "max_position_embeddings": 512,
133
+ "max_relative_positions": -1,
134
+ "model_type": "deberta-v2",
135
+ "norm_rel_ebd": "layer_norm",
136
+ "num_attention_heads": 12,
137
+ "num_hidden_layers": 12,
138
+ "pad_token_id": 0,
139
+ "pooler_dropout": 0,
140
+ "pooler_hidden_act": "gelu",
141
+ "pooler_hidden_size": 768,
142
+ "pos_att_type": [
143
+ "p2c",
144
+ "c2p"
145
+ ],
146
+ "position_biased_input": false,
147
+ "position_buckets": 256,
148
+ "relative_attention": true,
149
+ "share_att_key": true,
150
+ "tie_word_embeddings": true,
151
+ "transformers_version": "5.3.0",
152
+ "type_vocab_size": 0,
153
+ "use_cache": false,
154
+ "vocab_size": 128100
155
+ }
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac1ac6db5767fe7430d934d49bc6cc500d526c15b8853c7a087b381fba5a038d
3
+ size 741501240
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc06ecba421418d13cb7580b32a5bb160ffb8bf5b915d53b3bf8a63c8e0a379b
3
+ size 1483126155
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c5a4ff5bdb20a21502e0fc7fd177a60cf20307d1b1a02da3631468c6a3e79555
3
+ size 14645
nemotron-pii-ready-augmented/SpanBased-CustomDeBERTa/checkpoint-1500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:527176780910d5e7798bd16f788595042abe87cae37934b740541112f2182d09
3
+ size 1465