Hugging Face's logo

PuxAI
/
GenPII-framework

Safetensors
Model card Files
xet
 Community
PuxAI commited on
Commit
5f10349
·
verified ·
1 Parent(s): 03caa22

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. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/config.json +85 -0
  2. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/model.safetensors +3 -0
  3. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/optimizer.pt +3 -0
  4. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/rng_state.pth +3 -0
  5. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/scheduler.pt +3 -0
  6. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer.json +0 -0
  7. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer_config.json +23 -0
  8. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/trainer_state.json +48 -0
  9. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/training_args.bin +3 -0
  10. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/config.json +85 -0
  11. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/model.safetensors +3 -0
  12. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/optimizer.pt +3 -0
  13. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/rng_state.pth +3 -0
  14. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/scheduler.pt +3 -0
  15. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/tokenizer.json +0 -0
  16. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/tokenizer_config.json +23 -0
  17. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/trainer_state.json +174 -0
  18. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/training_args.bin +3 -0
  19. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/config.json +85 -0
  20. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/model.safetensors +3 -0
  21. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/optimizer.pt +3 -0
  22. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/rng_state.pth +3 -0
  23. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/scheduler.pt +3 -0
  24. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/tokenizer.json +0 -0
  25. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/tokenizer_config.json +23 -0
  26. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/trainer_state.json +181 -0
  27. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/training_args.bin +3 -0
  28. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/config.json +85 -0
  29. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/model.safetensors +3 -0
  30. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/optimizer.pt +3 -0
  31. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/rng_state.pth +3 -0
  32. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/scheduler.pt +3 -0
  33. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/tokenizer.json +0 -0
  34. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/tokenizer_config.json +23 -0
  35. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/trainer_state.json +188 -0
  36. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/training_args.bin +3 -0
  37. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/config.json +85 -0
  38. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/model.safetensors +3 -0
  39. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/optimizer.pt +3 -0
  40. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/rng_state.pth +3 -0
  41. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/scheduler.pt +3 -0
  42. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/tokenizer.json +0 -0
  43. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/tokenizer_config.json +23 -0
  44. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/trainer_state.json +195 -0
  45. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/training_args.bin +3 -0
  46. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-12000/config.json +85 -0
  47. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-12000/model.safetensors +3 -0
  48. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-12000/optimizer.pt +3 -0
  49. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-12000/rng_state.pth +3 -0
  50. open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-12000/scheduler.pt +3 -0
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/config.json ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "AGE",
15
+ "2": "BUILDINGNUM",
16
+ "3": "CITY",
17
+ "4": "CREDITCARDNUMBER",
18
+ "5": "DATE",
19
+ "6": "DRIVERLICENSENUM",
20
+ "7": "EMAIL",
21
+ "8": "GENDER",
22
+ "9": "GIVENNAME",
23
+ "10": "IDCARDNUM",
24
+ "11": "PASSPORTNUM",
25
+ "12": "SEX",
26
+ "13": "SOCIALNUM",
27
+ "14": "STREET",
28
+ "15": "SURNAME",
29
+ "16": "TAXNUM",
30
+ "17": "TELEPHONENUM",
31
+ "18": "TIME",
32
+ "19": "TITLE",
33
+ "20": "ZIPCODE"
34
+ },
35
+ "initializer_range": 0.02,
36
+ "intermediate_size": 3072,
37
+ "label2id": {
38
+ "AGE": 1,
39
+ "BUILDINGNUM": 2,
40
+ "CITY": 3,
41
+ "CREDITCARDNUMBER": 4,
42
+ "DATE": 5,
43
+ "DRIVERLICENSENUM": 6,
44
+ "EMAIL": 7,
45
+ "GENDER": 8,
46
+ "GIVENNAME": 9,
47
+ "IDCARDNUM": 10,
48
+ "O": 0,
49
+ "PASSPORTNUM": 11,
50
+ "SEX": 12,
51
+ "SOCIALNUM": 13,
52
+ "STREET": 14,
53
+ "SURNAME": 15,
54
+ "TAXNUM": 16,
55
+ "TELEPHONENUM": 17,
56
+ "TIME": 18,
57
+ "TITLE": 19,
58
+ "ZIPCODE": 20
59
+ },
60
+ "layer_norm_eps": 1e-07,
61
+ "legacy": true,
62
+ "max_position_embeddings": 512,
63
+ "max_relative_positions": -1,
64
+ "model_type": "deberta-v2",
65
+ "norm_rel_ebd": "layer_norm",
66
+ "num_attention_heads": 12,
67
+ "num_hidden_layers": 12,
68
+ "pad_token_id": 0,
69
+ "pooler_dropout": 0,
70
+ "pooler_hidden_act": "gelu",
71
+ "pooler_hidden_size": 768,
72
+ "pos_att_type": [
73
+ "p2c",
74
+ "c2p"
75
+ ],
76
+ "position_biased_input": false,
77
+ "position_buckets": 256,
78
+ "relative_attention": true,
79
+ "share_att_key": true,
80
+ "tie_word_embeddings": true,
81
+ "transformers_version": "5.3.0",
82
+ "type_vocab_size": 0,
83
+ "use_cache": false,
84
+ "vocab_size": 128100
85
+ }
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25bbee78c7b7f7d9e63dd9a2f27afeda35e9b8f933cfcbb8920155dc9a56559e
3
+ size 741393580
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1163d947def5c018329dfa659bb2e219d43d4fe2aeb51519fa23a7cba02eaa32
3
+ size 1482910859
open-pii-masking-500k-ai4privacy-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:b9aade56e3e84878bba6426e697a1e034760b0c2e4d19f96a3ee635656fc702d
3
+ size 14645
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb29b0b568a9ca311334725bc00f7b6737788038f422d136dbdc511567d29375
3
+ size 1465
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
open-pii-masking-500k-ai4privacy-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
+ }
open-pii-masking-500k-ai4privacy-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": 0.038300969014516066,
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": 0.019150484507258033,
14
+ "grad_norm": 0.034156788140535355,
15
+ "learning_rate": 1.9808878164617566e-05,
16
+ "loss": 0.00871822738647461,
17
+ "step": 500
18
+ },
19
+ {
20
+ "epoch": 0.038300969014516066,
21
+ "grad_norm": 0.025637200102210045,
22
+ "learning_rate": 1.9617373319544984e-05,
23
+ "loss": 0.0022422654628753663,
24
+ "step": 1000
25
+ }
26
+ ],
27
+ "logging_steps": 500,
28
+ "max_steps": 52218,
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": 1426039699124736.0,
45
+ "train_batch_size": 32,
46
+ "trial_name": null,
47
+ "trial_params": null
48
+ }
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-1000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ed9b96f34da7af5119d6d27027508ecb06ae553eb0a6df303658da8d30a7825
3
+ size 5265
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/config.json ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "AGE",
15
+ "2": "BUILDINGNUM",
16
+ "3": "CITY",
17
+ "4": "CREDITCARDNUMBER",
18
+ "5": "DATE",
19
+ "6": "DRIVERLICENSENUM",
20
+ "7": "EMAIL",
21
+ "8": "GENDER",
22
+ "9": "GIVENNAME",
23
+ "10": "IDCARDNUM",
24
+ "11": "PASSPORTNUM",
25
+ "12": "SEX",
26
+ "13": "SOCIALNUM",
27
+ "14": "STREET",
28
+ "15": "SURNAME",
29
+ "16": "TAXNUM",
30
+ "17": "TELEPHONENUM",
31
+ "18": "TIME",
32
+ "19": "TITLE",
33
+ "20": "ZIPCODE"
34
+ },
35
+ "initializer_range": 0.02,
36
+ "intermediate_size": 3072,
37
+ "label2id": {
38
+ "AGE": 1,
39
+ "BUILDINGNUM": 2,
40
+ "CITY": 3,
41
+ "CREDITCARDNUMBER": 4,
42
+ "DATE": 5,
43
+ "DRIVERLICENSENUM": 6,
44
+ "EMAIL": 7,
45
+ "GENDER": 8,
46
+ "GIVENNAME": 9,
47
+ "IDCARDNUM": 10,
48
+ "O": 0,
49
+ "PASSPORTNUM": 11,
50
+ "SEX": 12,
51
+ "SOCIALNUM": 13,
52
+ "STREET": 14,
53
+ "SURNAME": 15,
54
+ "TAXNUM": 16,
55
+ "TELEPHONENUM": 17,
56
+ "TIME": 18,
57
+ "TITLE": 19,
58
+ "ZIPCODE": 20
59
+ },
60
+ "layer_norm_eps": 1e-07,
61
+ "legacy": true,
62
+ "max_position_embeddings": 512,
63
+ "max_relative_positions": -1,
64
+ "model_type": "deberta-v2",
65
+ "norm_rel_ebd": "layer_norm",
66
+ "num_attention_heads": 12,
67
+ "num_hidden_layers": 12,
68
+ "pad_token_id": 0,
69
+ "pooler_dropout": 0,
70
+ "pooler_hidden_act": "gelu",
71
+ "pooler_hidden_size": 768,
72
+ "pos_att_type": [
73
+ "p2c",
74
+ "c2p"
75
+ ],
76
+ "position_biased_input": false,
77
+ "position_buckets": 256,
78
+ "relative_attention": true,
79
+ "share_att_key": true,
80
+ "tie_word_embeddings": true,
81
+ "transformers_version": "5.3.0",
82
+ "type_vocab_size": 0,
83
+ "use_cache": false,
84
+ "vocab_size": 128100
85
+ }
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fa29a73ea7dbab2a3ed5a3770953b6a00ea14057d0f777126fcab2c9bc3309a
3
+ size 741393580
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59eb7bcc9e794b48e983e40f67f2ed6d75d6200a9d68f8bfaa7ed9e66632d454
3
+ size 1482910859
open-pii-masking-500k-ai4privacy-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:eb643f902b3d4ee51966121a595e4a18fa0fa9dea4e99fa7692dc83d75302bfa
3
+ size 14645
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:404885bcb576a7205a6087373645a3d183a0a4053ece9f916eef4b525203e3f4
3
+ size 1465
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
open-pii-masking-500k-ai4privacy-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
+ }
open-pii-masking-500k-ai4privacy-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": 0.3830096901451607,
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": 0.019150484507258033,
14
+ "grad_norm": 0.034156788140535355,
15
+ "learning_rate": 1.9808878164617566e-05,
16
+ "loss": 0.00871822738647461,
17
+ "step": 500
18
+ },
19
+ {
20
+ "epoch": 0.038300969014516066,
21
+ "grad_norm": 0.025637200102210045,
22
+ "learning_rate": 1.9617373319544984e-05,
23
+ "loss": 0.0022422654628753663,
24
+ "step": 1000
25
+ },
26
+ {
27
+ "epoch": 0.0574514535217741,
28
+ "grad_norm": 0.014048685319721699,
29
+ "learning_rate": 1.9425868474472406e-05,
30
+ "loss": 0.001473078727722168,
31
+ "step": 1500
32
+ },
33
+ {
34
+ "epoch": 0.07660193802903213,
35
+ "grad_norm": 0.01930215023458004,
36
+ "learning_rate": 1.9234363629399828e-05,
37
+ "loss": 0.0011630326509475708,
38
+ "step": 2000
39
+ },
40
+ {
41
+ "epoch": 0.09575242253629017,
42
+ "grad_norm": 0.009974177926778793,
43
+ "learning_rate": 1.9042858784327246e-05,
44
+ "loss": 0.0010202296972274781,
45
+ "step": 2500
46
+ },
47
+ {
48
+ "epoch": 0.1149029070435482,
49
+ "grad_norm": 0.02064693532884121,
50
+ "learning_rate": 1.8851353939254664e-05,
51
+ "loss": 0.0009012192487716675,
52
+ "step": 3000
53
+ },
54
+ {
55
+ "epoch": 0.13405339155080623,
56
+ "grad_norm": 0.01888507604598999,
57
+ "learning_rate": 1.8659849094182086e-05,
58
+ "loss": 0.0008294718861579895,
59
+ "step": 3500
60
+ },
61
+ {
62
+ "epoch": 0.15320387605806426,
63
+ "grad_norm": 0.009450356476008892,
64
+ "learning_rate": 1.8468344249109504e-05,
65
+ "loss": 0.0007994875311851501,
66
+ "step": 4000
67
+ },
68
+ {
69
+ "epoch": 0.1723543605653223,
70
+ "grad_norm": 0.009589815512299538,
71
+ "learning_rate": 1.8276839404036922e-05,
72
+ "loss": 0.0007291412353515624,
73
+ "step": 4500
74
+ },
75
+ {
76
+ "epoch": 0.19150484507258034,
77
+ "grad_norm": 0.011395294219255447,
78
+ "learning_rate": 1.8085334558964344e-05,
79
+ "loss": 0.0006778205633163452,
80
+ "step": 5000
81
+ },
82
+ {
83
+ "epoch": 0.21065532957983837,
84
+ "grad_norm": 0.016027547419071198,
85
+ "learning_rate": 1.7893829713891762e-05,
86
+ "loss": 0.0006764695644378662,
87
+ "step": 5500
88
+ },
89
+ {
90
+ "epoch": 0.2298058140870964,
91
+ "grad_norm": 0.013933046720921993,
92
+ "learning_rate": 1.7702324868819184e-05,
93
+ "loss": 0.0006260576844215393,
94
+ "step": 6000
95
+ },
96
+ {
97
+ "epoch": 0.24895629859435445,
98
+ "grad_norm": 0.016653377562761307,
99
+ "learning_rate": 1.7510820023746602e-05,
100
+ "loss": 0.0006070020198822022,
101
+ "step": 6500
102
+ },
103
+ {
104
+ "epoch": 0.26810678310161246,
105
+ "grad_norm": 0.014872554689645767,
106
+ "learning_rate": 1.731931517867402e-05,
107
+ "loss": 0.0005897992253303528,
108
+ "step": 7000
109
+ },
110
+ {
111
+ "epoch": 0.2872572676088705,
112
+ "grad_norm": 0.014473496936261654,
113
+ "learning_rate": 1.7127810333601442e-05,
114
+ "loss": 0.0005676191449165344,
115
+ "step": 7500
116
+ },
117
+ {
118
+ "epoch": 0.30640775211612853,
119
+ "grad_norm": 0.015456922352313995,
120
+ "learning_rate": 1.693630548852886e-05,
121
+ "loss": 0.0005725777745246887,
122
+ "step": 8000
123
+ },
124
+ {
125
+ "epoch": 0.32555823662338657,
126
+ "grad_norm": 0.004569775890558958,
127
+ "learning_rate": 1.674480064345628e-05,
128
+ "loss": 0.0005602359175682068,
129
+ "step": 8500
130
+ },
131
+ {
132
+ "epoch": 0.3447087211306446,
133
+ "grad_norm": 0.008353209123015404,
134
+ "learning_rate": 1.65532957983837e-05,
135
+ "loss": 0.0005334309339523316,
136
+ "step": 9000
137
+ },
138
+ {
139
+ "epoch": 0.36385920563790264,
140
+ "grad_norm": 0.01133146695792675,
141
+ "learning_rate": 1.6361790953311122e-05,
142
+ "loss": 0.0005262523889541626,
143
+ "step": 9500
144
+ },
145
+ {
146
+ "epoch": 0.3830096901451607,
147
+ "grad_norm": 0.0170753076672554,
148
+ "learning_rate": 1.617028610823854e-05,
149
+ "loss": 0.00048073449730873106,
150
+ "step": 10000
151
+ }
152
+ ],
153
+ "logging_steps": 500,
154
+ "max_steps": 52218,
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": 1.4178273748576128e+16,
171
+ "train_batch_size": 32,
172
+ "trial_name": null,
173
+ "trial_params": null
174
+ }
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ed9b96f34da7af5119d6d27027508ecb06ae553eb0a6df303658da8d30a7825
3
+ size 5265
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/config.json ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "AGE",
15
+ "2": "BUILDINGNUM",
16
+ "3": "CITY",
17
+ "4": "CREDITCARDNUMBER",
18
+ "5": "DATE",
19
+ "6": "DRIVERLICENSENUM",
20
+ "7": "EMAIL",
21
+ "8": "GENDER",
22
+ "9": "GIVENNAME",
23
+ "10": "IDCARDNUM",
24
+ "11": "PASSPORTNUM",
25
+ "12": "SEX",
26
+ "13": "SOCIALNUM",
27
+ "14": "STREET",
28
+ "15": "SURNAME",
29
+ "16": "TAXNUM",
30
+ "17": "TELEPHONENUM",
31
+ "18": "TIME",
32
+ "19": "TITLE",
33
+ "20": "ZIPCODE"
34
+ },
35
+ "initializer_range": 0.02,
36
+ "intermediate_size": 3072,
37
+ "label2id": {
38
+ "AGE": 1,
39
+ "BUILDINGNUM": 2,
40
+ "CITY": 3,
41
+ "CREDITCARDNUMBER": 4,
42
+ "DATE": 5,
43
+ "DRIVERLICENSENUM": 6,
44
+ "EMAIL": 7,
45
+ "GENDER": 8,
46
+ "GIVENNAME": 9,
47
+ "IDCARDNUM": 10,
48
+ "O": 0,
49
+ "PASSPORTNUM": 11,
50
+ "SEX": 12,
51
+ "SOCIALNUM": 13,
52
+ "STREET": 14,
53
+ "SURNAME": 15,
54
+ "TAXNUM": 16,
55
+ "TELEPHONENUM": 17,
56
+ "TIME": 18,
57
+ "TITLE": 19,
58
+ "ZIPCODE": 20
59
+ },
60
+ "layer_norm_eps": 1e-07,
61
+ "legacy": true,
62
+ "max_position_embeddings": 512,
63
+ "max_relative_positions": -1,
64
+ "model_type": "deberta-v2",
65
+ "norm_rel_ebd": "layer_norm",
66
+ "num_attention_heads": 12,
67
+ "num_hidden_layers": 12,
68
+ "pad_token_id": 0,
69
+ "pooler_dropout": 0,
70
+ "pooler_hidden_act": "gelu",
71
+ "pooler_hidden_size": 768,
72
+ "pos_att_type": [
73
+ "p2c",
74
+ "c2p"
75
+ ],
76
+ "position_biased_input": false,
77
+ "position_buckets": 256,
78
+ "relative_attention": true,
79
+ "share_att_key": true,
80
+ "tie_word_embeddings": true,
81
+ "transformers_version": "5.3.0",
82
+ "type_vocab_size": 0,
83
+ "use_cache": false,
84
+ "vocab_size": 128100
85
+ }
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:55da4bb1edbf3104f66ae98ffe6e3b451c2b69fd549ced94fee13fa00866c536
3
+ size 741393580
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:13bd6780bdc0bd8f4057a89d018ea2c9d45c10b7af8bae2129a3ba0a0a193554
3
+ size 1482910859
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:71152e2006c765d6467c0ab3ff96dd2a9fca87eb793e96912046835fb489b5aa
3
+ size 14645
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:43bf1713b9af2a162bd902e1ea73b0d1d57106ba34d82792a57b772c83aecd10
3
+ size 1465
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
open-pii-masking-500k-ai4privacy-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
+ }
open-pii-masking-500k-ai4privacy-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": 0.4021601746524187,
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": 0.019150484507258033,
14
+ "grad_norm": 0.034156788140535355,
15
+ "learning_rate": 1.9808878164617566e-05,
16
+ "loss": 0.00871822738647461,
17
+ "step": 500
18
+ },
19
+ {
20
+ "epoch": 0.038300969014516066,
21
+ "grad_norm": 0.025637200102210045,
22
+ "learning_rate": 1.9617373319544984e-05,
23
+ "loss": 0.0022422654628753663,
24
+ "step": 1000
25
+ },
26
+ {
27
+ "epoch": 0.0574514535217741,
28
+ "grad_norm": 0.014048685319721699,
29
+ "learning_rate": 1.9425868474472406e-05,
30
+ "loss": 0.001473078727722168,
31
+ "step": 1500
32
+ },
33
+ {
34
+ "epoch": 0.07660193802903213,
35
+ "grad_norm": 0.01930215023458004,
36
+ "learning_rate": 1.9234363629399828e-05,
37
+ "loss": 0.0011630326509475708,
38
+ "step": 2000
39
+ },
40
+ {
41
+ "epoch": 0.09575242253629017,
42
+ "grad_norm": 0.009974177926778793,
43
+ "learning_rate": 1.9042858784327246e-05,
44
+ "loss": 0.0010202296972274781,
45
+ "step": 2500
46
+ },
47
+ {
48
+ "epoch": 0.1149029070435482,
49
+ "grad_norm": 0.02064693532884121,
50
+ "learning_rate": 1.8851353939254664e-05,
51
+ "loss": 0.0009012192487716675,
52
+ "step": 3000
53
+ },
54
+ {
55
+ "epoch": 0.13405339155080623,
56
+ "grad_norm": 0.01888507604598999,
57
+ "learning_rate": 1.8659849094182086e-05,
58
+ "loss": 0.0008294718861579895,
59
+ "step": 3500
60
+ },
61
+ {
62
+ "epoch": 0.15320387605806426,
63
+ "grad_norm": 0.009450356476008892,
64
+ "learning_rate": 1.8468344249109504e-05,
65
+ "loss": 0.0007994875311851501,
66
+ "step": 4000
67
+ },
68
+ {
69
+ "epoch": 0.1723543605653223,
70
+ "grad_norm": 0.009589815512299538,
71
+ "learning_rate": 1.8276839404036922e-05,
72
+ "loss": 0.0007291412353515624,
73
+ "step": 4500
74
+ },
75
+ {
76
+ "epoch": 0.19150484507258034,
77
+ "grad_norm": 0.011395294219255447,
78
+ "learning_rate": 1.8085334558964344e-05,
79
+ "loss": 0.0006778205633163452,
80
+ "step": 5000
81
+ },
82
+ {
83
+ "epoch": 0.21065532957983837,
84
+ "grad_norm": 0.016027547419071198,
85
+ "learning_rate": 1.7893829713891762e-05,
86
+ "loss": 0.0006764695644378662,
87
+ "step": 5500
88
+ },
89
+ {
90
+ "epoch": 0.2298058140870964,
91
+ "grad_norm": 0.013933046720921993,
92
+ "learning_rate": 1.7702324868819184e-05,
93
+ "loss": 0.0006260576844215393,
94
+ "step": 6000
95
+ },
96
+ {
97
+ "epoch": 0.24895629859435445,
98
+ "grad_norm": 0.016653377562761307,
99
+ "learning_rate": 1.7510820023746602e-05,
100
+ "loss": 0.0006070020198822022,
101
+ "step": 6500
102
+ },
103
+ {
104
+ "epoch": 0.26810678310161246,
105
+ "grad_norm": 0.014872554689645767,
106
+ "learning_rate": 1.731931517867402e-05,
107
+ "loss": 0.0005897992253303528,
108
+ "step": 7000
109
+ },
110
+ {
111
+ "epoch": 0.2872572676088705,
112
+ "grad_norm": 0.014473496936261654,
113
+ "learning_rate": 1.7127810333601442e-05,
114
+ "loss": 0.0005676191449165344,
115
+ "step": 7500
116
+ },
117
+ {
118
+ "epoch": 0.30640775211612853,
119
+ "grad_norm": 0.015456922352313995,
120
+ "learning_rate": 1.693630548852886e-05,
121
+ "loss": 0.0005725777745246887,
122
+ "step": 8000
123
+ },
124
+ {
125
+ "epoch": 0.32555823662338657,
126
+ "grad_norm": 0.004569775890558958,
127
+ "learning_rate": 1.674480064345628e-05,
128
+ "loss": 0.0005602359175682068,
129
+ "step": 8500
130
+ },
131
+ {
132
+ "epoch": 0.3447087211306446,
133
+ "grad_norm": 0.008353209123015404,
134
+ "learning_rate": 1.65532957983837e-05,
135
+ "loss": 0.0005334309339523316,
136
+ "step": 9000
137
+ },
138
+ {
139
+ "epoch": 0.36385920563790264,
140
+ "grad_norm": 0.01133146695792675,
141
+ "learning_rate": 1.6361790953311122e-05,
142
+ "loss": 0.0005262523889541626,
143
+ "step": 9500
144
+ },
145
+ {
146
+ "epoch": 0.3830096901451607,
147
+ "grad_norm": 0.0170753076672554,
148
+ "learning_rate": 1.617028610823854e-05,
149
+ "loss": 0.00048073449730873106,
150
+ "step": 10000
151
+ },
152
+ {
153
+ "epoch": 0.4021601746524187,
154
+ "grad_norm": 0.010415528900921345,
155
+ "learning_rate": 1.597878126316596e-05,
156
+ "loss": 0.0005041568279266357,
157
+ "step": 10500
158
+ }
159
+ ],
160
+ "logging_steps": 500,
161
+ "max_steps": 52218,
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": 1.487316400328736e+16,
178
+ "train_batch_size": 32,
179
+ "trial_name": null,
180
+ "trial_params": null
181
+ }
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-10500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ed9b96f34da7af5119d6d27027508ecb06ae553eb0a6df303658da8d30a7825
3
+ size 5265
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/config.json ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "AGE",
15
+ "2": "BUILDINGNUM",
16
+ "3": "CITY",
17
+ "4": "CREDITCARDNUMBER",
18
+ "5": "DATE",
19
+ "6": "DRIVERLICENSENUM",
20
+ "7": "EMAIL",
21
+ "8": "GENDER",
22
+ "9": "GIVENNAME",
23
+ "10": "IDCARDNUM",
24
+ "11": "PASSPORTNUM",
25
+ "12": "SEX",
26
+ "13": "SOCIALNUM",
27
+ "14": "STREET",
28
+ "15": "SURNAME",
29
+ "16": "TAXNUM",
30
+ "17": "TELEPHONENUM",
31
+ "18": "TIME",
32
+ "19": "TITLE",
33
+ "20": "ZIPCODE"
34
+ },
35
+ "initializer_range": 0.02,
36
+ "intermediate_size": 3072,
37
+ "label2id": {
38
+ "AGE": 1,
39
+ "BUILDINGNUM": 2,
40
+ "CITY": 3,
41
+ "CREDITCARDNUMBER": 4,
42
+ "DATE": 5,
43
+ "DRIVERLICENSENUM": 6,
44
+ "EMAIL": 7,
45
+ "GENDER": 8,
46
+ "GIVENNAME": 9,
47
+ "IDCARDNUM": 10,
48
+ "O": 0,
49
+ "PASSPORTNUM": 11,
50
+ "SEX": 12,
51
+ "SOCIALNUM": 13,
52
+ "STREET": 14,
53
+ "SURNAME": 15,
54
+ "TAXNUM": 16,
55
+ "TELEPHONENUM": 17,
56
+ "TIME": 18,
57
+ "TITLE": 19,
58
+ "ZIPCODE": 20
59
+ },
60
+ "layer_norm_eps": 1e-07,
61
+ "legacy": true,
62
+ "max_position_embeddings": 512,
63
+ "max_relative_positions": -1,
64
+ "model_type": "deberta-v2",
65
+ "norm_rel_ebd": "layer_norm",
66
+ "num_attention_heads": 12,
67
+ "num_hidden_layers": 12,
68
+ "pad_token_id": 0,
69
+ "pooler_dropout": 0,
70
+ "pooler_hidden_act": "gelu",
71
+ "pooler_hidden_size": 768,
72
+ "pos_att_type": [
73
+ "p2c",
74
+ "c2p"
75
+ ],
76
+ "position_biased_input": false,
77
+ "position_buckets": 256,
78
+ "relative_attention": true,
79
+ "share_att_key": true,
80
+ "tie_word_embeddings": true,
81
+ "transformers_version": "5.3.0",
82
+ "type_vocab_size": 0,
83
+ "use_cache": false,
84
+ "vocab_size": 128100
85
+ }
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5cc26b9d5e012ab532ecf2ac719f9bf57a3cbef5ac752b51d6f2ab96261d9031
3
+ size 741393580
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b394c6cf05f8c84b8115a19325f85c53849dd6e89d1755ecded313918e4b7371
3
+ size 1482910859
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3dd581634902db729c297fbfbce04ab8864bbaafe84449835bc8d95e0de163ef
3
+ size 14645
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:baf211d6006dfdc25bf07dc4c8f637f3813b96cfc98bd2d705a5d57417aa1dc8
3
+ size 1465
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
open-pii-masking-500k-ai4privacy-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
+ }
open-pii-masking-500k-ai4privacy-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": 0.42131065915967675,
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": 0.019150484507258033,
14
+ "grad_norm": 0.034156788140535355,
15
+ "learning_rate": 1.9808878164617566e-05,
16
+ "loss": 0.00871822738647461,
17
+ "step": 500
18
+ },
19
+ {
20
+ "epoch": 0.038300969014516066,
21
+ "grad_norm": 0.025637200102210045,
22
+ "learning_rate": 1.9617373319544984e-05,
23
+ "loss": 0.0022422654628753663,
24
+ "step": 1000
25
+ },
26
+ {
27
+ "epoch": 0.0574514535217741,
28
+ "grad_norm": 0.014048685319721699,
29
+ "learning_rate": 1.9425868474472406e-05,
30
+ "loss": 0.001473078727722168,
31
+ "step": 1500
32
+ },
33
+ {
34
+ "epoch": 0.07660193802903213,
35
+ "grad_norm": 0.01930215023458004,
36
+ "learning_rate": 1.9234363629399828e-05,
37
+ "loss": 0.0011630326509475708,
38
+ "step": 2000
39
+ },
40
+ {
41
+ "epoch": 0.09575242253629017,
42
+ "grad_norm": 0.009974177926778793,
43
+ "learning_rate": 1.9042858784327246e-05,
44
+ "loss": 0.0010202296972274781,
45
+ "step": 2500
46
+ },
47
+ {
48
+ "epoch": 0.1149029070435482,
49
+ "grad_norm": 0.02064693532884121,
50
+ "learning_rate": 1.8851353939254664e-05,
51
+ "loss": 0.0009012192487716675,
52
+ "step": 3000
53
+ },
54
+ {
55
+ "epoch": 0.13405339155080623,
56
+ "grad_norm": 0.01888507604598999,
57
+ "learning_rate": 1.8659849094182086e-05,
58
+ "loss": 0.0008294718861579895,
59
+ "step": 3500
60
+ },
61
+ {
62
+ "epoch": 0.15320387605806426,
63
+ "grad_norm": 0.009450356476008892,
64
+ "learning_rate": 1.8468344249109504e-05,
65
+ "loss": 0.0007994875311851501,
66
+ "step": 4000
67
+ },
68
+ {
69
+ "epoch": 0.1723543605653223,
70
+ "grad_norm": 0.009589815512299538,
71
+ "learning_rate": 1.8276839404036922e-05,
72
+ "loss": 0.0007291412353515624,
73
+ "step": 4500
74
+ },
75
+ {
76
+ "epoch": 0.19150484507258034,
77
+ "grad_norm": 0.011395294219255447,
78
+ "learning_rate": 1.8085334558964344e-05,
79
+ "loss": 0.0006778205633163452,
80
+ "step": 5000
81
+ },
82
+ {
83
+ "epoch": 0.21065532957983837,
84
+ "grad_norm": 0.016027547419071198,
85
+ "learning_rate": 1.7893829713891762e-05,
86
+ "loss": 0.0006764695644378662,
87
+ "step": 5500
88
+ },
89
+ {
90
+ "epoch": 0.2298058140870964,
91
+ "grad_norm": 0.013933046720921993,
92
+ "learning_rate": 1.7702324868819184e-05,
93
+ "loss": 0.0006260576844215393,
94
+ "step": 6000
95
+ },
96
+ {
97
+ "epoch": 0.24895629859435445,
98
+ "grad_norm": 0.016653377562761307,
99
+ "learning_rate": 1.7510820023746602e-05,
100
+ "loss": 0.0006070020198822022,
101
+ "step": 6500
102
+ },
103
+ {
104
+ "epoch": 0.26810678310161246,
105
+ "grad_norm": 0.014872554689645767,
106
+ "learning_rate": 1.731931517867402e-05,
107
+ "loss": 0.0005897992253303528,
108
+ "step": 7000
109
+ },
110
+ {
111
+ "epoch": 0.2872572676088705,
112
+ "grad_norm": 0.014473496936261654,
113
+ "learning_rate": 1.7127810333601442e-05,
114
+ "loss": 0.0005676191449165344,
115
+ "step": 7500
116
+ },
117
+ {
118
+ "epoch": 0.30640775211612853,
119
+ "grad_norm": 0.015456922352313995,
120
+ "learning_rate": 1.693630548852886e-05,
121
+ "loss": 0.0005725777745246887,
122
+ "step": 8000
123
+ },
124
+ {
125
+ "epoch": 0.32555823662338657,
126
+ "grad_norm": 0.004569775890558958,
127
+ "learning_rate": 1.674480064345628e-05,
128
+ "loss": 0.0005602359175682068,
129
+ "step": 8500
130
+ },
131
+ {
132
+ "epoch": 0.3447087211306446,
133
+ "grad_norm": 0.008353209123015404,
134
+ "learning_rate": 1.65532957983837e-05,
135
+ "loss": 0.0005334309339523316,
136
+ "step": 9000
137
+ },
138
+ {
139
+ "epoch": 0.36385920563790264,
140
+ "grad_norm": 0.01133146695792675,
141
+ "learning_rate": 1.6361790953311122e-05,
142
+ "loss": 0.0005262523889541626,
143
+ "step": 9500
144
+ },
145
+ {
146
+ "epoch": 0.3830096901451607,
147
+ "grad_norm": 0.0170753076672554,
148
+ "learning_rate": 1.617028610823854e-05,
149
+ "loss": 0.00048073449730873106,
150
+ "step": 10000
151
+ },
152
+ {
153
+ "epoch": 0.4021601746524187,
154
+ "grad_norm": 0.010415528900921345,
155
+ "learning_rate": 1.597878126316596e-05,
156
+ "loss": 0.0005041568279266357,
157
+ "step": 10500
158
+ },
159
+ {
160
+ "epoch": 0.42131065915967675,
161
+ "grad_norm": 0.010244589298963547,
162
+ "learning_rate": 1.5787276418093377e-05,
163
+ "loss": 0.0004949500858783722,
164
+ "step": 11000
165
+ }
166
+ ],
167
+ "logging_steps": 500,
168
+ "max_steps": 52218,
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": 1.5568802041470912e+16,
185
+ "train_batch_size": 32,
186
+ "trial_name": null,
187
+ "trial_params": null
188
+ }
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ed9b96f34da7af5119d6d27027508ecb06ae553eb0a6df303658da8d30a7825
3
+ size 5265
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/config.json ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "AGE",
15
+ "2": "BUILDINGNUM",
16
+ "3": "CITY",
17
+ "4": "CREDITCARDNUMBER",
18
+ "5": "DATE",
19
+ "6": "DRIVERLICENSENUM",
20
+ "7": "EMAIL",
21
+ "8": "GENDER",
22
+ "9": "GIVENNAME",
23
+ "10": "IDCARDNUM",
24
+ "11": "PASSPORTNUM",
25
+ "12": "SEX",
26
+ "13": "SOCIALNUM",
27
+ "14": "STREET",
28
+ "15": "SURNAME",
29
+ "16": "TAXNUM",
30
+ "17": "TELEPHONENUM",
31
+ "18": "TIME",
32
+ "19": "TITLE",
33
+ "20": "ZIPCODE"
34
+ },
35
+ "initializer_range": 0.02,
36
+ "intermediate_size": 3072,
37
+ "label2id": {
38
+ "AGE": 1,
39
+ "BUILDINGNUM": 2,
40
+ "CITY": 3,
41
+ "CREDITCARDNUMBER": 4,
42
+ "DATE": 5,
43
+ "DRIVERLICENSENUM": 6,
44
+ "EMAIL": 7,
45
+ "GENDER": 8,
46
+ "GIVENNAME": 9,
47
+ "IDCARDNUM": 10,
48
+ "O": 0,
49
+ "PASSPORTNUM": 11,
50
+ "SEX": 12,
51
+ "SOCIALNUM": 13,
52
+ "STREET": 14,
53
+ "SURNAME": 15,
54
+ "TAXNUM": 16,
55
+ "TELEPHONENUM": 17,
56
+ "TIME": 18,
57
+ "TITLE": 19,
58
+ "ZIPCODE": 20
59
+ },
60
+ "layer_norm_eps": 1e-07,
61
+ "legacy": true,
62
+ "max_position_embeddings": 512,
63
+ "max_relative_positions": -1,
64
+ "model_type": "deberta-v2",
65
+ "norm_rel_ebd": "layer_norm",
66
+ "num_attention_heads": 12,
67
+ "num_hidden_layers": 12,
68
+ "pad_token_id": 0,
69
+ "pooler_dropout": 0,
70
+ "pooler_hidden_act": "gelu",
71
+ "pooler_hidden_size": 768,
72
+ "pos_att_type": [
73
+ "p2c",
74
+ "c2p"
75
+ ],
76
+ "position_biased_input": false,
77
+ "position_buckets": 256,
78
+ "relative_attention": true,
79
+ "share_att_key": true,
80
+ "tie_word_embeddings": true,
81
+ "transformers_version": "5.3.0",
82
+ "type_vocab_size": 0,
83
+ "use_cache": false,
84
+ "vocab_size": 128100
85
+ }
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f33808c8054727deb25a1f947b56458c18028316e7cead8f4fa9afa27b8b8069
3
+ size 741393580
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:08367ff0dac373ff07e3c694949df674217552fd7586574c59d83adba8e45743
3
+ size 1482910859
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:18199688298cd4cbbd465b0f50865c7d4b83cec08ffd037c7f7bf686e6e0b024
3
+ size 14645
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd496d4feb817e56d069986796f959fdd146a8a41de32503343a9e499bdbe536
3
+ size 1465
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/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
+ }
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/trainer_state.json ADDED
@@ -0,0 +1,195 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.4404611436669348,
6
+ "eval_steps": 500,
7
+ "global_step": 11500,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.019150484507258033,
14
+ "grad_norm": 0.034156788140535355,
15
+ "learning_rate": 1.9808878164617566e-05,
16
+ "loss": 0.00871822738647461,
17
+ "step": 500
18
+ },
19
+ {
20
+ "epoch": 0.038300969014516066,
21
+ "grad_norm": 0.025637200102210045,
22
+ "learning_rate": 1.9617373319544984e-05,
23
+ "loss": 0.0022422654628753663,
24
+ "step": 1000
25
+ },
26
+ {
27
+ "epoch": 0.0574514535217741,
28
+ "grad_norm": 0.014048685319721699,
29
+ "learning_rate": 1.9425868474472406e-05,
30
+ "loss": 0.001473078727722168,
31
+ "step": 1500
32
+ },
33
+ {
34
+ "epoch": 0.07660193802903213,
35
+ "grad_norm": 0.01930215023458004,
36
+ "learning_rate": 1.9234363629399828e-05,
37
+ "loss": 0.0011630326509475708,
38
+ "step": 2000
39
+ },
40
+ {
41
+ "epoch": 0.09575242253629017,
42
+ "grad_norm": 0.009974177926778793,
43
+ "learning_rate": 1.9042858784327246e-05,
44
+ "loss": 0.0010202296972274781,
45
+ "step": 2500
46
+ },
47
+ {
48
+ "epoch": 0.1149029070435482,
49
+ "grad_norm": 0.02064693532884121,
50
+ "learning_rate": 1.8851353939254664e-05,
51
+ "loss": 0.0009012192487716675,
52
+ "step": 3000
53
+ },
54
+ {
55
+ "epoch": 0.13405339155080623,
56
+ "grad_norm": 0.01888507604598999,
57
+ "learning_rate": 1.8659849094182086e-05,
58
+ "loss": 0.0008294718861579895,
59
+ "step": 3500
60
+ },
61
+ {
62
+ "epoch": 0.15320387605806426,
63
+ "grad_norm": 0.009450356476008892,
64
+ "learning_rate": 1.8468344249109504e-05,
65
+ "loss": 0.0007994875311851501,
66
+ "step": 4000
67
+ },
68
+ {
69
+ "epoch": 0.1723543605653223,
70
+ "grad_norm": 0.009589815512299538,
71
+ "learning_rate": 1.8276839404036922e-05,
72
+ "loss": 0.0007291412353515624,
73
+ "step": 4500
74
+ },
75
+ {
76
+ "epoch": 0.19150484507258034,
77
+ "grad_norm": 0.011395294219255447,
78
+ "learning_rate": 1.8085334558964344e-05,
79
+ "loss": 0.0006778205633163452,
80
+ "step": 5000
81
+ },
82
+ {
83
+ "epoch": 0.21065532957983837,
84
+ "grad_norm": 0.016027547419071198,
85
+ "learning_rate": 1.7893829713891762e-05,
86
+ "loss": 0.0006764695644378662,
87
+ "step": 5500
88
+ },
89
+ {
90
+ "epoch": 0.2298058140870964,
91
+ "grad_norm": 0.013933046720921993,
92
+ "learning_rate": 1.7702324868819184e-05,
93
+ "loss": 0.0006260576844215393,
94
+ "step": 6000
95
+ },
96
+ {
97
+ "epoch": 0.24895629859435445,
98
+ "grad_norm": 0.016653377562761307,
99
+ "learning_rate": 1.7510820023746602e-05,
100
+ "loss": 0.0006070020198822022,
101
+ "step": 6500
102
+ },
103
+ {
104
+ "epoch": 0.26810678310161246,
105
+ "grad_norm": 0.014872554689645767,
106
+ "learning_rate": 1.731931517867402e-05,
107
+ "loss": 0.0005897992253303528,
108
+ "step": 7000
109
+ },
110
+ {
111
+ "epoch": 0.2872572676088705,
112
+ "grad_norm": 0.014473496936261654,
113
+ "learning_rate": 1.7127810333601442e-05,
114
+ "loss": 0.0005676191449165344,
115
+ "step": 7500
116
+ },
117
+ {
118
+ "epoch": 0.30640775211612853,
119
+ "grad_norm": 0.015456922352313995,
120
+ "learning_rate": 1.693630548852886e-05,
121
+ "loss": 0.0005725777745246887,
122
+ "step": 8000
123
+ },
124
+ {
125
+ "epoch": 0.32555823662338657,
126
+ "grad_norm": 0.004569775890558958,
127
+ "learning_rate": 1.674480064345628e-05,
128
+ "loss": 0.0005602359175682068,
129
+ "step": 8500
130
+ },
131
+ {
132
+ "epoch": 0.3447087211306446,
133
+ "grad_norm": 0.008353209123015404,
134
+ "learning_rate": 1.65532957983837e-05,
135
+ "loss": 0.0005334309339523316,
136
+ "step": 9000
137
+ },
138
+ {
139
+ "epoch": 0.36385920563790264,
140
+ "grad_norm": 0.01133146695792675,
141
+ "learning_rate": 1.6361790953311122e-05,
142
+ "loss": 0.0005262523889541626,
143
+ "step": 9500
144
+ },
145
+ {
146
+ "epoch": 0.3830096901451607,
147
+ "grad_norm": 0.0170753076672554,
148
+ "learning_rate": 1.617028610823854e-05,
149
+ "loss": 0.00048073449730873106,
150
+ "step": 10000
151
+ },
152
+ {
153
+ "epoch": 0.4021601746524187,
154
+ "grad_norm": 0.010415528900921345,
155
+ "learning_rate": 1.597878126316596e-05,
156
+ "loss": 0.0005041568279266357,
157
+ "step": 10500
158
+ },
159
+ {
160
+ "epoch": 0.42131065915967675,
161
+ "grad_norm": 0.010244589298963547,
162
+ "learning_rate": 1.5787276418093377e-05,
163
+ "loss": 0.0004949500858783722,
164
+ "step": 11000
165
+ },
166
+ {
167
+ "epoch": 0.4404611436669348,
168
+ "grad_norm": 0.007186885457485914,
169
+ "learning_rate": 1.55957715730208e-05,
170
+ "loss": 0.00047441959381103513,
171
+ "step": 11500
172
+ }
173
+ ],
174
+ "logging_steps": 500,
175
+ "max_steps": 52218,
176
+ "num_input_tokens_seen": 0,
177
+ "num_train_epochs": 2,
178
+ "save_steps": 500,
179
+ "stateful_callbacks": {
180
+ "TrainerControl": {
181
+ "args": {
182
+ "should_epoch_stop": false,
183
+ "should_evaluate": false,
184
+ "should_log": false,
185
+ "should_save": true,
186
+ "should_training_stop": false
187
+ },
188
+ "attributes": {}
189
+ }
190
+ },
191
+ "total_flos": 1.6259936761410048e+16,
192
+ "train_batch_size": 32,
193
+ "trial_name": null,
194
+ "trial_params": null
195
+ }
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-11500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ed9b96f34da7af5119d6d27027508ecb06ae553eb0a6df303658da8d30a7825
3
+ size 5265
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-12000/config.json ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "AGE",
15
+ "2": "BUILDINGNUM",
16
+ "3": "CITY",
17
+ "4": "CREDITCARDNUMBER",
18
+ "5": "DATE",
19
+ "6": "DRIVERLICENSENUM",
20
+ "7": "EMAIL",
21
+ "8": "GENDER",
22
+ "9": "GIVENNAME",
23
+ "10": "IDCARDNUM",
24
+ "11": "PASSPORTNUM",
25
+ "12": "SEX",
26
+ "13": "SOCIALNUM",
27
+ "14": "STREET",
28
+ "15": "SURNAME",
29
+ "16": "TAXNUM",
30
+ "17": "TELEPHONENUM",
31
+ "18": "TIME",
32
+ "19": "TITLE",
33
+ "20": "ZIPCODE"
34
+ },
35
+ "initializer_range": 0.02,
36
+ "intermediate_size": 3072,
37
+ "label2id": {
38
+ "AGE": 1,
39
+ "BUILDINGNUM": 2,
40
+ "CITY": 3,
41
+ "CREDITCARDNUMBER": 4,
42
+ "DATE": 5,
43
+ "DRIVERLICENSENUM": 6,
44
+ "EMAIL": 7,
45
+ "GENDER": 8,
46
+ "GIVENNAME": 9,
47
+ "IDCARDNUM": 10,
48
+ "O": 0,
49
+ "PASSPORTNUM": 11,
50
+ "SEX": 12,
51
+ "SOCIALNUM": 13,
52
+ "STREET": 14,
53
+ "SURNAME": 15,
54
+ "TAXNUM": 16,
55
+ "TELEPHONENUM": 17,
56
+ "TIME": 18,
57
+ "TITLE": 19,
58
+ "ZIPCODE": 20
59
+ },
60
+ "layer_norm_eps": 1e-07,
61
+ "legacy": true,
62
+ "max_position_embeddings": 512,
63
+ "max_relative_positions": -1,
64
+ "model_type": "deberta-v2",
65
+ "norm_rel_ebd": "layer_norm",
66
+ "num_attention_heads": 12,
67
+ "num_hidden_layers": 12,
68
+ "pad_token_id": 0,
69
+ "pooler_dropout": 0,
70
+ "pooler_hidden_act": "gelu",
71
+ "pooler_hidden_size": 768,
72
+ "pos_att_type": [
73
+ "p2c",
74
+ "c2p"
75
+ ],
76
+ "position_biased_input": false,
77
+ "position_buckets": 256,
78
+ "relative_attention": true,
79
+ "share_att_key": true,
80
+ "tie_word_embeddings": true,
81
+ "transformers_version": "5.3.0",
82
+ "type_vocab_size": 0,
83
+ "use_cache": false,
84
+ "vocab_size": 128100
85
+ }
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-12000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab42777dc1d7114925dbc1d95e873772f74cda042b72aa4157644d1dff24cc86
3
+ size 741393580
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-12000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e03696541732929c417f098e0f45843ab55bfceb150b6453c59292006bd21873
3
+ size 1482910859
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-12000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca0d6364e2b2162ad2b9810815971c34036fc8660ac9deb6820108caaf6fc9e7
3
+ size 14645
open-pii-masking-500k-ai4privacy-augmented/SpanBased-CustomDeBERTa/checkpoint-12000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fbda0ca3516cb8ca1a5dbc706a7f2816f0cd8776cc8fd813d48fa3241cef68b4
3
+ size 1465