Hugging Face's logo

PuxAI
/
GenPII-framework-Data

Safetensors
Model card Files
xet
 Community
PuxAI commited on
Commit
fcb4102
·
verified ·
1 Parent(s): cd2086c

Upload folder using huggingface_hub

Browse files
Files changed (12) hide show
  1. financial-pii-ready/TokenBased-CRF/checkpoint-63/config.json +23 -99
  2. financial-pii-ready/TokenBased-CRF/checkpoint-63/model.safetensors +2 -2
  3. financial-pii-ready/TokenBased-CRF/checkpoint-63/optimizer.pt +2 -2
  4. financial-pii-ready/TokenBased-CRF/checkpoint-63/rng_state.pth +1 -1
  5. financial-pii-ready/TokenBased-CRF/checkpoint-63/tokenizer.json +0 -0
  6. financial-pii-ready/TokenBased-CRF/checkpoint-63/tokenizer_config.json +5 -7
  7. financial-pii-ready/TokenBased-CRF/checkpoint-63/trainer_state.json +11 -2
  8. financial-pii-ready/TokenBased-CRF/checkpoint-63/training_args.bin +1 -1
  9. financial-pii-ready/TokenBased-CRF/config.json +23 -99
  10. financial-pii-ready/TokenBased-CRF/pytorch_model.bin +2 -2
  11. financial-pii-ready/TokenBased-CRF/tokenizer.json +0 -0
  12. financial-pii-ready/TokenBased-CRF/tokenizer_config.json +5 -7
financial-pii-ready/TokenBased-CRF/checkpoint-63/config.json CHANGED
@@ -1,116 +1,40 @@
1
  {
 
2
  "architectures": [
3
  "TransformerCrfForTokenClassification"
4
  ],
5
- "attention_bias": false,
6
- "attention_dropout": 0.0,
7
- "bos_token_id": 50281,
8
- "classifier_activation": "gelu",
9
- "classifier_bias": false,
10
- "classifier_dropout": 0.0,
11
- "classifier_pooling": "mean",
12
- "cls_token_id": 50281,
13
- "decoder_bias": true,
14
- "deterministic_flash_attn": false,
15
  "dtype": "float32",
16
- "embedding_dropout": 0.0,
17
- "eos_token_id": 50282,
18
- "global_attn_every_n_layers": 3,
19
- "gradient_checkpointing": false,
20
- "hidden_activation": "gelu",
21
  "hidden_size": 768,
22
  "id2label": {
23
- "0": "O",
24
- "1": "B-ADDRESS",
25
- "2": "I-ADDRESS",
26
- "3": "B-COMPANY",
27
- "4": "I-COMPANY",
28
- "5": "B-CREDIT_CARD",
29
- "6": "I-CREDIT_CARD",
30
- "7": "B-EMAIL",
31
- "8": "I-EMAIL",
32
- "9": "B-NAME",
33
- "10": "I-NAME",
34
- "11": "B-PHONE",
35
- "12": "I-PHONE",
36
- "13": "B-SSN",
37
- "14": "I-SSN",
38
- "15": "B-URL",
39
- "16": "I-URL"
40
  },
41
- "initializer_cutoff_factor": 2.0,
42
  "initializer_range": 0.02,
43
- "intermediate_size": 1152,
 
44
  "label2id": {
45
- "B-ADDRESS": 1,
46
- "B-COMPANY": 3,
47
- "B-CREDIT_CARD": 5,
48
- "B-EMAIL": 7,
49
- "B-NAME": 9,
50
- "B-PHONE": 11,
51
- "B-SSN": 13,
52
- "B-URL": 15,
53
- "I-ADDRESS": 2,
54
- "I-COMPANY": 4,
55
- "I-CREDIT_CARD": 6,
56
- "I-EMAIL": 8,
57
- "I-NAME": 10,
58
- "I-PHONE": 12,
59
- "I-SSN": 14,
60
- "I-URL": 16,
61
  "O": 0
62
  },
63
- "layer_norm_eps": 1e-05,
64
- "layer_types": [
65
- "full_attention",
66
- "sliding_attention",
67
- "sliding_attention",
68
- "full_attention",
69
- "sliding_attention",
70
- "sliding_attention",
71
- "full_attention",
72
- "sliding_attention",
73
- "sliding_attention",
74
- "full_attention",
75
- "sliding_attention",
76
- "sliding_attention",
77
- "full_attention",
78
- "sliding_attention",
79
- "sliding_attention",
80
- "full_attention",
81
- "sliding_attention",
82
- "sliding_attention",
83
- "full_attention",
84
- "sliding_attention",
85
- "sliding_attention",
86
- "full_attention"
87
- ],
88
- "local_attention": 128,
89
- "max_position_embeddings": 8192,
90
- "mlp_bias": false,
91
- "mlp_dropout": 0.0,
92
- "model_type": "modernbert",
93
- "norm_bias": false,
94
- "norm_eps": 1e-05,
95
  "num_attention_heads": 12,
96
- "num_hidden_layers": 22,
97
- "pad_token_id": 50283,
98
- "position_embedding_type": "absolute",
99
- "rope_parameters": {
100
- "full_attention": {
101
- "rope_theta": 160000.0,
102
- "rope_type": "default"
103
- },
104
- "sliding_attention": {
105
- "rope_theta": 10000.0,
106
- "rope_type": "default"
107
- }
108
- },
109
- "sep_token_id": 50282,
110
- "sparse_pred_ignore_index": -100,
111
- "sparse_prediction": false,
112
  "tie_word_embeddings": true,
113
  "transformers_version": "5.3.0",
 
114
  "use_cache": false,
115
- "vocab_size": 50368
116
  }
 
1
  {
2
+ "add_cross_attention": false,
3
  "architectures": [
4
  "TransformerCrfForTokenClassification"
5
  ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": null,
8
+ "classifier_dropout": null,
9
+ "directionality": "bidi",
 
 
 
 
 
 
10
  "dtype": "float32",
11
+ "eos_token_id": null,
12
+ "hidden_act": "gelu",
13
+ "hidden_dropout_prob": 0.1,
 
 
14
  "hidden_size": 768,
15
  "id2label": {
16
+ "0": "O"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
17
  },
 
18
  "initializer_range": 0.02,
19
+ "intermediate_size": 3072,
20
+ "is_decoder": false,
21
  "label2id": {
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
22
  "O": 0
23
  },
24
+ "layer_norm_eps": 1e-12,
25
+ "max_position_embeddings": 512,
26
+ "model_type": "bert",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
27
  "num_attention_heads": 12,
28
+ "num_hidden_layers": 12,
29
+ "pad_token_id": 0,
30
+ "pooler_fc_size": 768,
31
+ "pooler_num_attention_heads": 12,
32
+ "pooler_num_fc_layers": 3,
33
+ "pooler_size_per_head": 128,
34
+ "pooler_type": "first_token_transform",
 
 
 
 
 
 
 
 
 
35
  "tie_word_embeddings": true,
36
  "transformers_version": "5.3.0",
37
+ "type_vocab_size": 2,
38
  "use_cache": false,
39
+ "vocab_size": 119547
40
  }
financial-pii-ready/TokenBased-CRF/checkpoint-63/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:bb2de5121940ea499d69515b14943ddc906365e18614190136df170dd83d7d15
3
- size 596125728
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a33fa8f20989e1678a3edc77b7b2fddfda14c9f0fb76b514af560bf54babbac
3
+ size 711441992
financial-pii-ready/TokenBased-CRF/checkpoint-63/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f50d75c63f240db17404009f1824339ab3d3605828257e6872700b30067727cb
3
- size 1192340747
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:79148412b64f6a2290a7249211de11a1587d63b7a027284d03c3ba7d2f16c7e8
3
+ size 1418281163
financial-pii-ready/TokenBased-CRF/checkpoint-63/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7991aa545aa2a9819a120cc19b48bfbf88dcaec121a00f803b9fbabf38d57d55
3
  size 14645
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f39eab8ed980549bfffcd8b948e8852bef979d820a1c47a898b2c9f270cc3986
3
  size 14645
financial-pii-ready/TokenBased-CRF/checkpoint-63/tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff
 
financial-pii-ready/TokenBased-CRF/checkpoint-63/tokenizer_config.json CHANGED
@@ -1,17 +1,15 @@
1
  {
2
  "add_prefix_space": true,
3
  "backend": "tokenizers",
4
- "clean_up_tokenization_spaces": true,
5
  "cls_token": "[CLS]",
 
6
  "is_local": false,
7
  "mask_token": "[MASK]",
8
- "model_input_names": [
9
- "input_ids",
10
- "attention_mask"
11
- ],
12
- "model_max_length": 8192,
13
  "pad_token": "[PAD]",
14
  "sep_token": "[SEP]",
15
- "tokenizer_class": "TokenizersBackend",
 
 
16
  "unk_token": "[UNK]"
17
  }
 
1
  {
2
  "add_prefix_space": true,
3
  "backend": "tokenizers",
 
4
  "cls_token": "[CLS]",
5
+ "do_lower_case": false,
6
  "is_local": false,
7
  "mask_token": "[MASK]",
8
+ "model_max_length": 512,
 
 
 
 
9
  "pad_token": "[PAD]",
10
  "sep_token": "[SEP]",
11
+ "strip_accents": null,
12
+ "tokenize_chinese_chars": true,
13
+ "tokenizer_class": "BertTokenizer",
14
  "unk_token": "[UNK]"
15
  }
financial-pii-ready/TokenBased-CRF/checkpoint-63/trainer_state.json CHANGED
@@ -8,7 +8,16 @@
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
11
- "log_history": [],
 
 
 
 
 
 
 
 
 
12
  "logging_steps": 500,
13
  "max_steps": 63,
14
  "num_input_tokens_seen": 0,
@@ -26,7 +35,7 @@
26
  "attributes": {}
27
  }
28
  },
29
- "total_flos": 169489987584000.0,
30
  "train_batch_size": 8,
31
  "trial_name": null,
32
  "trial_params": null
 
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 1.0,
14
+ "eval_loss": 0.0,
15
+ "eval_runtime": 0.5944,
16
+ "eval_samples_per_second": 50.475,
17
+ "eval_steps_per_second": 6.73,
18
+ "step": 63
19
+ }
20
+ ],
21
  "logging_steps": 500,
22
  "max_steps": 63,
23
  "num_input_tokens_seen": 0,
 
35
  "attributes": {}
36
  }
37
  },
38
+ "total_flos": 131554351104000.0,
39
  "train_batch_size": 8,
40
  "trial_name": null,
41
  "trial_params": null
financial-pii-ready/TokenBased-CRF/checkpoint-63/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5a3e9ae25a3c620ff30c32b0562f13e8c74a0630c6e5f3c6ff751337dbed4ff7
3
  size 5201
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:431f28d8e4182178155b38bfeaf84a2fb1983fcfb72ede91d8c85d444114f61a
3
  size 5201
financial-pii-ready/TokenBased-CRF/config.json CHANGED
@@ -1,116 +1,40 @@
1
  {
 
2
  "architectures": [
3
  "TransformerCrfForTokenClassification"
4
  ],
5
- "attention_bias": false,
6
- "attention_dropout": 0.0,
7
- "bos_token_id": 50281,
8
- "classifier_activation": "gelu",
9
- "classifier_bias": false,
10
- "classifier_dropout": 0.0,
11
- "classifier_pooling": "mean",
12
- "cls_token_id": 50281,
13
- "decoder_bias": true,
14
- "deterministic_flash_attn": false,
15
  "dtype": "float32",
16
- "embedding_dropout": 0.0,
17
- "eos_token_id": 50282,
18
- "global_attn_every_n_layers": 3,
19
- "gradient_checkpointing": false,
20
- "hidden_activation": "gelu",
21
  "hidden_size": 768,
22
  "id2label": {
23
- "0": "O",
24
- "1": "B-ADDRESS",
25
- "2": "I-ADDRESS",
26
- "3": "B-COMPANY",
27
- "4": "I-COMPANY",
28
- "5": "B-CREDIT_CARD",
29
- "6": "I-CREDIT_CARD",
30
- "7": "B-EMAIL",
31
- "8": "I-EMAIL",
32
- "9": "B-NAME",
33
- "10": "I-NAME",
34
- "11": "B-PHONE",
35
- "12": "I-PHONE",
36
- "13": "B-SSN",
37
- "14": "I-SSN",
38
- "15": "B-URL",
39
- "16": "I-URL"
40
  },
41
- "initializer_cutoff_factor": 2.0,
42
  "initializer_range": 0.02,
43
- "intermediate_size": 1152,
 
44
  "label2id": {
45
- "B-ADDRESS": 1,
46
- "B-COMPANY": 3,
47
- "B-CREDIT_CARD": 5,
48
- "B-EMAIL": 7,
49
- "B-NAME": 9,
50
- "B-PHONE": 11,
51
- "B-SSN": 13,
52
- "B-URL": 15,
53
- "I-ADDRESS": 2,
54
- "I-COMPANY": 4,
55
- "I-CREDIT_CARD": 6,
56
- "I-EMAIL": 8,
57
- "I-NAME": 10,
58
- "I-PHONE": 12,
59
- "I-SSN": 14,
60
- "I-URL": 16,
61
  "O": 0
62
  },
63
- "layer_norm_eps": 1e-05,
64
- "layer_types": [
65
- "full_attention",
66
- "sliding_attention",
67
- "sliding_attention",
68
- "full_attention",
69
- "sliding_attention",
70
- "sliding_attention",
71
- "full_attention",
72
- "sliding_attention",
73
- "sliding_attention",
74
- "full_attention",
75
- "sliding_attention",
76
- "sliding_attention",
77
- "full_attention",
78
- "sliding_attention",
79
- "sliding_attention",
80
- "full_attention",
81
- "sliding_attention",
82
- "sliding_attention",
83
- "full_attention",
84
- "sliding_attention",
85
- "sliding_attention",
86
- "full_attention"
87
- ],
88
- "local_attention": 128,
89
- "max_position_embeddings": 8192,
90
- "mlp_bias": false,
91
- "mlp_dropout": 0.0,
92
- "model_type": "modernbert",
93
- "norm_bias": false,
94
- "norm_eps": 1e-05,
95
  "num_attention_heads": 12,
96
- "num_hidden_layers": 22,
97
- "pad_token_id": 50283,
98
- "position_embedding_type": "absolute",
99
- "rope_parameters": {
100
- "full_attention": {
101
- "rope_theta": 160000.0,
102
- "rope_type": "default"
103
- },
104
- "sliding_attention": {
105
- "rope_theta": 10000.0,
106
- "rope_type": "default"
107
- }
108
- },
109
- "sep_token_id": 50282,
110
- "sparse_pred_ignore_index": -100,
111
- "sparse_prediction": false,
112
  "tie_word_embeddings": true,
113
  "transformers_version": "5.3.0",
 
114
  "use_cache": false,
115
- "vocab_size": 50368
116
  }
 
1
  {
2
+ "add_cross_attention": false,
3
  "architectures": [
4
  "TransformerCrfForTokenClassification"
5
  ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": null,
8
+ "classifier_dropout": null,
9
+ "directionality": "bidi",
 
 
 
 
 
 
10
  "dtype": "float32",
11
+ "eos_token_id": null,
12
+ "hidden_act": "gelu",
13
+ "hidden_dropout_prob": 0.1,
 
 
14
  "hidden_size": 768,
15
  "id2label": {
16
+ "0": "O"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
17
  },
 
18
  "initializer_range": 0.02,
19
+ "intermediate_size": 3072,
20
+ "is_decoder": false,
21
  "label2id": {
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
22
  "O": 0
23
  },
24
+ "layer_norm_eps": 1e-12,
25
+ "max_position_embeddings": 512,
26
+ "model_type": "bert",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
27
  "num_attention_heads": 12,
28
+ "num_hidden_layers": 12,
29
+ "pad_token_id": 0,
30
+ "pooler_fc_size": 768,
31
+ "pooler_num_attention_heads": 12,
32
+ "pooler_num_fc_layers": 3,
33
+ "pooler_size_per_head": 128,
34
+ "pooler_type": "first_token_transform",
 
 
 
 
 
 
 
 
 
35
  "tie_word_embeddings": true,
36
  "transformers_version": "5.3.0",
37
+ "type_vocab_size": 2,
38
  "use_cache": false,
39
+ "vocab_size": 119547
40
  }
financial-pii-ready/TokenBased-CRF/pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:798dac6bd4c7223f4c9479c9b3366a3d5d59047b2246056d2cb3f12c52a9b5f0
3
- size 596171855
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f62b92b941f41460837da75a5cc8e4a3ca9b95fa49700be682eb87c0ff17c349
3
+ size 711504083
financial-pii-ready/TokenBased-CRF/tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff
 
financial-pii-ready/TokenBased-CRF/tokenizer_config.json CHANGED
@@ -1,17 +1,15 @@
1
  {
2
  "add_prefix_space": true,
3
  "backend": "tokenizers",
4
- "clean_up_tokenization_spaces": true,
5
  "cls_token": "[CLS]",
 
6
  "is_local": false,
7
  "mask_token": "[MASK]",
8
- "model_input_names": [
9
- "input_ids",
10
- "attention_mask"
11
- ],
12
- "model_max_length": 8192,
13
  "pad_token": "[PAD]",
14
  "sep_token": "[SEP]",
15
- "tokenizer_class": "TokenizersBackend",
 
 
16
  "unk_token": "[UNK]"
17
  }
 
1
  {
2
  "add_prefix_space": true,
3
  "backend": "tokenizers",
 
4
  "cls_token": "[CLS]",
5
+ "do_lower_case": false,
6
  "is_local": false,
7
  "mask_token": "[MASK]",
8
+ "model_max_length": 512,
 
 
 
 
9
  "pad_token": "[PAD]",
10
  "sep_token": "[SEP]",
11
+ "strip_accents": null,
12
+ "tokenize_chinese_chars": true,
13
+ "tokenizer_class": "BertTokenizer",
14
  "unk_token": "[UNK]"
15
  }