Hugging Face's logo

PuxAI
/
GenPII-framework

Safetensors
Model card Files
xet
 Community
PuxAI commited on
Commit
bad289a
·
verified ·
1 Parent(s): 6df3758

Delete gorkem-pii-ready

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/gliner_config.json +0 -105
  2. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/optimizer.pt +0 -3
  3. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/pytorch_model.bin +0 -3
  4. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/rng_state.pth +0 -3
  5. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/scheduler.pt +0 -3
  6. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/tokenizer.json +0 -0
  7. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/tokenizer_config.json +0 -24
  8. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/trainer_state.json +0 -41
  9. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/gliner_config.json +0 -105
  10. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/optimizer.pt +0 -3
  11. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/pytorch_model.bin +0 -3
  12. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/rng_state.pth +0 -3
  13. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/scheduler.pt +0 -3
  14. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/tokenizer.json +0 -0
  15. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/tokenizer_config.json +0 -24
  16. gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/trainer_state.json +0 -41
  17. gorkem-pii-ready/GLiNER-NVIDIA/gliner_config.json +0 -105
  18. gorkem-pii-ready/GLiNER-NVIDIA/pytorch_model.bin +0 -3
  19. gorkem-pii-ready/GLiNER-NVIDIA/tokenizer.json +0 -0
  20. gorkem-pii-ready/GLiNER-NVIDIA/tokenizer_config.json +0 -24
  21. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/config.json +0 -63
  22. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/model.safetensors +0 -3
  23. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/optimizer.pt +0 -3
  24. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/rng_state.pth +0 -3
  25. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/scheduler.pt +0 -3
  26. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer.json +0 -0
  27. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer_config.json +0 -23
  28. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/trainer_state.json +0 -48
  29. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/training_args.bin +0 -3
  30. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/config.json +0 -63
  31. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/model.safetensors +0 -3
  32. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/optimizer.pt +0 -3
  33. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/rng_state.pth +0 -3
  34. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/scheduler.pt +0 -3
  35. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/tokenizer.json +0 -0
  36. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/tokenizer_config.json +0 -23
  37. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/trainer_state.json +0 -48
  38. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/training_args.bin +0 -3
  39. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/config.json +0 -63
  40. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/model.safetensors +0 -3
  41. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/optimizer.pt +0 -3
  42. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/rng_state.pth +0 -3
  43. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/scheduler.pt +0 -3
  44. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/tokenizer.json +0 -0
  45. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/tokenizer_config.json +0 -23
  46. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/trainer_state.json +0 -41
  47. gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/training_args.bin +0 -3
  48. gorkem-pii-ready/SpanBased-CustomDeBERTa/config.json +0 -63
  49. gorkem-pii-ready/SpanBased-CustomDeBERTa/pytorch_model.bin +0 -3
  50. gorkem-pii-ready/SpanBased-CustomDeBERTa/tokenizer.json +0 -0
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/gliner_config.json DELETED
@@ -1,105 +0,0 @@
1
- {
2
- "bos_token_id": 1,
3
- "class_token_index": 128002,
4
- "dropout": 0.4,
5
- "embed_ent_token": true,
6
- "encoder_config": {
7
- "_name_or_path": "microsoft/deberta-v3-large",
8
- "add_cross_attention": false,
9
- "architectures": null,
10
- "attention_probs_dropout_prob": 0.1,
11
- "bos_token_id": null,
12
- "chunk_size_feed_forward": 0,
13
- "cross_attention_hidden_size": null,
14
- "decoder_start_token_id": null,
15
- "dtype": null,
16
- "eos_token_id": null,
17
- "finetuning_task": null,
18
- "hidden_act": "gelu",
19
- "hidden_dropout_prob": 0.1,
20
- "hidden_size": 1024,
21
- "id2label": {
22
- "0": "LABEL_0",
23
- "1": "LABEL_1"
24
- },
25
- "initializer_range": 0.02,
26
- "intermediate_size": 4096,
27
- "is_decoder": false,
28
- "is_encoder_decoder": false,
29
- "label2id": {
30
- "LABEL_0": 0,
31
- "LABEL_1": 1
32
- },
33
- "layer_norm_eps": 1e-07,
34
- "legacy": true,
35
- "max_position_embeddings": 512,
36
- "max_relative_positions": -1,
37
- "model_type": "deberta-v2",
38
- "norm_rel_ebd": "layer_norm",
39
- "num_attention_heads": 16,
40
- "num_hidden_layers": 24,
41
- "output_attentions": false,
42
- "output_hidden_states": false,
43
- "pad_token_id": 0,
44
- "pooler_dropout": 0,
45
- "pooler_hidden_act": "gelu",
46
- "pooler_hidden_size": 1024,
47
- "pos_att_type": [
48
- "p2c",
49
- "c2p"
50
- ],
51
- "position_biased_input": false,
52
- "position_buckets": 256,
53
- "prefix": null,
54
- "problem_type": null,
55
- "pruned_heads": {},
56
- "relative_attention": true,
57
- "return_dict": true,
58
- "sep_token_id": null,
59
- "share_att_key": true,
60
- "task_specific_params": null,
61
- "tf_legacy_loss": false,
62
- "tie_encoder_decoder": false,
63
- "tie_word_embeddings": true,
64
- "tokenizer_class": null,
65
- "torchscript": false,
66
- "type_vocab_size": 0,
67
- "use_bfloat16": false,
68
- "vocab_size": 128004
69
- },
70
- "ent_token": "<<ENT>>",
71
- "eos_token_id": 2,
72
- "eval_every": 5000,
73
- "fine_tune": true,
74
- "fuse_layers": false,
75
- "has_rnn": true,
76
- "hidden_size": 512,
77
- "labels_encoder": null,
78
- "labels_encoder_config": null,
79
- "lr_encoder": "1e-5",
80
- "lr_others": "5e-5",
81
- "max_len": 384,
82
- "max_neg_type_ratio": 1,
83
- "max_types": 25,
84
- "max_width": 12,
85
- "model_name": "microsoft/deberta-v3-large",
86
- "model_type": null,
87
- "name": "correct",
88
- "num_post_fusion_layers": 1,
89
- "num_rnn_layers": 1,
90
- "num_steps": 30000,
91
- "pad_token_id": 0,
92
- "post_fusion_schema": "",
93
- "random_drop": true,
94
- "sep_token": "<<SEP>>",
95
- "shuffle_types": true,
96
- "size_sup": -1,
97
- "span_mode": "markerV0",
98
- "subtoken_pooling": "first",
99
- "train_batch_size": 8,
100
- "transformers_version": "5.3.0",
101
- "use_cache": false,
102
- "vocab_size": 128004,
103
- "warmup_ratio": 3000,
104
- "words_splitter_type": "whitespace"
105
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/optimizer.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:038c61d8db5902a8ac43adb2f4efca69e93c8241a7dda16554ebc410e2996f9c
3
- size 2873
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/pytorch_model.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:0abe4edc77f96f88e966f61f2c72b785f8d6f1d5aca2bd845e7ac15dac597a5d
3
- size 1782005383
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/rng_state.pth DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:cc631f12a7c31a4ed81063d3b2f93ccff8c7e040567f6b9352ca937bffc67c88
3
- size 14645
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/scheduler.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:dc6c17958d31e8c41c9b9c1c8ecfb1ef18aae011c17974b49b8d96d889c3416e
3
- size 1529
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/tokenizer.json DELETED
The diff for this file is too large to render. See raw diff
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/tokenizer_config.json DELETED
@@ -1,24 +0,0 @@
1
- {
2
- "add_prefix_space": true,
3
- "backend": "tokenizers",
4
- "bos_token": "[CLS]",
5
- "clean_up_tokenization_spaces": false,
6
- "cls_token": "[CLS]",
7
- "do_lower_case": false,
8
- "eos_token": "[SEP]",
9
- "is_local": true,
10
- "mask_token": "[MASK]",
11
- "max_length": null,
12
- "model_max_length": 1000000000000000019884624838656,
13
- "pad_to_multiple_of": null,
14
- "pad_token": "[PAD]",
15
- "pad_token_type_id": 0,
16
- "padding_side": "right",
17
- "sep_token": "[SEP]",
18
- "sp_model_kwargs": {},
19
- "split_by_punct": false,
20
- "tokenizer_class": "DebertaV2Tokenizer",
21
- "unk_id": 3,
22
- "unk_token": "[UNK]",
23
- "vocab_type": "spm"
24
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-500/trainer_state.json DELETED
@@ -1,41 +0,0 @@
1
- {
2
- "best_global_step": null,
3
- "best_metric": null,
4
- "best_model_checkpoint": null,
5
- "epoch": 0.6915629322268326,
6
- "eval_steps": 500,
7
- "global_step": 500,
8
- "is_hyper_param_search": false,
9
- "is_local_process_zero": true,
10
- "is_world_process_zero": true,
11
- "log_history": [
12
- {
13
- "epoch": 0.6915629322268326,
14
- "grad_norm": 0.0,
15
- "learning_rate": 3.0982019363762106e-06,
16
- "loss": 0.0,
17
- "step": 500
18
- }
19
- ],
20
- "logging_steps": 500,
21
- "max_steps": 723,
22
- "num_input_tokens_seen": 0,
23
- "num_train_epochs": 1,
24
- "save_steps": 500,
25
- "stateful_callbacks": {
26
- "TrainerControl": {
27
- "args": {
28
- "should_epoch_stop": false,
29
- "should_evaluate": false,
30
- "should_log": false,
31
- "should_save": true,
32
- "should_training_stop": false
33
- },
34
- "attributes": {}
35
- }
36
- },
37
- "total_flos": 0.0,
38
- "train_batch_size": 16,
39
- "trial_name": null,
40
- "trial_params": null
41
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/gliner_config.json DELETED
@@ -1,105 +0,0 @@
1
- {
2
- "bos_token_id": 1,
3
- "class_token_index": 128002,
4
- "dropout": 0.4,
5
- "embed_ent_token": true,
6
- "encoder_config": {
7
- "_name_or_path": "microsoft/deberta-v3-large",
8
- "add_cross_attention": false,
9
- "architectures": null,
10
- "attention_probs_dropout_prob": 0.1,
11
- "bos_token_id": null,
12
- "chunk_size_feed_forward": 0,
13
- "cross_attention_hidden_size": null,
14
- "decoder_start_token_id": null,
15
- "dtype": null,
16
- "eos_token_id": null,
17
- "finetuning_task": null,
18
- "hidden_act": "gelu",
19
- "hidden_dropout_prob": 0.1,
20
- "hidden_size": 1024,
21
- "id2label": {
22
- "0": "LABEL_0",
23
- "1": "LABEL_1"
24
- },
25
- "initializer_range": 0.02,
26
- "intermediate_size": 4096,
27
- "is_decoder": false,
28
- "is_encoder_decoder": false,
29
- "label2id": {
30
- "LABEL_0": 0,
31
- "LABEL_1": 1
32
- },
33
- "layer_norm_eps": 1e-07,
34
- "legacy": true,
35
- "max_position_embeddings": 512,
36
- "max_relative_positions": -1,
37
- "model_type": "deberta-v2",
38
- "norm_rel_ebd": "layer_norm",
39
- "num_attention_heads": 16,
40
- "num_hidden_layers": 24,
41
- "output_attentions": false,
42
- "output_hidden_states": false,
43
- "pad_token_id": 0,
44
- "pooler_dropout": 0,
45
- "pooler_hidden_act": "gelu",
46
- "pooler_hidden_size": 1024,
47
- "pos_att_type": [
48
- "p2c",
49
- "c2p"
50
- ],
51
- "position_biased_input": false,
52
- "position_buckets": 256,
53
- "prefix": null,
54
- "problem_type": null,
55
- "pruned_heads": {},
56
- "relative_attention": true,
57
- "return_dict": true,
58
- "sep_token_id": null,
59
- "share_att_key": true,
60
- "task_specific_params": null,
61
- "tf_legacy_loss": false,
62
- "tie_encoder_decoder": false,
63
- "tie_word_embeddings": true,
64
- "tokenizer_class": null,
65
- "torchscript": false,
66
- "type_vocab_size": 0,
67
- "use_bfloat16": false,
68
- "vocab_size": 128004
69
- },
70
- "ent_token": "<<ENT>>",
71
- "eos_token_id": 2,
72
- "eval_every": 5000,
73
- "fine_tune": true,
74
- "fuse_layers": false,
75
- "has_rnn": true,
76
- "hidden_size": 512,
77
- "labels_encoder": null,
78
- "labels_encoder_config": null,
79
- "lr_encoder": "1e-5",
80
- "lr_others": "5e-5",
81
- "max_len": 384,
82
- "max_neg_type_ratio": 1,
83
- "max_types": 25,
84
- "max_width": 12,
85
- "model_name": "microsoft/deberta-v3-large",
86
- "model_type": null,
87
- "name": "correct",
88
- "num_post_fusion_layers": 1,
89
- "num_rnn_layers": 1,
90
- "num_steps": 30000,
91
- "pad_token_id": 0,
92
- "post_fusion_schema": "",
93
- "random_drop": true,
94
- "sep_token": "<<SEP>>",
95
- "shuffle_types": true,
96
- "size_sup": -1,
97
- "span_mode": "markerV0",
98
- "subtoken_pooling": "first",
99
- "train_batch_size": 8,
100
- "transformers_version": "5.3.0",
101
- "use_cache": false,
102
- "vocab_size": 128004,
103
- "warmup_ratio": 3000,
104
- "words_splitter_type": "whitespace"
105
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/optimizer.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:95160d2070a1730f206cc65410042a7a4495f7c3dba2d5c0fa679181dd51fad6
3
- size 2873
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/pytorch_model.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:0abe4edc77f96f88e966f61f2c72b785f8d6f1d5aca2bd845e7ac15dac597a5d
3
- size 1782005383
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/rng_state.pth DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:c4f2a5faa9e3eebe16aec0927eee8c4d9069b7e7d4993617667c596fb97d3878
3
- size 14645
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/scheduler.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:0843d80c2f81a09a5c2f8c1c64e1d7b7fce1d1e73e5b3dda05763cced875850e
3
- size 1529
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/tokenizer.json DELETED
The diff for this file is too large to render. See raw diff
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/tokenizer_config.json DELETED
@@ -1,24 +0,0 @@
1
- {
2
- "add_prefix_space": true,
3
- "backend": "tokenizers",
4
- "bos_token": "[CLS]",
5
- "clean_up_tokenization_spaces": false,
6
- "cls_token": "[CLS]",
7
- "do_lower_case": false,
8
- "eos_token": "[SEP]",
9
- "is_local": true,
10
- "mask_token": "[MASK]",
11
- "max_length": null,
12
- "model_max_length": 1000000000000000019884624838656,
13
- "pad_to_multiple_of": null,
14
- "pad_token": "[PAD]",
15
- "pad_token_type_id": 0,
16
- "padding_side": "right",
17
- "sep_token": "[SEP]",
18
- "sp_model_kwargs": {},
19
- "split_by_punct": false,
20
- "tokenizer_class": "DebertaV2Tokenizer",
21
- "unk_id": 3,
22
- "unk_token": "[UNK]",
23
- "vocab_type": "spm"
24
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/checkpoint-723/trainer_state.json DELETED
@@ -1,41 +0,0 @@
1
- {
2
- "best_global_step": null,
3
- "best_metric": null,
4
- "best_model_checkpoint": null,
5
- "epoch": 1.0,
6
- "eval_steps": 500,
7
- "global_step": 723,
8
- "is_hyper_param_search": false,
9
- "is_local_process_zero": true,
10
- "is_world_process_zero": true,
11
- "log_history": [
12
- {
13
- "epoch": 0.6915629322268326,
14
- "grad_norm": 0.0,
15
- "learning_rate": 3.0982019363762106e-06,
16
- "loss": 0.0,
17
- "step": 500
18
- }
19
- ],
20
- "logging_steps": 500,
21
- "max_steps": 723,
22
- "num_input_tokens_seen": 0,
23
- "num_train_epochs": 1,
24
- "save_steps": 500,
25
- "stateful_callbacks": {
26
- "TrainerControl": {
27
- "args": {
28
- "should_epoch_stop": false,
29
- "should_evaluate": false,
30
- "should_log": false,
31
- "should_save": true,
32
- "should_training_stop": true
33
- },
34
- "attributes": {}
35
- }
36
- },
37
- "total_flos": 0.0,
38
- "train_batch_size": 16,
39
- "trial_name": null,
40
- "trial_params": null
41
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/gliner_config.json DELETED
@@ -1,105 +0,0 @@
1
- {
2
- "bos_token_id": 1,
3
- "class_token_index": 128002,
4
- "dropout": 0.4,
5
- "embed_ent_token": true,
6
- "encoder_config": {
7
- "_name_or_path": "microsoft/deberta-v3-large",
8
- "add_cross_attention": false,
9
- "architectures": null,
10
- "attention_probs_dropout_prob": 0.1,
11
- "bos_token_id": null,
12
- "chunk_size_feed_forward": 0,
13
- "cross_attention_hidden_size": null,
14
- "decoder_start_token_id": null,
15
- "dtype": null,
16
- "eos_token_id": null,
17
- "finetuning_task": null,
18
- "hidden_act": "gelu",
19
- "hidden_dropout_prob": 0.1,
20
- "hidden_size": 1024,
21
- "id2label": {
22
- "0": "LABEL_0",
23
- "1": "LABEL_1"
24
- },
25
- "initializer_range": 0.02,
26
- "intermediate_size": 4096,
27
- "is_decoder": false,
28
- "is_encoder_decoder": false,
29
- "label2id": {
30
- "LABEL_0": 0,
31
- "LABEL_1": 1
32
- },
33
- "layer_norm_eps": 1e-07,
34
- "legacy": true,
35
- "max_position_embeddings": 512,
36
- "max_relative_positions": -1,
37
- "model_type": "deberta-v2",
38
- "norm_rel_ebd": "layer_norm",
39
- "num_attention_heads": 16,
40
- "num_hidden_layers": 24,
41
- "output_attentions": false,
42
- "output_hidden_states": false,
43
- "pad_token_id": 0,
44
- "pooler_dropout": 0,
45
- "pooler_hidden_act": "gelu",
46
- "pooler_hidden_size": 1024,
47
- "pos_att_type": [
48
- "p2c",
49
- "c2p"
50
- ],
51
- "position_biased_input": false,
52
- "position_buckets": 256,
53
- "prefix": null,
54
- "problem_type": null,
55
- "pruned_heads": {},
56
- "relative_attention": true,
57
- "return_dict": true,
58
- "sep_token_id": null,
59
- "share_att_key": true,
60
- "task_specific_params": null,
61
- "tf_legacy_loss": false,
62
- "tie_encoder_decoder": false,
63
- "tie_word_embeddings": true,
64
- "tokenizer_class": null,
65
- "torchscript": false,
66
- "type_vocab_size": 0,
67
- "use_bfloat16": false,
68
- "vocab_size": 128004
69
- },
70
- "ent_token": "<<ENT>>",
71
- "eos_token_id": 2,
72
- "eval_every": 5000,
73
- "fine_tune": true,
74
- "fuse_layers": false,
75
- "has_rnn": true,
76
- "hidden_size": 512,
77
- "labels_encoder": null,
78
- "labels_encoder_config": null,
79
- "lr_encoder": "1e-5",
80
- "lr_others": "5e-5",
81
- "max_len": 384,
82
- "max_neg_type_ratio": 1,
83
- "max_types": 25,
84
- "max_width": 12,
85
- "model_name": "microsoft/deberta-v3-large",
86
- "model_type": null,
87
- "name": "correct",
88
- "num_post_fusion_layers": 1,
89
- "num_rnn_layers": 1,
90
- "num_steps": 30000,
91
- "pad_token_id": 0,
92
- "post_fusion_schema": "",
93
- "random_drop": true,
94
- "sep_token": "<<SEP>>",
95
- "shuffle_types": true,
96
- "size_sup": -1,
97
- "span_mode": "markerV0",
98
- "subtoken_pooling": "first",
99
- "train_batch_size": 8,
100
- "transformers_version": "5.3.0",
101
- "use_cache": false,
102
- "vocab_size": 128004,
103
- "warmup_ratio": 3000,
104
- "words_splitter_type": "whitespace"
105
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/pytorch_model.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:0abe4edc77f96f88e966f61f2c72b785f8d6f1d5aca2bd845e7ac15dac597a5d
3
- size 1782005383
 
 
 
 
gorkem-pii-ready/GLiNER-NVIDIA/tokenizer.json DELETED
The diff for this file is too large to render. See raw diff
 
gorkem-pii-ready/GLiNER-NVIDIA/tokenizer_config.json DELETED
@@ -1,24 +0,0 @@
1
- {
2
- "add_prefix_space": true,
3
- "backend": "tokenizers",
4
- "bos_token": "[CLS]",
5
- "clean_up_tokenization_spaces": false,
6
- "cls_token": "[CLS]",
7
- "do_lower_case": false,
8
- "eos_token": "[SEP]",
9
- "is_local": true,
10
- "mask_token": "[MASK]",
11
- "max_length": null,
12
- "model_max_length": 1000000000000000019884624838656,
13
- "pad_to_multiple_of": null,
14
- "pad_token": "[PAD]",
15
- "pad_token_type_id": 0,
16
- "padding_side": "right",
17
- "sep_token": "[SEP]",
18
- "sp_model_kwargs": {},
19
- "split_by_punct": false,
20
- "tokenizer_class": "DebertaV2Tokenizer",
21
- "unk_id": 3,
22
- "unk_token": "[UNK]",
23
- "vocab_type": "spm"
24
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/config.json DELETED
@@ -1,63 +0,0 @@
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": "ADDRESS",
15
- "2": "CREDIT_CARD",
16
- "3": "CRYPTO_ADDRESS",
17
- "4": "EMAIL",
18
- "5": "IBAN",
19
- "6": "OFF_PLATFORM_ATTEMPT",
20
- "7": "PHONE",
21
- "8": "SOCIAL_MEDIA",
22
- "9": "URL"
23
- },
24
- "initializer_range": 0.02,
25
- "intermediate_size": 3072,
26
- "label2id": {
27
- "ADDRESS": 1,
28
- "CREDIT_CARD": 2,
29
- "CRYPTO_ADDRESS": 3,
30
- "EMAIL": 4,
31
- "IBAN": 5,
32
- "O": 0,
33
- "OFF_PLATFORM_ATTEMPT": 6,
34
- "PHONE": 7,
35
- "SOCIAL_MEDIA": 8,
36
- "URL": 9
37
- },
38
- "layer_norm_eps": 1e-07,
39
- "legacy": true,
40
- "max_position_embeddings": 512,
41
- "max_relative_positions": -1,
42
- "model_type": "deberta-v2",
43
- "norm_rel_ebd": "layer_norm",
44
- "num_attention_heads": 12,
45
- "num_hidden_layers": 12,
46
- "pad_token_id": 0,
47
- "pooler_dropout": 0,
48
- "pooler_hidden_act": "gelu",
49
- "pooler_hidden_size": 768,
50
- "pos_att_type": [
51
- "p2c",
52
- "c2p"
53
- ],
54
- "position_biased_input": false,
55
- "position_buckets": 256,
56
- "relative_attention": true,
57
- "share_att_key": true,
58
- "tie_word_embeddings": true,
59
- "transformers_version": "5.3.0",
60
- "type_vocab_size": 0,
61
- "use_cache": false,
62
- "vocab_size": 128100
63
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/model.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:b50cd5ab3c7657588b9a8806b8067cff40a5a88f08af036cb65a7c350c6e3910
3
- size 741359744
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/optimizer.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:81efb579de78cb0c15910daed7b2813085eb412a6429ddf08ef0fd3f95424b29
3
- size 1482843147
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/rng_state.pth DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:71cb42a9e4441862670a2ab5ca064012a2845f17a7eaf1f9dcd3ae09778900b8
3
- size 14645
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/scheduler.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:1ed3b7879c9d08310fa68bd1af5c1b9a3843a452ce133fa854cced73ec663d43
3
- size 1465
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer.json DELETED
The diff for this file is too large to render. See raw diff
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/tokenizer_config.json DELETED
@@ -1,23 +0,0 @@
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
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/trainer_state.json DELETED
@@ -1,48 +0,0 @@
1
- {
2
- "best_global_step": null,
3
- "best_metric": null,
4
- "best_model_checkpoint": null,
5
- "epoch": 1.3831258644536653,
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": 0.6915629322268326,
14
- "grad_norm": 0.06808845698833466,
15
- "learning_rate": 1.3098201936376211e-05,
16
- "loss": 0.012068235397338867,
17
- "step": 500
18
- },
19
- {
20
- "epoch": 1.3831258644536653,
21
- "grad_norm": 0.031969357281923294,
22
- "learning_rate": 6.182572614107885e-06,
23
- "loss": 0.0036826863288879396,
24
- "step": 1000
25
- }
26
- ],
27
- "logging_steps": 500,
28
- "max_steps": 1446,
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": 243931167235368.0,
45
- "train_batch_size": 16,
46
- "trial_name": null,
47
- "trial_params": null
48
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1000/training_args.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:7e609457181ac50440df7d6728c3518d3b9316719fead1acbcffdb0f4941dd06
3
- size 5201
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/config.json DELETED
@@ -1,63 +0,0 @@
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": "ADDRESS",
15
- "2": "CREDIT_CARD",
16
- "3": "CRYPTO_ADDRESS",
17
- "4": "EMAIL",
18
- "5": "IBAN",
19
- "6": "OFF_PLATFORM_ATTEMPT",
20
- "7": "PHONE",
21
- "8": "SOCIAL_MEDIA",
22
- "9": "URL"
23
- },
24
- "initializer_range": 0.02,
25
- "intermediate_size": 3072,
26
- "label2id": {
27
- "ADDRESS": 1,
28
- "CREDIT_CARD": 2,
29
- "CRYPTO_ADDRESS": 3,
30
- "EMAIL": 4,
31
- "IBAN": 5,
32
- "O": 0,
33
- "OFF_PLATFORM_ATTEMPT": 6,
34
- "PHONE": 7,
35
- "SOCIAL_MEDIA": 8,
36
- "URL": 9
37
- },
38
- "layer_norm_eps": 1e-07,
39
- "legacy": true,
40
- "max_position_embeddings": 512,
41
- "max_relative_positions": -1,
42
- "model_type": "deberta-v2",
43
- "norm_rel_ebd": "layer_norm",
44
- "num_attention_heads": 12,
45
- "num_hidden_layers": 12,
46
- "pad_token_id": 0,
47
- "pooler_dropout": 0,
48
- "pooler_hidden_act": "gelu",
49
- "pooler_hidden_size": 768,
50
- "pos_att_type": [
51
- "p2c",
52
- "c2p"
53
- ],
54
- "position_biased_input": false,
55
- "position_buckets": 256,
56
- "relative_attention": true,
57
- "share_att_key": true,
58
- "tie_word_embeddings": true,
59
- "transformers_version": "5.3.0",
60
- "type_vocab_size": 0,
61
- "use_cache": false,
62
- "vocab_size": 128100
63
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/model.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:e0451ac8b4df0b6b6d533b31ec15c485262a83d16df5fbbf3ea11188baee6592
3
- size 741359744
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/optimizer.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:f24e1abe2c8864c29da0680eee30c4c32cb4197b8014d068f5c3a96e0ec8bbab
3
- size 1482843147
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/rng_state.pth DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:35e65ddbf7e3c4fc0d4d3c23af53d056bf5459ff4970dcb0db5997c61b26527e
3
- size 14645
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/scheduler.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:b8742f08c44d981613576954e914b525feeeddfe29bf72a4b0ac0e765a99171c
3
- size 1465
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/tokenizer.json DELETED
The diff for this file is too large to render. See raw diff
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/tokenizer_config.json DELETED
@@ -1,23 +0,0 @@
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
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/trainer_state.json DELETED
@@ -1,48 +0,0 @@
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": 1446,
8
- "is_hyper_param_search": false,
9
- "is_local_process_zero": true,
10
- "is_world_process_zero": true,
11
- "log_history": [
12
- {
13
- "epoch": 0.6915629322268326,
14
- "grad_norm": 0.06808845698833466,
15
- "learning_rate": 1.3098201936376211e-05,
16
- "loss": 0.012068235397338867,
17
- "step": 500
18
- },
19
- {
20
- "epoch": 1.3831258644536653,
21
- "grad_norm": 0.031969357281923294,
22
- "learning_rate": 6.182572614107885e-06,
23
- "loss": 0.0036826863288879396,
24
- "step": 1000
25
- }
26
- ],
27
- "logging_steps": 500,
28
- "max_steps": 1446,
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": true
40
- },
41
- "attributes": {}
42
- }
43
- },
44
- "total_flos": 353148288762204.0,
45
- "train_batch_size": 16,
46
- "trial_name": null,
47
- "trial_params": null
48
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-1446/training_args.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:7e609457181ac50440df7d6728c3518d3b9316719fead1acbcffdb0f4941dd06
3
- size 5201
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/config.json DELETED
@@ -1,63 +0,0 @@
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": "ADDRESS",
15
- "2": "CREDIT_CARD",
16
- "3": "CRYPTO_ADDRESS",
17
- "4": "EMAIL",
18
- "5": "IBAN",
19
- "6": "OFF_PLATFORM_ATTEMPT",
20
- "7": "PHONE",
21
- "8": "SOCIAL_MEDIA",
22
- "9": "URL"
23
- },
24
- "initializer_range": 0.02,
25
- "intermediate_size": 3072,
26
- "label2id": {
27
- "ADDRESS": 1,
28
- "CREDIT_CARD": 2,
29
- "CRYPTO_ADDRESS": 3,
30
- "EMAIL": 4,
31
- "IBAN": 5,
32
- "O": 0,
33
- "OFF_PLATFORM_ATTEMPT": 6,
34
- "PHONE": 7,
35
- "SOCIAL_MEDIA": 8,
36
- "URL": 9
37
- },
38
- "layer_norm_eps": 1e-07,
39
- "legacy": true,
40
- "max_position_embeddings": 512,
41
- "max_relative_positions": -1,
42
- "model_type": "deberta-v2",
43
- "norm_rel_ebd": "layer_norm",
44
- "num_attention_heads": 12,
45
- "num_hidden_layers": 12,
46
- "pad_token_id": 0,
47
- "pooler_dropout": 0,
48
- "pooler_hidden_act": "gelu",
49
- "pooler_hidden_size": 768,
50
- "pos_att_type": [
51
- "p2c",
52
- "c2p"
53
- ],
54
- "position_biased_input": false,
55
- "position_buckets": 256,
56
- "relative_attention": true,
57
- "share_att_key": true,
58
- "tie_word_embeddings": true,
59
- "transformers_version": "5.3.0",
60
- "type_vocab_size": 0,
61
- "use_cache": false,
62
- "vocab_size": 128100
63
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/model.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:eaea070db4574cbd17227d4d3af5838d33529749916280fa241994ee375ce290
3
- size 741359744
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/optimizer.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:09076ea10a3d61e7c2e1df3f6041fa3881cd48d41cc55b7dd24e43bbfab79c04
3
- size 1482843147
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/rng_state.pth DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:cd28b96146b0c352b1c54b6d9faac47cc39eb31de8b2a770e08db7f2e62d4d5b
3
- size 14645
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/scheduler.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:7d1f43e93fb7b4c7684484d3ae04ff8f2520e0f67f3b6d0d2947cb202aa52ee2
3
- size 1465
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/tokenizer.json DELETED
The diff for this file is too large to render. See raw diff
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/tokenizer_config.json DELETED
@@ -1,23 +0,0 @@
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
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/trainer_state.json DELETED
@@ -1,41 +0,0 @@
1
- {
2
- "best_global_step": null,
3
- "best_metric": null,
4
- "best_model_checkpoint": null,
5
- "epoch": 0.6915629322268326,
6
- "eval_steps": 500,
7
- "global_step": 500,
8
- "is_hyper_param_search": false,
9
- "is_local_process_zero": true,
10
- "is_world_process_zero": true,
11
- "log_history": [
12
- {
13
- "epoch": 0.6915629322268326,
14
- "grad_norm": 0.06808845698833466,
15
- "learning_rate": 1.3098201936376211e-05,
16
- "loss": 0.012068235397338867,
17
- "step": 500
18
- }
19
- ],
20
- "logging_steps": 500,
21
- "max_steps": 1446,
22
- "num_input_tokens_seen": 0,
23
- "num_train_epochs": 2,
24
- "save_steps": 500,
25
- "stateful_callbacks": {
26
- "TrainerControl": {
27
- "args": {
28
- "should_epoch_stop": false,
29
- "should_evaluate": false,
30
- "should_log": false,
31
- "should_save": true,
32
- "should_training_stop": false
33
- },
34
- "attributes": {}
35
- }
36
- },
37
- "total_flos": 121295240486400.0,
38
- "train_batch_size": 16,
39
- "trial_name": null,
40
- "trial_params": null
41
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/checkpoint-500/training_args.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:7e609457181ac50440df7d6728c3518d3b9316719fead1acbcffdb0f4941dd06
3
- size 5201
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/config.json DELETED
@@ -1,63 +0,0 @@
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": "ADDRESS",
15
- "2": "CREDIT_CARD",
16
- "3": "CRYPTO_ADDRESS",
17
- "4": "EMAIL",
18
- "5": "IBAN",
19
- "6": "OFF_PLATFORM_ATTEMPT",
20
- "7": "PHONE",
21
- "8": "SOCIAL_MEDIA",
22
- "9": "URL"
23
- },
24
- "initializer_range": 0.02,
25
- "intermediate_size": 3072,
26
- "label2id": {
27
- "ADDRESS": 1,
28
- "CREDIT_CARD": 2,
29
- "CRYPTO_ADDRESS": 3,
30
- "EMAIL": 4,
31
- "IBAN": 5,
32
- "O": 0,
33
- "OFF_PLATFORM_ATTEMPT": 6,
34
- "PHONE": 7,
35
- "SOCIAL_MEDIA": 8,
36
- "URL": 9
37
- },
38
- "layer_norm_eps": 1e-07,
39
- "legacy": true,
40
- "max_position_embeddings": 512,
41
- "max_relative_positions": -1,
42
- "model_type": "deberta-v2",
43
- "norm_rel_ebd": "layer_norm",
44
- "num_attention_heads": 12,
45
- "num_hidden_layers": 12,
46
- "pad_token_id": 0,
47
- "pooler_dropout": 0,
48
- "pooler_hidden_act": "gelu",
49
- "pooler_hidden_size": 768,
50
- "pos_att_type": [
51
- "p2c",
52
- "c2p"
53
- ],
54
- "position_biased_input": false,
55
- "position_buckets": 256,
56
- "relative_attention": true,
57
- "share_att_key": true,
58
- "tie_word_embeddings": true,
59
- "transformers_version": "5.3.0",
60
- "type_vocab_size": 0,
61
- "use_cache": false,
62
- "vocab_size": 128100
63
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/pytorch_model.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:f9d7fb44d2ab60a24cf4a3e1dce9b3baf23b551a5b131bda58d827ac494d5b1e
3
- size 741421647
 
 
 
 
gorkem-pii-ready/SpanBased-CustomDeBERTa/tokenizer.json DELETED
The diff for this file is too large to render. See raw diff