Hugging Face's logo

Tokyosaurus
/
taglish-mbert-binary

Safetensors
bert
Model card Files
xet
 Community
Tokyosaurus commited on
Commit
4da0070
·
verified ·
1 Parent(s): 01beb6c

Upload folder using huggingface_hub

Browse files
Files changed (26) hide show
  1. checkpoint-340/config.json +43 -0
  2. checkpoint-340/model.safetensors +3 -0
  3. checkpoint-340/optimizer.pt +3 -0
  4. checkpoint-340/rng_state.pth +3 -0
  5. checkpoint-340/scaler.pt +3 -0
  6. checkpoint-340/scheduler.pt +3 -0
  7. checkpoint-340/tokenizer.json +0 -0
  8. checkpoint-340/tokenizer_config.json +14 -0
  9. checkpoint-340/trainer_state.json +139 -0
  10. checkpoint-340/training_args.bin +3 -0
  11. checkpoint-425/config.json +43 -0
  12. checkpoint-425/model.safetensors +3 -0
  13. checkpoint-425/optimizer.pt +3 -0
  14. checkpoint-425/rng_state.pth +3 -0
  15. checkpoint-425/scaler.pt +3 -0
  16. checkpoint-425/scheduler.pt +3 -0
  17. checkpoint-425/tokenizer.json +0 -0
  18. checkpoint-425/tokenizer_config.json +14 -0
  19. checkpoint-425/trainer_state.json +163 -0
  20. checkpoint-425/training_args.bin +3 -0
  21. comprehensive_results_binary.json +79 -0
  22. config.json +43 -0
  23. model.safetensors +3 -0
  24. tokenizer.json +0 -0
  25. tokenizer_config.json +14 -0
  26. training_args.bin +3 -0
checkpoint-340/config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_cross_attention": false,
3
+ "architectures": [
4
+ "BertForSequenceClassification"
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": "Non-Gaslighting",
17
+ "1": "Gaslighting"
18
+ },
19
+ "initializer_range": 0.02,
20
+ "intermediate_size": 3072,
21
+ "is_decoder": false,
22
+ "label2id": {
23
+ "Gaslighting": 1,
24
+ "Non-Gaslighting": 0
25
+ },
26
+ "layer_norm_eps": 1e-12,
27
+ "max_position_embeddings": 512,
28
+ "model_type": "bert",
29
+ "num_attention_heads": 12,
30
+ "num_hidden_layers": 12,
31
+ "pad_token_id": 0,
32
+ "pooler_fc_size": 768,
33
+ "pooler_num_attention_heads": 12,
34
+ "pooler_num_fc_layers": 3,
35
+ "pooler_size_per_head": 128,
36
+ "pooler_type": "first_token_transform",
37
+ "problem_type": "single_label_classification",
38
+ "tie_word_embeddings": true,
39
+ "transformers_version": "5.0.0",
40
+ "type_vocab_size": 2,
41
+ "use_cache": false,
42
+ "vocab_size": 105879
43
+ }
checkpoint-340/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f0085f3383719f007ef8fe7c40858ff5a2b5370a15802d468b4f45043e902fc
3
+ size 669455336
checkpoint-340/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:009acdd20bfce1e989ef058825aa64c4a7be626b6890484206fcc5cb2c838e95
3
+ size 1339032203
checkpoint-340/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:47fb45a7b2139ee3f5b084679b9edeba65d0dc98ebdb3f8c4b6e7947626ff113
3
+ size 14645
checkpoint-340/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:41d03862e2cc4e9749501d6c736e76da5165fe491f53b662caa2d77a3068a2b1
3
+ size 1383
checkpoint-340/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:81f9ccb04b432af0119e04f8e965bb80397941335599a88c3c3d75493d2a79ba
3
+ size 1465
checkpoint-340/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-340/tokenizer_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "cls_token": "[CLS]",
4
+ "do_lower_case": true,
5
+ "is_local": false,
6
+ "mask_token": "[MASK]",
7
+ "model_max_length": 512,
8
+ "pad_token": "[PAD]",
9
+ "sep_token": "[SEP]",
10
+ "strip_accents": null,
11
+ "tokenize_chinese_chars": true,
12
+ "tokenizer_class": "BertTokenizer",
13
+ "unk_token": "[UNK]"
14
+ }
checkpoint-340/trainer_state.json ADDED
@@ -0,0 +1,139 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 340,
3
+ "best_metric": 0.9236111111111112,
4
+ "best_model_checkpoint": "D:/Thesis/Taglish_Gaslighting_V3\\model_outputs\\mbert_binary\\checkpoint-340",
5
+ "epoch": 4.0,
6
+ "eval_steps": 500,
7
+ "global_step": 340,
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
+ "grad_norm": 13.322940826416016,
15
+ "learning_rate": 1.8010471204188483e-05,
16
+ "loss": 0.5703998341279871,
17
+ "step": 85
18
+ },
19
+ {
20
+ "epoch": 1.0,
21
+ "eval_accuracy": 0.8062283737024222,
22
+ "eval_f1": 0.8082191780821918,
23
+ "eval_loss": 0.4196648597717285,
24
+ "eval_macro_f1": 0.806207491138998,
25
+ "eval_non_gas_f1": 0.8041958041958042,
26
+ "eval_non_gas_precision": 0.8273381294964028,
27
+ "eval_non_gas_recall": 0.782312925170068,
28
+ "eval_precision": 0.7866666666666666,
29
+ "eval_recall": 0.8309859154929577,
30
+ "eval_roc_auc": 0.8890485771773498,
31
+ "eval_runtime": 13.237,
32
+ "eval_samples_per_second": 21.833,
33
+ "eval_steps_per_second": 1.435,
34
+ "step": 85
35
+ },
36
+ {
37
+ "epoch": 2.0,
38
+ "grad_norm": 0.4100542664527893,
39
+ "learning_rate": 1.356020942408377e-05,
40
+ "loss": 0.31691517549402576,
41
+ "step": 170
42
+ },
43
+ {
44
+ "epoch": 2.0,
45
+ "eval_accuracy": 0.903114186851211,
46
+ "eval_f1": 0.9014084507042254,
47
+ "eval_loss": 0.2633134126663208,
48
+ "eval_macro_f1": 0.9030851777330651,
49
+ "eval_non_gas_f1": 0.9047619047619048,
50
+ "eval_non_gas_precision": 0.9047619047619048,
51
+ "eval_non_gas_recall": 0.9047619047619048,
52
+ "eval_precision": 0.9014084507042254,
53
+ "eval_recall": 0.9014084507042254,
54
+ "eval_roc_auc": 0.9666570853693591,
55
+ "eval_runtime": 14.0294,
56
+ "eval_samples_per_second": 20.6,
57
+ "eval_steps_per_second": 1.354,
58
+ "step": 170
59
+ },
60
+ {
61
+ "epoch": 3.0,
62
+ "grad_norm": 0.385506808757782,
63
+ "learning_rate": 9.109947643979057e-06,
64
+ "loss": 0.15622444152832032,
65
+ "step": 255
66
+ },
67
+ {
68
+ "epoch": 3.0,
69
+ "eval_accuracy": 0.8961937716262975,
70
+ "eval_f1": 0.8913043478260869,
71
+ "eval_loss": 0.2805224657058716,
72
+ "eval_macro_f1": 0.8959832997408581,
73
+ "eval_non_gas_f1": 0.9006622516556292,
74
+ "eval_non_gas_precision": 0.8774193548387097,
75
+ "eval_non_gas_recall": 0.9251700680272109,
76
+ "eval_precision": 0.917910447761194,
77
+ "eval_recall": 0.8661971830985915,
78
+ "eval_roc_auc": 0.9686691578039667,
79
+ "eval_runtime": 13.2177,
80
+ "eval_samples_per_second": 21.865,
81
+ "eval_steps_per_second": 1.437,
82
+ "step": 255
83
+ },
84
+ {
85
+ "epoch": 4.0,
86
+ "grad_norm": 0.06804927438497543,
87
+ "learning_rate": 4.659685863874346e-06,
88
+ "loss": 0.09398165871115292,
89
+ "step": 340
90
+ },
91
+ {
92
+ "epoch": 4.0,
93
+ "eval_accuracy": 0.9238754325259516,
94
+ "eval_f1": 0.9236111111111112,
95
+ "eval_loss": 0.3195452392101288,
96
+ "eval_macro_f1": 0.923874521072797,
97
+ "eval_non_gas_f1": 0.9241379310344827,
98
+ "eval_non_gas_precision": 0.9370629370629371,
99
+ "eval_non_gas_recall": 0.9115646258503401,
100
+ "eval_precision": 0.910958904109589,
101
+ "eval_recall": 0.9366197183098591,
102
+ "eval_roc_auc": 0.9701542588866533,
103
+ "eval_runtime": 13.0948,
104
+ "eval_samples_per_second": 22.07,
105
+ "eval_steps_per_second": 1.451,
106
+ "step": 340
107
+ }
108
+ ],
109
+ "logging_steps": 500,
110
+ "max_steps": 425,
111
+ "num_input_tokens_seen": 0,
112
+ "num_train_epochs": 5,
113
+ "save_steps": 500,
114
+ "stateful_callbacks": {
115
+ "EarlyStoppingCallback": {
116
+ "args": {
117
+ "early_stopping_patience": 2,
118
+ "early_stopping_threshold": 0.0
119
+ },
120
+ "attributes": {
121
+ "early_stopping_patience_counter": 0
122
+ }
123
+ },
124
+ "TrainerControl": {
125
+ "args": {
126
+ "should_epoch_stop": false,
127
+ "should_evaluate": false,
128
+ "should_log": false,
129
+ "should_save": true,
130
+ "should_training_stop": false
131
+ },
132
+ "attributes": {}
133
+ }
134
+ },
135
+ "total_flos": 354673702625280.0,
136
+ "train_batch_size": 16,
137
+ "trial_name": null,
138
+ "trial_params": null
139
+ }
checkpoint-340/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:309858438c0025fd1aebdb6d019a6805f49bd4a7a22c10642321af45f7e715bb
3
+ size 5265
checkpoint-425/config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_cross_attention": false,
3
+ "architectures": [
4
+ "BertForSequenceClassification"
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": "Non-Gaslighting",
17
+ "1": "Gaslighting"
18
+ },
19
+ "initializer_range": 0.02,
20
+ "intermediate_size": 3072,
21
+ "is_decoder": false,
22
+ "label2id": {
23
+ "Gaslighting": 1,
24
+ "Non-Gaslighting": 0
25
+ },
26
+ "layer_norm_eps": 1e-12,
27
+ "max_position_embeddings": 512,
28
+ "model_type": "bert",
29
+ "num_attention_heads": 12,
30
+ "num_hidden_layers": 12,
31
+ "pad_token_id": 0,
32
+ "pooler_fc_size": 768,
33
+ "pooler_num_attention_heads": 12,
34
+ "pooler_num_fc_layers": 3,
35
+ "pooler_size_per_head": 128,
36
+ "pooler_type": "first_token_transform",
37
+ "problem_type": "single_label_classification",
38
+ "tie_word_embeddings": true,
39
+ "transformers_version": "5.0.0",
40
+ "type_vocab_size": 2,
41
+ "use_cache": false,
42
+ "vocab_size": 105879
43
+ }
checkpoint-425/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a112f9cc253eccd3e41e7fddc033c1a352eba4b43ccaf585a4d933e0e0081570
3
+ size 669455336
checkpoint-425/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c31218f116faf8bf1ccad2cfff8128c1966f78627b77acd2bf644c53a8141c30
3
+ size 1339032203
checkpoint-425/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:968d97a2add53d7d13b10ff013ed088813225c5be4cf09b17d8906eb2d21c161
3
+ size 14645
checkpoint-425/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:89602be0ec0530c0e15f95c976ba44f2e4bbbaf178676b868efef493d2e0581c
3
+ size 1383
checkpoint-425/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9be970c4e82c4cd2d70c0d430ca30097388f2071fe3ef9cd3c0fa569f26f3373
3
+ size 1465
checkpoint-425/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-425/tokenizer_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "cls_token": "[CLS]",
4
+ "do_lower_case": true,
5
+ "is_local": false,
6
+ "mask_token": "[MASK]",
7
+ "model_max_length": 512,
8
+ "pad_token": "[PAD]",
9
+ "sep_token": "[SEP]",
10
+ "strip_accents": null,
11
+ "tokenize_chinese_chars": true,
12
+ "tokenizer_class": "BertTokenizer",
13
+ "unk_token": "[UNK]"
14
+ }
checkpoint-425/trainer_state.json ADDED
@@ -0,0 +1,163 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 340,
3
+ "best_metric": 0.9236111111111112,
4
+ "best_model_checkpoint": "D:/Thesis/Taglish_Gaslighting_V3\\model_outputs\\mbert_binary\\checkpoint-340",
5
+ "epoch": 5.0,
6
+ "eval_steps": 500,
7
+ "global_step": 425,
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
+ "grad_norm": 13.322940826416016,
15
+ "learning_rate": 1.8010471204188483e-05,
16
+ "loss": 0.5703998341279871,
17
+ "step": 85
18
+ },
19
+ {
20
+ "epoch": 1.0,
21
+ "eval_accuracy": 0.8062283737024222,
22
+ "eval_f1": 0.8082191780821918,
23
+ "eval_loss": 0.4196648597717285,
24
+ "eval_macro_f1": 0.806207491138998,
25
+ "eval_non_gas_f1": 0.8041958041958042,
26
+ "eval_non_gas_precision": 0.8273381294964028,
27
+ "eval_non_gas_recall": 0.782312925170068,
28
+ "eval_precision": 0.7866666666666666,
29
+ "eval_recall": 0.8309859154929577,
30
+ "eval_roc_auc": 0.8890485771773498,
31
+ "eval_runtime": 13.237,
32
+ "eval_samples_per_second": 21.833,
33
+ "eval_steps_per_second": 1.435,
34
+ "step": 85
35
+ },
36
+ {
37
+ "epoch": 2.0,
38
+ "grad_norm": 0.4100542664527893,
39
+ "learning_rate": 1.356020942408377e-05,
40
+ "loss": 0.31691517549402576,
41
+ "step": 170
42
+ },
43
+ {
44
+ "epoch": 2.0,
45
+ "eval_accuracy": 0.903114186851211,
46
+ "eval_f1": 0.9014084507042254,
47
+ "eval_loss": 0.2633134126663208,
48
+ "eval_macro_f1": 0.9030851777330651,
49
+ "eval_non_gas_f1": 0.9047619047619048,
50
+ "eval_non_gas_precision": 0.9047619047619048,
51
+ "eval_non_gas_recall": 0.9047619047619048,
52
+ "eval_precision": 0.9014084507042254,
53
+ "eval_recall": 0.9014084507042254,
54
+ "eval_roc_auc": 0.9666570853693591,
55
+ "eval_runtime": 14.0294,
56
+ "eval_samples_per_second": 20.6,
57
+ "eval_steps_per_second": 1.354,
58
+ "step": 170
59
+ },
60
+ {
61
+ "epoch": 3.0,
62
+ "grad_norm": 0.385506808757782,
63
+ "learning_rate": 9.109947643979057e-06,
64
+ "loss": 0.15622444152832032,
65
+ "step": 255
66
+ },
67
+ {
68
+ "epoch": 3.0,
69
+ "eval_accuracy": 0.8961937716262975,
70
+ "eval_f1": 0.8913043478260869,
71
+ "eval_loss": 0.2805224657058716,
72
+ "eval_macro_f1": 0.8959832997408581,
73
+ "eval_non_gas_f1": 0.9006622516556292,
74
+ "eval_non_gas_precision": 0.8774193548387097,
75
+ "eval_non_gas_recall": 0.9251700680272109,
76
+ "eval_precision": 0.917910447761194,
77
+ "eval_recall": 0.8661971830985915,
78
+ "eval_roc_auc": 0.9686691578039667,
79
+ "eval_runtime": 13.2177,
80
+ "eval_samples_per_second": 21.865,
81
+ "eval_steps_per_second": 1.437,
82
+ "step": 255
83
+ },
84
+ {
85
+ "epoch": 4.0,
86
+ "grad_norm": 0.06804927438497543,
87
+ "learning_rate": 4.659685863874346e-06,
88
+ "loss": 0.09398165871115292,
89
+ "step": 340
90
+ },
91
+ {
92
+ "epoch": 4.0,
93
+ "eval_accuracy": 0.9238754325259516,
94
+ "eval_f1": 0.9236111111111112,
95
+ "eval_loss": 0.3195452392101288,
96
+ "eval_macro_f1": 0.923874521072797,
97
+ "eval_non_gas_f1": 0.9241379310344827,
98
+ "eval_non_gas_precision": 0.9370629370629371,
99
+ "eval_non_gas_recall": 0.9115646258503401,
100
+ "eval_precision": 0.910958904109589,
101
+ "eval_recall": 0.9366197183098591,
102
+ "eval_roc_auc": 0.9701542588866533,
103
+ "eval_runtime": 13.0948,
104
+ "eval_samples_per_second": 22.07,
105
+ "eval_steps_per_second": 1.451,
106
+ "step": 340
107
+ },
108
+ {
109
+ "epoch": 5.0,
110
+ "grad_norm": 0.07191683351993561,
111
+ "learning_rate": 2.0942408376963353e-07,
112
+ "loss": 0.05241534289191751,
113
+ "step": 425
114
+ },
115
+ {
116
+ "epoch": 5.0,
117
+ "eval_accuracy": 0.903114186851211,
118
+ "eval_f1": 0.8985507246376812,
119
+ "eval_loss": 0.38448667526245117,
120
+ "eval_macro_f1": 0.9029177464248008,
121
+ "eval_non_gas_f1": 0.9072847682119205,
122
+ "eval_non_gas_precision": 0.8838709677419355,
123
+ "eval_non_gas_recall": 0.9319727891156463,
124
+ "eval_precision": 0.9253731343283582,
125
+ "eval_recall": 0.8732394366197183,
126
+ "eval_roc_auc": 0.9692679888856952,
127
+ "eval_runtime": 12.9711,
128
+ "eval_samples_per_second": 22.28,
129
+ "eval_steps_per_second": 1.465,
130
+ "step": 425
131
+ }
132
+ ],
133
+ "logging_steps": 500,
134
+ "max_steps": 425,
135
+ "num_input_tokens_seen": 0,
136
+ "num_train_epochs": 5,
137
+ "save_steps": 500,
138
+ "stateful_callbacks": {
139
+ "EarlyStoppingCallback": {
140
+ "args": {
141
+ "early_stopping_patience": 2,
142
+ "early_stopping_threshold": 0.0
143
+ },
144
+ "attributes": {
145
+ "early_stopping_patience_counter": 1
146
+ }
147
+ },
148
+ "TrainerControl": {
149
+ "args": {
150
+ "should_epoch_stop": false,
151
+ "should_evaluate": false,
152
+ "should_log": false,
153
+ "should_save": true,
154
+ "should_training_stop": true
155
+ },
156
+ "attributes": {}
157
+ }
158
+ },
159
+ "total_flos": 443342128281600.0,
160
+ "train_batch_size": 16,
161
+ "trial_name": null,
162
+ "trial_params": null
163
+ }
checkpoint-425/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:309858438c0025fd1aebdb6d019a6805f49bd4a7a22c10642321af45f7e715bb
3
+ size 5265
comprehensive_results_binary.json ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "val": {
3
+ "loss": 0.3197322487831116,
4
+ "accuracy": 0.9238754325259516,
5
+ "precision": 0.910958904109589,
6
+ "recall": 0.9366197183098591,
7
+ "f1": 0.9236111111111112,
8
+ "macro_f1": 0.923874521072797,
9
+ "roc_auc": 0.9701782121299224,
10
+ "non_gas_precision": 0.9370629370629371,
11
+ "non_gas_recall": 0.9115646258503401,
12
+ "non_gas_f1": 0.9241379310344827,
13
+ "runtime": 12.9943,
14
+ "samples_per_second": 22.241,
15
+ "steps_per_second": 1.462,
16
+ "confusion_matrix": [
17
+ [
18
+ 134,
19
+ 13
20
+ ],
21
+ [
22
+ 9,
23
+ 133
24
+ ]
25
+ ]
26
+ },
27
+ "test_id": {
28
+ "loss": 0.336225688457489,
29
+ "accuracy": 0.9065743944636678,
30
+ "precision": 0.9020979020979021,
31
+ "recall": 0.9084507042253521,
32
+ "f1": 0.9052631578947369,
33
+ "macro_f1": 0.9065564936231363,
34
+ "roc_auc": 0.9741304972693302,
35
+ "non_gas_precision": 0.910958904109589,
36
+ "non_gas_recall": 0.9047619047619048,
37
+ "non_gas_f1": 0.9078498293515358,
38
+ "runtime": 13.0057,
39
+ "samples_per_second": 22.221,
40
+ "steps_per_second": 1.461,
41
+ "confusion_matrix": [
42
+ [
43
+ 133,
44
+ 14
45
+ ],
46
+ [
47
+ 13,
48
+ 129
49
+ ]
50
+ ]
51
+ },
52
+ "test_ood": {
53
+ "loss": 0.5277629494667053,
54
+ "accuracy": 0.8695652173913043,
55
+ "precision": 0.994579945799458,
56
+ "recall": 0.8398169336384439,
57
+ "f1": 0.9106699751861043,
58
+ "macro_f1": 0.8345296184655353,
59
+ "roc_auc": 0.9881404835339767,
60
+ "non_gas_precision": 0.6174863387978142,
61
+ "non_gas_recall": 0.9826086956521739,
62
+ "non_gas_f1": 0.7583892617449665,
63
+ "runtime": 13.2146,
64
+ "samples_per_second": 41.772,
65
+ "steps_per_second": 2.649,
66
+ "confusion_matrix": [
67
+ [
68
+ 113,
69
+ 2
70
+ ],
71
+ [
72
+ 70,
73
+ 367
74
+ ]
75
+ ]
76
+ },
77
+ "delta_f1": -0.005406817291367383,
78
+ "delta_f1_interpretation": "\u2705 Robust generalization (minimal domain sensitivity)"
79
+ }
config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_cross_attention": false,
3
+ "architectures": [
4
+ "BertForSequenceClassification"
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": "Non-Gaslighting",
17
+ "1": "Gaslighting"
18
+ },
19
+ "initializer_range": 0.02,
20
+ "intermediate_size": 3072,
21
+ "is_decoder": false,
22
+ "label2id": {
23
+ "Gaslighting": 1,
24
+ "Non-Gaslighting": 0
25
+ },
26
+ "layer_norm_eps": 1e-12,
27
+ "max_position_embeddings": 512,
28
+ "model_type": "bert",
29
+ "num_attention_heads": 12,
30
+ "num_hidden_layers": 12,
31
+ "pad_token_id": 0,
32
+ "pooler_fc_size": 768,
33
+ "pooler_num_attention_heads": 12,
34
+ "pooler_num_fc_layers": 3,
35
+ "pooler_size_per_head": 128,
36
+ "pooler_type": "first_token_transform",
37
+ "problem_type": "single_label_classification",
38
+ "tie_word_embeddings": true,
39
+ "transformers_version": "5.0.0",
40
+ "type_vocab_size": 2,
41
+ "use_cache": false,
42
+ "vocab_size": 105879
43
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e81ea3fdf2bdea9f09e3939e1ceeda7d1fd9fe807059ae3e491d660db73cb428
3
+ size 669455336
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "cls_token": "[CLS]",
4
+ "do_lower_case": true,
5
+ "is_local": false,
6
+ "mask_token": "[MASK]",
7
+ "model_max_length": 512,
8
+ "pad_token": "[PAD]",
9
+ "sep_token": "[SEP]",
10
+ "strip_accents": null,
11
+ "tokenize_chinese_chars": true,
12
+ "tokenizer_class": "BertTokenizer",
13
+ "unk_token": "[UNK]"
14
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:309858438c0025fd1aebdb6d019a6805f49bd4a7a22c10642321af45f7e715bb
3
+ size 5265