phantomkidding commited on
Commit
e77047e
·
verified ·
1 Parent(s): bcaac37

Add files using upload-large-folder tool

Browse files
Files changed (50) hide show
  1. special_tokens_map.json +37 -0
  2. tokenizer.json +0 -0
  3. tokenizer_config.json +56 -0
  4. toxicn/README.md +39 -0
  5. toxicn/expression_ie/0/checkpoint-266/config.json +45 -0
  6. toxicn/expression_ie/0/checkpoint-266/special_tokens_map.json +37 -0
  7. toxicn/expression_ie/0/checkpoint-266/tokenizer.json +0 -0
  8. toxicn/expression_ie/0/checkpoint-266/tokenizer_config.json +56 -0
  9. toxicn/expression_ie/0/checkpoint-266/trainer_state.json +181 -0
  10. toxicn/expression_ie/0/checkpoint-266/vocab.txt +0 -0
  11. toxicn/expression_ie/0/config.json +45 -0
  12. toxicn/expression_ie/0/special_tokens_map.json +37 -0
  13. toxicn/expression_ie/0/tokenizer.json +0 -0
  14. toxicn/expression_ie/0/tokenizer_config.json +56 -0
  15. toxicn/expression_ie/0/vocab.txt +0 -0
  16. toxicn/expression_ie/1/checkpoint-304/config.json +45 -0
  17. toxicn/expression_ie/1/checkpoint-304/special_tokens_map.json +37 -0
  18. toxicn/expression_ie/1/checkpoint-304/tokenizer.json +0 -0
  19. toxicn/expression_ie/1/checkpoint-304/tokenizer_config.json +56 -0
  20. toxicn/expression_ie/1/checkpoint-304/trainer_state.json +202 -0
  21. toxicn/expression_ie/1/checkpoint-304/vocab.txt +0 -0
  22. toxicn/expression_ie/1/config.json +45 -0
  23. toxicn/expression_ie/1/special_tokens_map.json +37 -0
  24. toxicn/expression_ie/1/tokenizer.json +0 -0
  25. toxicn/expression_ie/1/tokenizer_config.json +56 -0
  26. toxicn/expression_ie/1/vocab.txt +0 -0
  27. toxicn/expression_ie/2/checkpoint-152/config.json +45 -0
  28. toxicn/expression_ie/2/checkpoint-152/special_tokens_map.json +37 -0
  29. toxicn/expression_ie/2/checkpoint-152/tokenizer.json +0 -0
  30. toxicn/expression_ie/2/checkpoint-152/tokenizer_config.json +56 -0
  31. toxicn/expression_ie/2/checkpoint-152/trainer_state.json +118 -0
  32. toxicn/expression_ie/2/checkpoint-152/vocab.txt +0 -0
  33. toxicn/expression_ie/2/config.json +45 -0
  34. toxicn/expression_ie/2/special_tokens_map.json +37 -0
  35. toxicn/expression_ie/2/tokenizer.json +0 -0
  36. toxicn/expression_ie/2/tokenizer_config.json +56 -0
  37. toxicn/expression_ie/2/vocab.txt +0 -0
  38. toxicn/expression_ie/3/checkpoint-190/config.json +45 -0
  39. toxicn/expression_ie/3/checkpoint-190/special_tokens_map.json +37 -0
  40. toxicn/expression_ie/3/checkpoint-190/tokenizer.json +0 -0
  41. toxicn/expression_ie/3/checkpoint-190/tokenizer_config.json +56 -0
  42. toxicn/expression_ie/3/checkpoint-190/trainer_state.json +139 -0
  43. toxicn/expression_ie/3/checkpoint-190/vocab.txt +0 -0
  44. toxicn/expression_ie/3/config.json +45 -0
  45. toxicn/expression_ie/3/special_tokens_map.json +37 -0
  46. toxicn/expression_ie/3/tokenizer.json +0 -0
  47. toxicn/expression_ie/3/tokenizer_config.json +56 -0
  48. toxicn/expression_ie/3/vocab.txt +0 -0
  49. toxicn/expression_ie/4/config.json +45 -0
  50. vocab.txt +0 -0
special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 1000000000000000019884624838656,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
toxicn/README.md ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Performance Report
2
+ F1-macro (accuracy)
3
+ ## expression_ie
4
+ Combined implicit and reporting labels.
5
+ | | epoch | overall | non-hate | explicit | implicit |
6
+ |:---:|:-------:|:------------------:|:------------------:|:------------------:|:------------------:|
7
+ | 0 | 7 | 0.7110<br>(74.70%) | 0.8306<br>(83.10%) | 0.8090<br>(86.52%) | 0.7271<br>(79.78%) |
8
+ | 1 | 8 | 0.7105<br>(74.60%) | 0.8193<br>(82.01%) | 0.8031<br>(86.30%) | 0.7396<br>(80.89%) |
9
+ | 2 | 4 | 0.7091<br>(74.48%) | 0.8164<br>(81.72%) | 0.8151<br>(86.93%) | 0.7272<br>(80.31%) |
10
+ | 3 | 5 | 0.7096<br>(74.77%) | 0.8294<br>(83.01%) | 0.8075<br>(86.55%) | 0.7272<br>(79.98%) |
11
+ | 4 | 8 | 0.7219<br>(75.40%) | 0.8214<br>(82.22%) | 0.8194<br>(87.64%) | 0.7439<br>(80.93%) |
12
+ | avg | 6.4 | 0.7124<br>(74.79%) | 0.8234<br>(82.42%) | 0.8108<br>(86.79%) | 0.7330<br>(80.38%) |
13
+ ## expression
14
+ | | epoch | overall | non-hate | explicit | implicit | reporting |
15
+ |:---:|:-------:|:------------------:|:------------------:|:------------------:|:------------------:|:------------------:|
16
+ | 0 | 7 | 0.6425<br>(72.45%) | 0.8206<br>(82.19%) | 0.8090<br>(86.85%) | 0.6693<br>(81.94%) | 0.7762<br>(93.92%) |
17
+ | 1 | 5 | 0.6522<br>(73.52%) | 0.8224<br>(82.39%) | 0.8091<br>(86.18%) | 0.6797<br>(84.22%) | 0.7896<br>(94.25%) |
18
+ | 2 | 5 | 0.6442<br>(72.48%) | 0.8261<br>(82.68%) | 0.7969<br>(85.30%) | 0.6795<br>(83.39%) | 0.7789<br>(93.59%) |
19
+ | 3 | 7 | 0.6636<br>(73.90%) | 0.8270<br>(82.76%) | 0.8344<br>(88.13%) | 0.6797<br>(83.14%) | 0.7887<br>(93.76%) |
20
+ | 4 | 7 | 0.6455<br>(72.19%) | 0.8163<br>(81.72%) | 0.8006<br>(86.05%) | 0.6832<br>(82.89%) | 0.7795<br>(93.71%) |
21
+ | avg | 6.2 | 0.6496<br>(72.91%) | 0.8225<br>(82.35%) | 0.8100<br>(86.50%) | 0.6783<br>(83.12%) | 0.7826<br>(93.85%) |
22
+ ## toxic
23
+ | | epoch | overall | non-toxic | toxic |
24
+ |:---:|:-------:|:------------------:|:------------------:|:------------------:|
25
+ | 0 | 4 | 0.8292<br>(83.02%) | 0.8292<br>(83.02%) | 0.8292<br>(83.02%) |
26
+ | 1 | 6 | 0.8134<br>(81.43%) | 0.8134<br>(81.43%) | 0.8134<br>(81.43%) |
27
+ | 2 | 6 | 0.8182<br>(81.85%) | 0.8182<br>(81.85%) | 0.8182<br>(81.85%) |
28
+ | 3 | 5 | 0.8153<br>(81.56%) | 0.8153<br>(81.56%) | 0.8153<br>(81.56%) |
29
+ | 4 | 6 | 0.8175<br>(81.81%) | 0.8175<br>(81.81%) | 0.8175<br>(81.81%) |
30
+ | avg | 5.4 | 0.8187<br>(81.93%) | 0.8187<br>(81.93%) | 0.8187<br>(81.93%) |
31
+ ## toxic_type
32
+ | | epoch | overall | non-toxic | general | hate |
33
+ |:---:|:-------:|:------------------:|:------------------:|:------------------:|:------------------:|
34
+ | 0 | 7 | 0.7159<br>(79.40%) | 0.8201<br>(82.15%) | 0.7424<br>(93.47%) | 0.8315<br>(83.19%) |
35
+ | 1 | 6 | 0.6990<br>(79.27%) | 0.8152<br>(81.56%) | 0.7204<br>(93.96%) | 0.8293<br>(83.01%) |
36
+ | 2 | 5 | 0.7240<br>(80.52%) | 0.8312<br>(83.18%) | 0.7448<br>(93.84%) | 0.8393<br>(84.01%) |
37
+ | 3 | 6 | 0.7207<br>(79.48%) | 0.8219<br>(82.22%) | 0.7527<br>(94.05%) | 0.8255<br>(82.68%) |
38
+ | 4 | 5 | 0.7265<br>(78.89%) | 0.8133<br>(81.39%) | 0.7700<br>(94.30%) | 0.8199<br>(82.10%) |
39
+ | avg | 5.8 | 0.7172<br>(79.51%) | 0.8204<br>(82.10%) | 0.7461<br>(93.92%) | 0.8291<br>(83.00%) |
toxicn/expression_ie/0/checkpoint-266/config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "id2label": {
14
+ "0": "non-hate",
15
+ "1": "explicit",
16
+ "2": "implicit"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 4096,
20
+ "label2id": {
21
+ "explicit": 1,
22
+ "implicit": 2,
23
+ "non-hate": 0
24
+ },
25
+ "layer_norm_eps": 1e-12,
26
+ "max_position_embeddings": 512,
27
+ "model_type": "bert",
28
+ "note": "Combined implicit and reporting labels.",
29
+ "num_attention_heads": 16,
30
+ "num_hidden_layers": 24,
31
+ "output_past": true,
32
+ "pad_token_id": 0,
33
+ "pooler_fc_size": 768,
34
+ "pooler_num_attention_heads": 12,
35
+ "pooler_num_fc_layers": 3,
36
+ "pooler_size_per_head": 128,
37
+ "pooler_type": "first_token_transform",
38
+ "position_embedding_type": "absolute",
39
+ "problem_type": "single_label_classification",
40
+ "torch_dtype": "float32",
41
+ "transformers_version": "4.51.3",
42
+ "type_vocab_size": 2,
43
+ "use_cache": true,
44
+ "vocab_size": 21128
45
+ }
toxicn/expression_ie/0/checkpoint-266/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
toxicn/expression_ie/0/checkpoint-266/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/0/checkpoint-266/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 1000000000000000019884624838656,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
toxicn/expression_ie/0/checkpoint-266/trainer_state.json ADDED
@@ -0,0 +1,181 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": 0.7109539047628343,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 7.0,
6
+ "eval_steps": 500,
7
+ "global_step": 266,
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_accuracy": 0.6716604244694132,
15
+ "eval_explicit_accuracy": 0.8056595921764461,
16
+ "eval_explicit_f1": 0.5639589169000934,
17
+ "eval_explicit_f1_macro": 0.7194627248489758,
18
+ "eval_f1_macro": 0.6113793084743014,
19
+ "eval_implicit_accuracy": 0.7677902621722846,
20
+ "eval_implicit_f1": 0.47752808988764045,
21
+ "eval_implicit_f1_macro": 0.6641252006420546,
22
+ "eval_loss": 0.7282944917678833,
23
+ "eval_non-hate_accuracy": 0.7698709945900957,
24
+ "eval_non-hate_f1": 0.7926509186351706,
25
+ "eval_non-hate_f1_macro": 0.7670594471623726,
26
+ "eval_precision_macro": 0.6199485971322883,
27
+ "eval_recall_macro": 0.6065545570426408,
28
+ "eval_runtime": 11.7325,
29
+ "eval_samples_per_second": 204.816,
30
+ "eval_steps_per_second": 0.256,
31
+ "step": 38
32
+ },
33
+ {
34
+ "epoch": 2.0,
35
+ "eval_accuracy": 0.7020391177694548,
36
+ "eval_explicit_accuracy": 0.8456096545984186,
37
+ "eval_explicit_f1": 0.6032085561497326,
38
+ "eval_explicit_f1_macro": 0.753683844078483,
39
+ "eval_f1_macro": 0.631743065413863,
40
+ "eval_implicit_accuracy": 0.7806908031627132,
41
+ "eval_implicit_f1": 0.47873392680514343,
42
+ "eval_implicit_f1_macro": 0.6699334983169327,
43
+ "eval_loss": 0.6848538517951965,
44
+ "eval_non-hate_accuracy": 0.7777777777777778,
45
+ "eval_non-hate_f1": 0.8132867132867133,
46
+ "eval_non-hate_f1_macro": 0.7694388345467482,
47
+ "eval_precision_macro": 0.6749213890273745,
48
+ "eval_recall_macro": 0.6146648724635333,
49
+ "eval_runtime": 11.5543,
50
+ "eval_samples_per_second": 207.975,
51
+ "eval_steps_per_second": 0.26,
52
+ "step": 76
53
+ },
54
+ {
55
+ "epoch": 3.0,
56
+ "eval_accuracy": 0.7369954223886808,
57
+ "eval_explicit_accuracy": 0.8605909280066584,
58
+ "eval_explicit_f1": 0.6946216955332726,
59
+ "eval_explicit_f1_macro": 0.8021504271680922,
60
+ "eval_f1_macro": 0.6829045890481753,
61
+ "eval_implicit_accuracy": 0.8031627132750728,
62
+ "eval_implicit_f1": 0.5217391304347826,
63
+ "eval_implicit_f1_macro": 0.6989099110387169,
64
+ "eval_loss": 0.6469447612762451,
65
+ "eval_non-hate_accuracy": 0.8102372034956304,
66
+ "eval_non-hate_f1": 0.8323529411764706,
67
+ "eval_non-hate_f1_macro": 0.8068763747109582,
68
+ "eval_precision_macro": 0.7030637504984988,
69
+ "eval_recall_macro": 0.6759524893279935,
70
+ "eval_runtime": 11.6013,
71
+ "eval_samples_per_second": 207.132,
72
+ "eval_steps_per_second": 0.259,
73
+ "step": 114
74
+ },
75
+ {
76
+ "epoch": 4.0,
77
+ "eval_accuracy": 0.7469829379941739,
78
+ "eval_explicit_accuracy": 0.8622555139409072,
79
+ "eval_explicit_f1": 0.6982680036463081,
80
+ "eval_explicit_f1_macro": 0.8045128101272792,
81
+ "eval_f1_macro": 0.7064782941624405,
82
+ "eval_implicit_accuracy": 0.8027465667915106,
83
+ "eval_implicit_f1": 0.5805309734513274,
84
+ "eval_implicit_f1_macro": 0.7257932342773503,
85
+ "eval_loss": 0.654564380645752,
86
+ "eval_non-hate_accuracy": 0.8289637952559301,
87
+ "eval_non-hate_f1": 0.8406359053896859,
88
+ "eval_non-hate_f1_macro": 0.8280413473962349,
89
+ "eval_precision_macro": 0.7085777694030945,
90
+ "eval_recall_macro": 0.7047708072087131,
91
+ "eval_runtime": 11.7024,
92
+ "eval_samples_per_second": 205.342,
93
+ "eval_steps_per_second": 0.256,
94
+ "step": 152
95
+ },
96
+ {
97
+ "epoch": 5.0,
98
+ "eval_accuracy": 0.7478152309612984,
99
+ "eval_explicit_accuracy": 0.8605909280066584,
100
+ "eval_explicit_f1": 0.7129391602399314,
101
+ "eval_explicit_f1_macro": 0.8104404512384049,
102
+ "eval_f1_macro": 0.704145748753633,
103
+ "eval_implicit_accuracy": 0.807324178110695,
104
+ "eval_implicit_f1": 0.5594671741198858,
105
+ "eval_implicit_f1_macro": 0.7180824552357086,
106
+ "eval_loss": 0.703403651714325,
107
+ "eval_non-hate_accuracy": 0.8277153558052435,
108
+ "eval_non-hate_f1": 0.8400309119010819,
109
+ "eval_non-hate_f1_macro": 0.8266881340389088,
110
+ "eval_precision_macro": 0.7082762214834658,
111
+ "eval_recall_macro": 0.7061221314459502,
112
+ "eval_runtime": 11.6056,
113
+ "eval_samples_per_second": 207.056,
114
+ "eval_steps_per_second": 0.258,
115
+ "step": 190
116
+ },
117
+ {
118
+ "epoch": 6.0,
119
+ "eval_accuracy": 0.7453183520599251,
120
+ "eval_explicit_accuracy": 0.8614232209737828,
121
+ "eval_explicit_f1": 0.7050487156775908,
122
+ "eval_explicit_f1_macro": 0.8072428783718386,
123
+ "eval_f1_macro": 0.7084361126403143,
124
+ "eval_implicit_accuracy": 0.8023304203079484,
125
+ "eval_implicit_f1": 0.584426946631671,
126
+ "eval_implicit_f1_macro": 0.7273759084782707,
127
+ "eval_loss": 0.749207079410553,
128
+ "eval_non-hate_accuracy": 0.826883062838119,
129
+ "eval_non-hate_f1": 0.8358326756116812,
130
+ "eval_non-hate_f1_macro": 0.8263670420311927,
131
+ "eval_precision_macro": 0.7066026859513975,
132
+ "eval_recall_macro": 0.7107961576220431,
133
+ "eval_runtime": 11.327,
134
+ "eval_samples_per_second": 212.149,
135
+ "eval_steps_per_second": 0.265,
136
+ "step": 228
137
+ },
138
+ {
139
+ "epoch": 7.0,
140
+ "eval_accuracy": 0.7469829379941739,
141
+ "eval_explicit_accuracy": 0.8651685393258427,
142
+ "eval_explicit_f1": 0.7054545454545454,
143
+ "eval_explicit_f1_macro": 0.8090143747240348,
144
+ "eval_f1_macro": 0.7109539047628343,
145
+ "eval_implicit_accuracy": 0.797752808988764,
146
+ "eval_implicit_f1": 0.588135593220339,
147
+ "eval_implicit_f1_macro": 0.7270518010227454,
148
+ "eval_loss": 0.8124973177909851,
149
+ "eval_non-hate_accuracy": 0.8310445276737412,
150
+ "eval_non-hate_f1": 0.8392715756136183,
151
+ "eval_non-hate_f1_macro": 0.8306007000875109,
152
+ "eval_precision_macro": 0.709614249840807,
153
+ "eval_recall_macro": 0.7124334807287682,
154
+ "eval_runtime": 11.5245,
155
+ "eval_samples_per_second": 208.512,
156
+ "eval_steps_per_second": 0.26,
157
+ "step": 266
158
+ }
159
+ ],
160
+ "logging_steps": 500,
161
+ "max_steps": 380,
162
+ "num_input_tokens_seen": 0,
163
+ "num_train_epochs": 10,
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": 6.133158586208606e+16,
178
+ "train_batch_size": 256,
179
+ "trial_name": null,
180
+ "trial_params": null
181
+ }
toxicn/expression_ie/0/checkpoint-266/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/0/config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "id2label": {
14
+ "0": "non-hate",
15
+ "1": "explicit",
16
+ "2": "implicit"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 4096,
20
+ "label2id": {
21
+ "explicit": 1,
22
+ "implicit": 2,
23
+ "non-hate": 0
24
+ },
25
+ "layer_norm_eps": 1e-12,
26
+ "max_position_embeddings": 512,
27
+ "model_type": "bert",
28
+ "note": "Combined implicit and reporting labels.",
29
+ "num_attention_heads": 16,
30
+ "num_hidden_layers": 24,
31
+ "output_past": true,
32
+ "pad_token_id": 0,
33
+ "pooler_fc_size": 768,
34
+ "pooler_num_attention_heads": 12,
35
+ "pooler_num_fc_layers": 3,
36
+ "pooler_size_per_head": 128,
37
+ "pooler_type": "first_token_transform",
38
+ "position_embedding_type": "absolute",
39
+ "problem_type": "single_label_classification",
40
+ "torch_dtype": "float32",
41
+ "transformers_version": "4.51.3",
42
+ "type_vocab_size": 2,
43
+ "use_cache": true,
44
+ "vocab_size": 21128
45
+ }
toxicn/expression_ie/0/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
toxicn/expression_ie/0/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/0/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 1000000000000000019884624838656,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
toxicn/expression_ie/0/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/1/checkpoint-304/config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "id2label": {
14
+ "0": "non-hate",
15
+ "1": "explicit",
16
+ "2": "implicit"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 4096,
20
+ "label2id": {
21
+ "explicit": 1,
22
+ "implicit": 2,
23
+ "non-hate": 0
24
+ },
25
+ "layer_norm_eps": 1e-12,
26
+ "max_position_embeddings": 512,
27
+ "model_type": "bert",
28
+ "note": "Combined implicit and reporting labels.",
29
+ "num_attention_heads": 16,
30
+ "num_hidden_layers": 24,
31
+ "output_past": true,
32
+ "pad_token_id": 0,
33
+ "pooler_fc_size": 768,
34
+ "pooler_num_attention_heads": 12,
35
+ "pooler_num_fc_layers": 3,
36
+ "pooler_size_per_head": 128,
37
+ "pooler_type": "first_token_transform",
38
+ "position_embedding_type": "absolute",
39
+ "problem_type": "single_label_classification",
40
+ "torch_dtype": "float32",
41
+ "transformers_version": "4.51.3",
42
+ "type_vocab_size": 2,
43
+ "use_cache": true,
44
+ "vocab_size": 21128
45
+ }
toxicn/expression_ie/1/checkpoint-304/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
toxicn/expression_ie/1/checkpoint-304/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/1/checkpoint-304/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 1000000000000000019884624838656,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
toxicn/expression_ie/1/checkpoint-304/trainer_state.json ADDED
@@ -0,0 +1,202 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": 0.7104553736043854,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 8.0,
6
+ "eval_steps": 500,
7
+ "global_step": 304,
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_accuracy": 0.664029975020816,
15
+ "eval_explicit_accuracy": 0.8193172356369692,
16
+ "eval_explicit_f1": 0.5616161616161616,
17
+ "eval_explicit_f1_macro": 0.7239124331940273,
18
+ "eval_f1_macro": 0.5929869945225031,
19
+ "eval_implicit_accuracy": 0.7518734388009992,
20
+ "eval_implicit_f1": 0.42802303262955854,
21
+ "eval_implicit_f1_macro": 0.6347983318384369,
22
+ "eval_loss": 0.741411566734314,
23
+ "eval_non-hate_accuracy": 0.7568692756036636,
24
+ "eval_non-hate_f1": 0.7893217893217893,
25
+ "eval_non-hate_f1_macro": 0.7509601072593198,
26
+ "eval_precision_macro": 0.6140473306672011,
27
+ "eval_recall_macro": 0.5835918257416923,
28
+ "eval_runtime": 11.3831,
29
+ "eval_samples_per_second": 211.015,
30
+ "eval_steps_per_second": 0.264,
31
+ "step": 38
32
+ },
33
+ {
34
+ "epoch": 2.0,
35
+ "eval_accuracy": 0.7127393838467944,
36
+ "eval_explicit_accuracy": 0.8422148209825145,
37
+ "eval_explicit_f1": 0.633816425120773,
38
+ "eval_explicit_f1_macro": 0.7666296240753772,
39
+ "eval_f1_macro": 0.6639606192636317,
40
+ "eval_implicit_accuracy": 0.7847626977518735,
41
+ "eval_implicit_f1": 0.541259982253771,
42
+ "eval_implicit_f1_macro": 0.7003281145971059,
43
+ "eval_loss": 0.6645469069480896,
44
+ "eval_non-hate_accuracy": 0.7985012489592007,
45
+ "eval_non-hate_f1": 0.8168054504163512,
46
+ "eval_non-hate_f1_macro": 0.7964693302035503,
47
+ "eval_precision_macro": 0.6733102399957404,
48
+ "eval_recall_macro": 0.6570978123146175,
49
+ "eval_runtime": 11.287,
50
+ "eval_samples_per_second": 212.812,
51
+ "eval_steps_per_second": 0.266,
52
+ "step": 76
53
+ },
54
+ {
55
+ "epoch": 3.0,
56
+ "eval_accuracy": 0.7352206494587843,
57
+ "eval_explicit_accuracy": 0.8513738551207327,
58
+ "eval_explicit_f1": 0.6792452830188679,
59
+ "eval_explicit_f1_macro": 0.7912617636985426,
60
+ "eval_f1_macro": 0.688805197580855,
61
+ "eval_implicit_accuracy": 0.8118234804329725,
62
+ "eval_implicit_f1": 0.5611650485436893,
63
+ "eval_implicit_f1_macro": 0.7206991114472554,
64
+ "eval_loss": 0.6540467739105225,
65
+ "eval_non-hate_accuracy": 0.8072439633638634,
66
+ "eval_non-hate_f1": 0.8260052611800075,
67
+ "eval_non-hate_f1_macro": 0.8049764989987764,
68
+ "eval_precision_macro": 0.7015735173994445,
69
+ "eval_recall_macro": 0.683640208447628,
70
+ "eval_runtime": 11.3168,
71
+ "eval_samples_per_second": 212.25,
72
+ "eval_steps_per_second": 0.265,
73
+ "step": 114
74
+ },
75
+ {
76
+ "epoch": 4.0,
77
+ "eval_accuracy": 0.7447960033305578,
78
+ "eval_explicit_accuracy": 0.8572023313905079,
79
+ "eval_explicit_f1": 0.6779342723004694,
80
+ "eval_explicit_f1_macro": 0.7930992570381727,
81
+ "eval_f1_macro": 0.70043908459981,
82
+ "eval_implicit_accuracy": 0.8155703580349709,
83
+ "eval_implicit_f1": 0.5886722376973074,
84
+ "eval_implicit_f1_macro": 0.7349049409575885,
85
+ "eval_loss": 0.6789953112602234,
86
+ "eval_non-hate_accuracy": 0.8168193172356369,
87
+ "eval_non-hate_f1": 0.8347107438016529,
88
+ "eval_non-hate_f1_macro": 0.8146476221342531,
89
+ "eval_precision_macro": 0.7123904162781053,
90
+ "eval_recall_macro": 0.6924595106095306,
91
+ "eval_runtime": 11.1541,
92
+ "eval_samples_per_second": 215.347,
93
+ "eval_steps_per_second": 0.269,
94
+ "step": 152
95
+ },
96
+ {
97
+ "epoch": 5.0,
98
+ "eval_accuracy": 0.7485428809325562,
99
+ "eval_explicit_accuracy": 0.8601165695253955,
100
+ "eval_explicit_f1": 0.691743119266055,
101
+ "eval_explicit_f1_macro": 0.8006373108446592,
102
+ "eval_f1_macro": 0.7089311427925056,
103
+ "eval_implicit_accuracy": 0.8184845961698585,
104
+ "eval_implicit_f1": 0.6014625228519196,
105
+ "eval_implicit_f1_macro": 0.7419711536092481,
106
+ "eval_loss": 0.7061896324157715,
107
+ "eval_non-hate_accuracy": 0.8184845961698585,
108
+ "eval_non-hate_f1": 0.833587786259542,
109
+ "eval_non-hate_f1_macro": 0.8169770433129212,
110
+ "eval_precision_macro": 0.7158531747405051,
111
+ "eval_recall_macro": 0.7041073908112286,
112
+ "eval_runtime": 11.4298,
113
+ "eval_samples_per_second": 210.152,
114
+ "eval_steps_per_second": 0.262,
115
+ "step": 190
116
+ },
117
+ {
118
+ "epoch": 6.0,
119
+ "eval_accuracy": 0.7447960033305578,
120
+ "eval_explicit_accuracy": 0.858451290591174,
121
+ "eval_explicit_f1": 0.6660117878192534,
122
+ "eval_explicit_f1_macro": 0.7881036223829494,
123
+ "eval_f1_macro": 0.70072253061642,
124
+ "eval_implicit_accuracy": 0.8143213988343048,
125
+ "eval_implicit_f1": 0.6010733452593918,
126
+ "eval_implicit_f1_macro": 0.740037486520092,
127
+ "eval_loss": 0.7737213969230652,
128
+ "eval_non-hate_accuracy": 0.8168193172356369,
129
+ "eval_non-hate_f1": 0.8350824587706147,
130
+ "eval_non-hate_f1_macro": 0.8145449747036595,
131
+ "eval_precision_macro": 0.7150590676051517,
132
+ "eval_recall_macro": 0.6907206131246949,
133
+ "eval_runtime": 11.3014,
134
+ "eval_samples_per_second": 212.541,
135
+ "eval_steps_per_second": 0.265,
136
+ "step": 228
137
+ },
138
+ {
139
+ "epoch": 7.0,
140
+ "eval_accuracy": 0.7464612822647794,
141
+ "eval_explicit_accuracy": 0.8588676103247294,
142
+ "eval_explicit_f1": 0.6864014801110083,
143
+ "eval_explicit_f1_macro": 0.7976729398943438,
144
+ "eval_f1_macro": 0.7080506880983318,
145
+ "eval_implicit_accuracy": 0.8143213988343048,
146
+ "eval_implicit_f1": 0.6046099290780141,
147
+ "eval_implicit_f1_macro": 0.741641199577092,
148
+ "eval_loss": 0.818276047706604,
149
+ "eval_non-hate_accuracy": 0.8197335553705246,
150
+ "eval_non-hate_f1": 0.833140655105973,
151
+ "eval_non-hate_f1_macro": 0.8185621790695099,
152
+ "eval_precision_macro": 0.7123330442356773,
153
+ "eval_recall_macro": 0.7044436007488328,
154
+ "eval_runtime": 11.2527,
155
+ "eval_samples_per_second": 213.459,
156
+ "eval_steps_per_second": 0.267,
157
+ "step": 266
158
+ },
159
+ {
160
+ "epoch": 8.0,
161
+ "eval_accuracy": 0.746044962531224,
162
+ "eval_explicit_accuracy": 0.8630308076602831,
163
+ "eval_explicit_f1": 0.6945218198700093,
164
+ "eval_explicit_f1_macro": 0.8031235340294506,
165
+ "eval_f1_macro": 0.7104553736043854,
166
+ "eval_implicit_accuracy": 0.8089092422980849,
167
+ "eval_implicit_f1": 0.6053310404127257,
168
+ "eval_implicit_f1_macro": 0.7396333861772499,
169
+ "eval_loss": 0.8596600294113159,
170
+ "eval_non-hate_accuracy": 0.8201498751040799,
171
+ "eval_non-hate_f1": 0.8315132605304212,
172
+ "eval_non-hate_f1_macro": 0.8193280588366392,
173
+ "eval_precision_macro": 0.7124096185141853,
174
+ "eval_recall_macro": 0.7086441390890778,
175
+ "eval_runtime": 11.3099,
176
+ "eval_samples_per_second": 212.38,
177
+ "eval_steps_per_second": 0.265,
178
+ "step": 304
179
+ }
180
+ ],
181
+ "logging_steps": 500,
182
+ "max_steps": 380,
183
+ "num_input_tokens_seen": 0,
184
+ "num_train_epochs": 10,
185
+ "save_steps": 500,
186
+ "stateful_callbacks": {
187
+ "TrainerControl": {
188
+ "args": {
189
+ "should_epoch_stop": false,
190
+ "should_evaluate": false,
191
+ "should_log": false,
192
+ "should_save": true,
193
+ "should_training_stop": false
194
+ },
195
+ "attributes": {}
196
+ }
197
+ },
198
+ "total_flos": 7.03803787453139e+16,
199
+ "train_batch_size": 256,
200
+ "trial_name": null,
201
+ "trial_params": null
202
+ }
toxicn/expression_ie/1/checkpoint-304/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/1/config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "id2label": {
14
+ "0": "non-hate",
15
+ "1": "explicit",
16
+ "2": "implicit"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 4096,
20
+ "label2id": {
21
+ "explicit": 1,
22
+ "implicit": 2,
23
+ "non-hate": 0
24
+ },
25
+ "layer_norm_eps": 1e-12,
26
+ "max_position_embeddings": 512,
27
+ "model_type": "bert",
28
+ "note": "Combined implicit and reporting labels.",
29
+ "num_attention_heads": 16,
30
+ "num_hidden_layers": 24,
31
+ "output_past": true,
32
+ "pad_token_id": 0,
33
+ "pooler_fc_size": 768,
34
+ "pooler_num_attention_heads": 12,
35
+ "pooler_num_fc_layers": 3,
36
+ "pooler_size_per_head": 128,
37
+ "pooler_type": "first_token_transform",
38
+ "position_embedding_type": "absolute",
39
+ "problem_type": "single_label_classification",
40
+ "torch_dtype": "float32",
41
+ "transformers_version": "4.51.3",
42
+ "type_vocab_size": 2,
43
+ "use_cache": true,
44
+ "vocab_size": 21128
45
+ }
toxicn/expression_ie/1/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
toxicn/expression_ie/1/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/1/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 1000000000000000019884624838656,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
toxicn/expression_ie/1/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/2/checkpoint-152/config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "id2label": {
14
+ "0": "non-hate",
15
+ "1": "explicit",
16
+ "2": "implicit"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 4096,
20
+ "label2id": {
21
+ "explicit": 1,
22
+ "implicit": 2,
23
+ "non-hate": 0
24
+ },
25
+ "layer_norm_eps": 1e-12,
26
+ "max_position_embeddings": 512,
27
+ "model_type": "bert",
28
+ "note": "Combined implicit and reporting labels.",
29
+ "num_attention_heads": 16,
30
+ "num_hidden_layers": 24,
31
+ "output_past": true,
32
+ "pad_token_id": 0,
33
+ "pooler_fc_size": 768,
34
+ "pooler_num_attention_heads": 12,
35
+ "pooler_num_fc_layers": 3,
36
+ "pooler_size_per_head": 128,
37
+ "pooler_type": "first_token_transform",
38
+ "position_embedding_type": "absolute",
39
+ "problem_type": "single_label_classification",
40
+ "torch_dtype": "float32",
41
+ "transformers_version": "4.51.3",
42
+ "type_vocab_size": 2,
43
+ "use_cache": true,
44
+ "vocab_size": 21128
45
+ }
toxicn/expression_ie/2/checkpoint-152/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
toxicn/expression_ie/2/checkpoint-152/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/2/checkpoint-152/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 1000000000000000019884624838656,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
toxicn/expression_ie/2/checkpoint-152/trainer_state.json ADDED
@@ -0,0 +1,118 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": 0.7091022272313756,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 4.0,
6
+ "eval_steps": 500,
7
+ "global_step": 152,
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_accuracy": 0.6736053288925895,
15
+ "eval_explicit_accuracy": 0.8288925895087427,
16
+ "eval_explicit_f1": 0.5585392051557465,
17
+ "eval_explicit_f1_macro": 0.726209958890809,
18
+ "eval_f1_macro": 0.6026805494376805,
19
+ "eval_implicit_accuracy": 0.7560366361365529,
20
+ "eval_implicit_f1": 0.4533582089552239,
21
+ "eval_implicit_f1_macro": 0.6481689222696806,
22
+ "eval_loss": 0.7294162511825562,
23
+ "eval_non-hate_accuracy": 0.7622814321398834,
24
+ "eval_non-hate_f1": 0.7961442342020707,
25
+ "eval_non-hate_f1_macro": 0.755535921394595,
26
+ "eval_precision_macro": 0.634237914188363,
27
+ "eval_recall_macro": 0.589576479756159,
28
+ "eval_runtime": 11.2451,
29
+ "eval_samples_per_second": 213.603,
30
+ "eval_steps_per_second": 0.267,
31
+ "step": 38
32
+ },
33
+ {
34
+ "epoch": 2.0,
35
+ "eval_accuracy": 0.6998334721065779,
36
+ "eval_explicit_accuracy": 0.8517901748542881,
37
+ "eval_explicit_f1": 0.6,
38
+ "eval_explicit_f1_macro": 0.7545222278998467,
39
+ "eval_f1_macro": 0.6273678361411024,
40
+ "eval_implicit_accuracy": 0.7835137385512073,
41
+ "eval_implicit_f1": 0.47580645161290325,
42
+ "eval_implicit_f1_macro": 0.6696975594895576,
43
+ "eval_loss": 0.698452353477478,
44
+ "eval_non-hate_accuracy": 0.7643630308076603,
45
+ "eval_non-hate_f1": 0.8062970568104039,
46
+ "eval_non-hate_f1_macro": 0.7527765836655633,
47
+ "eval_precision_macro": 0.6894692529837108,
48
+ "eval_recall_macro": 0.6063294726478178,
49
+ "eval_runtime": 11.1108,
50
+ "eval_samples_per_second": 216.186,
51
+ "eval_steps_per_second": 0.27,
52
+ "step": 76
53
+ },
54
+ {
55
+ "epoch": 3.0,
56
+ "eval_accuracy": 0.7318900915903414,
57
+ "eval_explicit_accuracy": 0.8621981681931724,
58
+ "eval_explicit_f1": 0.6789524733268671,
59
+ "eval_explicit_f1_macro": 0.7956119376970938,
60
+ "eval_f1_macro": 0.6935716066296994,
61
+ "eval_implicit_accuracy": 0.7918401332223147,
62
+ "eval_implicit_f1": 0.5784148397976391,
63
+ "eval_implicit_f1_macro": 0.7201084702028548,
64
+ "eval_loss": 0.6404813528060913,
65
+ "eval_non-hate_accuracy": 0.8097418817651957,
66
+ "eval_non-hate_f1": 0.8233475067645922,
67
+ "eval_non-hate_f1_macro": 0.8086065454436403,
68
+ "eval_precision_macro": 0.7005078578915928,
69
+ "eval_recall_macro": 0.688602422743145,
70
+ "eval_runtime": 11.1676,
71
+ "eval_samples_per_second": 215.087,
72
+ "eval_steps_per_second": 0.269,
73
+ "step": 114
74
+ },
75
+ {
76
+ "epoch": 4.0,
77
+ "eval_accuracy": 0.7447960033305578,
78
+ "eval_explicit_accuracy": 0.8692756036636137,
79
+ "eval_explicit_f1": 0.7150635208711433,
80
+ "eval_explicit_f1_macro": 0.8151222520617196,
81
+ "eval_f1_macro": 0.7091022272313756,
82
+ "eval_implicit_accuracy": 0.8030807660283097,
83
+ "eval_implicit_f1": 0.5832599118942732,
84
+ "eval_implicit_f1_macro": 0.7271709752984585,
85
+ "eval_loss": 0.6510656476020813,
86
+ "eval_non-hate_accuracy": 0.8172356369691923,
87
+ "eval_non-hate_f1": 0.8289832489287106,
88
+ "eval_non-hate_f1_macro": 0.8163691389927414,
89
+ "eval_precision_macro": 0.7102399413592261,
90
+ "eval_recall_macro": 0.7082806408098472,
91
+ "eval_runtime": 11.0908,
92
+ "eval_samples_per_second": 216.577,
93
+ "eval_steps_per_second": 0.27,
94
+ "step": 152
95
+ }
96
+ ],
97
+ "logging_steps": 500,
98
+ "max_steps": 380,
99
+ "num_input_tokens_seen": 0,
100
+ "num_train_epochs": 10,
101
+ "save_steps": 500,
102
+ "stateful_callbacks": {
103
+ "TrainerControl": {
104
+ "args": {
105
+ "should_epoch_stop": false,
106
+ "should_evaluate": false,
107
+ "should_log": false,
108
+ "should_save": true,
109
+ "should_training_stop": false
110
+ },
111
+ "attributes": {}
112
+ }
113
+ },
114
+ "total_flos": 3.519018937265695e+16,
115
+ "train_batch_size": 256,
116
+ "trial_name": null,
117
+ "trial_params": null
118
+ }
toxicn/expression_ie/2/checkpoint-152/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/2/config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "id2label": {
14
+ "0": "non-hate",
15
+ "1": "explicit",
16
+ "2": "implicit"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 4096,
20
+ "label2id": {
21
+ "explicit": 1,
22
+ "implicit": 2,
23
+ "non-hate": 0
24
+ },
25
+ "layer_norm_eps": 1e-12,
26
+ "max_position_embeddings": 512,
27
+ "model_type": "bert",
28
+ "note": "Combined implicit and reporting labels.",
29
+ "num_attention_heads": 16,
30
+ "num_hidden_layers": 24,
31
+ "output_past": true,
32
+ "pad_token_id": 0,
33
+ "pooler_fc_size": 768,
34
+ "pooler_num_attention_heads": 12,
35
+ "pooler_num_fc_layers": 3,
36
+ "pooler_size_per_head": 128,
37
+ "pooler_type": "first_token_transform",
38
+ "position_embedding_type": "absolute",
39
+ "problem_type": "single_label_classification",
40
+ "torch_dtype": "float32",
41
+ "transformers_version": "4.51.3",
42
+ "type_vocab_size": 2,
43
+ "use_cache": true,
44
+ "vocab_size": 21128
45
+ }
toxicn/expression_ie/2/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
toxicn/expression_ie/2/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/2/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 1000000000000000019884624838656,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
toxicn/expression_ie/2/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/3/checkpoint-190/config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "id2label": {
14
+ "0": "non-hate",
15
+ "1": "explicit",
16
+ "2": "implicit"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 4096,
20
+ "label2id": {
21
+ "explicit": 1,
22
+ "implicit": 2,
23
+ "non-hate": 0
24
+ },
25
+ "layer_norm_eps": 1e-12,
26
+ "max_position_embeddings": 512,
27
+ "model_type": "bert",
28
+ "note": "Combined implicit and reporting labels.",
29
+ "num_attention_heads": 16,
30
+ "num_hidden_layers": 24,
31
+ "output_past": true,
32
+ "pad_token_id": 0,
33
+ "pooler_fc_size": 768,
34
+ "pooler_num_attention_heads": 12,
35
+ "pooler_num_fc_layers": 3,
36
+ "pooler_size_per_head": 128,
37
+ "pooler_type": "first_token_transform",
38
+ "position_embedding_type": "absolute",
39
+ "problem_type": "single_label_classification",
40
+ "torch_dtype": "float32",
41
+ "transformers_version": "4.51.3",
42
+ "type_vocab_size": 2,
43
+ "use_cache": true,
44
+ "vocab_size": 21128
45
+ }
toxicn/expression_ie/3/checkpoint-190/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
toxicn/expression_ie/3/checkpoint-190/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/3/checkpoint-190/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 1000000000000000019884624838656,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
toxicn/expression_ie/3/checkpoint-190/trainer_state.json ADDED
@@ -0,0 +1,139 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": 0.7095564074317182,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 5.0,
6
+ "eval_steps": 500,
7
+ "global_step": 190,
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_accuracy": 0.6910907577019151,
15
+ "eval_explicit_accuracy": 0.8309741881765196,
16
+ "eval_explicit_f1": 0.6019607843137255,
17
+ "eval_explicit_f1_macro": 0.7473334576959747,
18
+ "eval_f1_macro": 0.6297427953651995,
19
+ "eval_implicit_accuracy": 0.7772689425478768,
20
+ "eval_implicit_f1": 0.4850818094321463,
21
+ "eval_implicit_f1_macro": 0.6714917679298846,
22
+ "eval_loss": 0.7187609672546387,
23
+ "eval_non-hate_accuracy": 0.7739383846794338,
24
+ "eval_non-hate_f1": 0.8021857923497268,
25
+ "eval_non-hate_f1_macro": 0.7692327698999726,
26
+ "eval_precision_macro": 0.6494776339297957,
27
+ "eval_recall_macro": 0.6197063103274625,
28
+ "eval_runtime": 10.3367,
29
+ "eval_samples_per_second": 232.376,
30
+ "eval_steps_per_second": 0.29,
31
+ "step": 38
32
+ },
33
+ {
34
+ "epoch": 2.0,
35
+ "eval_accuracy": 0.7289758534554538,
36
+ "eval_explicit_accuracy": 0.8555370524562864,
37
+ "eval_explicit_f1": 0.6526526526526526,
38
+ "eval_explicit_f1_macro": 0.7807284288230412,
39
+ "eval_f1_macro": 0.6825812478290718,
40
+ "eval_implicit_accuracy": 0.7918401332223147,
41
+ "eval_implicit_f1": 0.5667244367417678,
42
+ "eval_implicit_f1_macro": 0.7148690676859524,
43
+ "eval_loss": 0.6403889656066895,
44
+ "eval_non-hate_accuracy": 0.8105745212323064,
45
+ "eval_non-hate_f1": 0.8283666540927952,
46
+ "eval_non-hate_f1_macro": 0.8085168152024589,
47
+ "eval_precision_macro": 0.6965747421165688,
48
+ "eval_recall_macro": 0.6734543967445351,
49
+ "eval_runtime": 10.2193,
50
+ "eval_samples_per_second": 235.046,
51
+ "eval_steps_per_second": 0.294,
52
+ "step": 76
53
+ },
54
+ {
55
+ "epoch": 3.0,
56
+ "eval_accuracy": 0.7402164862614488,
57
+ "eval_explicit_accuracy": 0.8601165695253955,
58
+ "eval_explicit_f1": 0.6712328767123288,
59
+ "eval_explicit_f1_macro": 0.7911954970552654,
60
+ "eval_f1_macro": 0.6930499968329465,
61
+ "eval_implicit_accuracy": 0.798917568692756,
62
+ "eval_implicit_f1": 0.5691347011596789,
63
+ "eval_implicit_f1_macro": 0.7189958056436461,
64
+ "eval_loss": 0.6372141242027283,
65
+ "eval_non-hate_accuracy": 0.821398834304746,
66
+ "eval_non-hate_f1": 0.8387824126268321,
67
+ "eval_non-hate_f1_macro": 0.8192978792018901,
68
+ "eval_precision_macro": 0.7059921933281124,
69
+ "eval_recall_macro": 0.6839453078498107,
70
+ "eval_runtime": 10.174,
71
+ "eval_samples_per_second": 236.093,
72
+ "eval_steps_per_second": 0.295,
73
+ "step": 114
74
+ },
75
+ {
76
+ "epoch": 4.0,
77
+ "eval_accuracy": 0.7514571190674438,
78
+ "eval_explicit_accuracy": 0.8717735220649458,
79
+ "eval_explicit_f1": 0.7072243346007605,
80
+ "eval_explicit_f1_macro": 0.8125673911809772,
81
+ "eval_f1_macro": 0.7090624611790778,
82
+ "eval_implicit_accuracy": 0.8047460449625312,
83
+ "eval_implicit_f1": 0.5778577857785778,
84
+ "eval_implicit_f1_macro": 0.7254303821933777,
85
+ "eval_loss": 0.6451897621154785,
86
+ "eval_non-hate_accuracy": 0.8263946711074105,
87
+ "eval_non-hate_f1": 0.8421052631578947,
88
+ "eval_non-hate_f1_macro": 0.8246587342141762,
89
+ "eval_precision_macro": 0.7187681511493437,
90
+ "eval_recall_macro": 0.7017155397435338,
91
+ "eval_runtime": 10.2699,
92
+ "eval_samples_per_second": 233.887,
93
+ "eval_steps_per_second": 0.292,
94
+ "step": 152
95
+ },
96
+ {
97
+ "epoch": 5.0,
98
+ "eval_accuracy": 0.7477102414654455,
99
+ "eval_explicit_accuracy": 0.8655287260616153,
100
+ "eval_explicit_f1": 0.7017543859649122,
101
+ "eval_explicit_f1_macro": 0.8074748817757913,
102
+ "eval_f1_macro": 0.7095564074317182,
103
+ "eval_implicit_accuracy": 0.7997502081598667,
104
+ "eval_implicit_f1": 0.586414445399828,
105
+ "eval_implicit_f1_macro": 0.7271539406345473,
106
+ "eval_loss": 0.6919171214103699,
107
+ "eval_non-hate_accuracy": 0.8301415487094088,
108
+ "eval_non-hate_f1": 0.8405003909304144,
109
+ "eval_non-hate_f1_macro": 0.829422056551583,
110
+ "eval_precision_macro": 0.7108167347861242,
111
+ "eval_recall_macro": 0.7083569695346545,
112
+ "eval_runtime": 10.217,
113
+ "eval_samples_per_second": 235.099,
114
+ "eval_steps_per_second": 0.294,
115
+ "step": 190
116
+ }
117
+ ],
118
+ "logging_steps": 500,
119
+ "max_steps": 380,
120
+ "num_input_tokens_seen": 0,
121
+ "num_train_epochs": 10,
122
+ "save_steps": 500,
123
+ "stateful_callbacks": {
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": 4.398773671582119e+16,
136
+ "train_batch_size": 256,
137
+ "trial_name": null,
138
+ "trial_params": null
139
+ }
toxicn/expression_ie/3/checkpoint-190/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/3/config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "id2label": {
14
+ "0": "non-hate",
15
+ "1": "explicit",
16
+ "2": "implicit"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 4096,
20
+ "label2id": {
21
+ "explicit": 1,
22
+ "implicit": 2,
23
+ "non-hate": 0
24
+ },
25
+ "layer_norm_eps": 1e-12,
26
+ "max_position_embeddings": 512,
27
+ "model_type": "bert",
28
+ "note": "Combined implicit and reporting labels.",
29
+ "num_attention_heads": 16,
30
+ "num_hidden_layers": 24,
31
+ "output_past": true,
32
+ "pad_token_id": 0,
33
+ "pooler_fc_size": 768,
34
+ "pooler_num_attention_heads": 12,
35
+ "pooler_num_fc_layers": 3,
36
+ "pooler_size_per_head": 128,
37
+ "pooler_type": "first_token_transform",
38
+ "position_embedding_type": "absolute",
39
+ "problem_type": "single_label_classification",
40
+ "torch_dtype": "float32",
41
+ "transformers_version": "4.51.3",
42
+ "type_vocab_size": 2,
43
+ "use_cache": true,
44
+ "vocab_size": 21128
45
+ }
toxicn/expression_ie/3/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
toxicn/expression_ie/3/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/3/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 1000000000000000019884624838656,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
toxicn/expression_ie/3/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
toxicn/expression_ie/4/config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "id2label": {
14
+ "0": "non-hate",
15
+ "1": "explicit",
16
+ "2": "implicit"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 4096,
20
+ "label2id": {
21
+ "explicit": 1,
22
+ "implicit": 2,
23
+ "non-hate": 0
24
+ },
25
+ "layer_norm_eps": 1e-12,
26
+ "max_position_embeddings": 512,
27
+ "model_type": "bert",
28
+ "note": "Combined implicit and reporting labels.",
29
+ "num_attention_heads": 16,
30
+ "num_hidden_layers": 24,
31
+ "output_past": true,
32
+ "pad_token_id": 0,
33
+ "pooler_fc_size": 768,
34
+ "pooler_num_attention_heads": 12,
35
+ "pooler_num_fc_layers": 3,
36
+ "pooler_size_per_head": 128,
37
+ "pooler_type": "first_token_transform",
38
+ "position_embedding_type": "absolute",
39
+ "problem_type": "single_label_classification",
40
+ "torch_dtype": "float32",
41
+ "transformers_version": "4.51.3",
42
+ "type_vocab_size": 2,
43
+ "use_cache": true,
44
+ "vocab_size": 21128
45
+ }
vocab.txt ADDED
The diff for this file is too large to render. See raw diff