Hugging Face's logo

imvladikon
/
t5-english-ner

Token Classification
Transformers
PyTorch
t5
Generated from Trainer
custom_code
text-generation-inference
Inference Endpoints
Model card Files Community
1
imvladikon commited on
Commit
f6b8a39
•
1 Parent(s): 981bba3

init

Browse files
Files changed (30) hide show
  1. README.md +0 -0
  2. added_tokens.json +3 -0
  3. all_results.json +125 -0
  4. config.json +103 -0
  5. eval_classification_report.csv +12 -0
  6. eval_classification_report.json +68 -0
  7. eval_confusion_matrix.csv +11 -0
  8. eval_results.json +44 -0
  9. eval_results_groups.json +85 -0
  10. eval_sk_classification_report.csv +21 -0
  11. label2id.json +21 -0
  12. predict_classification_report.csv +11 -0
  13. predict_classification_report.json +62 -0
  14. predict_confusion_matrix.csv +10 -0
  15. predict_results.json +38 -0
  16. predict_results_groups.json +63 -0
  17. predict_sk_classification_report.csv +19 -0
  18. pytorch_model.bin +3 -0
  19. special_tokens_map.json +108 -0
  20. spiece.model +3 -0
  21. tokenizer.json +0 -0
  22. tokenizer_config.json +115 -0
  23. train_classification_report.csv +13 -0
  24. train_classification_report.json +74 -0
  25. train_confusion_matrix.csv +12 -0
  26. train_results.json +46 -0
  27. train_results_groups.json +71 -0
  28. train_sk_classification_report.csv +23 -0
  29. trainer_state.json +0 -0
  30. training_args.bin +3 -0
README.md ADDED
The diff for this file is too large to render. See raw diff
 
added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "<s>": 32100
3
+ }
all_results.json ADDED
@@ -0,0 +1,125 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_COMMERCIAL_ITEM_f1": 0.0,
4
+ "eval_COMMERCIAL_ITEM_number": 1,
5
+ "eval_COMMERCIAL_ITEM_precision": 0.0,
6
+ "eval_COMMERCIAL_ITEM_recall": 0.0,
7
+ "eval_DATE_f1": 0.8666666666666666,
8
+ "eval_DATE_number": 14,
9
+ "eval_DATE_precision": 0.8125,
10
+ "eval_DATE_recall": 0.9285714285714286,
11
+ "eval_LOCATION_f1": 0.7317073170731706,
12
+ "eval_LOCATION_number": 20,
13
+ "eval_LOCATION_precision": 0.7142857142857143,
14
+ "eval_LOCATION_recall": 0.75,
15
+ "eval_ORGANIZATION_f1": 0.8848484848484848,
16
+ "eval_ORGANIZATION_number": 80,
17
+ "eval_ORGANIZATION_precision": 0.8588235294117647,
18
+ "eval_ORGANIZATION_recall": 0.9125,
19
+ "eval_OTHER_f1": 0.35,
20
+ "eval_OTHER_number": 21,
21
+ "eval_OTHER_precision": 0.3684210526315789,
22
+ "eval_OTHER_recall": 0.3333333333333333,
23
+ "eval_PERSON_f1": 0.8709677419354839,
24
+ "eval_PERSON_number": 29,
25
+ "eval_PERSON_precision": 0.8181818181818182,
26
+ "eval_PERSON_recall": 0.9310344827586207,
27
+ "eval_QUANTITY_f1": 0.8275862068965518,
28
+ "eval_QUANTITY_number": 14,
29
+ "eval_QUANTITY_precision": 0.8,
30
+ "eval_QUANTITY_recall": 0.8571428571428571,
31
+ "eval_TITLE_f1": 0.0,
32
+ "eval_TITLE_number": 7,
33
+ "eval_TITLE_precision": 0.0,
34
+ "eval_TITLE_recall": 0.0,
35
+ "eval_loss": 0.19558680057525635,
36
+ "eval_overall_accuracy": 0.9534472598703595,
37
+ "eval_overall_f1": 0.7696335078534032,
38
+ "eval_overall_precision": 0.75,
39
+ "eval_overall_recall": 0.7903225806451613,
40
+ "eval_runtime": 14.8553,
41
+ "eval_samples": 35,
42
+ "eval_samples_per_second": 2.356,
43
+ "eval_steps_per_second": 1.212,
44
+ "predict_DATE_f1": 0.9714285714285714,
45
+ "predict_DATE_number": 18,
46
+ "predict_DATE_precision": 1.0,
47
+ "predict_DATE_recall": 0.9444444444444444,
48
+ "predict_LOCATION_f1": 0.6666666666666666,
49
+ "predict_LOCATION_number": 9,
50
+ "predict_LOCATION_precision": 0.6666666666666666,
51
+ "predict_LOCATION_recall": 0.6666666666666666,
52
+ "predict_ORGANIZATION_f1": 0.9548387096774195,
53
+ "predict_ORGANIZATION_number": 78,
54
+ "predict_ORGANIZATION_precision": 0.961038961038961,
55
+ "predict_ORGANIZATION_recall": 0.9487179487179487,
56
+ "predict_OTHER_f1": 0.5490196078431373,
57
+ "predict_OTHER_number": 24,
58
+ "predict_OTHER_precision": 0.5185185185185185,
59
+ "predict_OTHER_recall": 0.5833333333333334,
60
+ "predict_PERSON_f1": 0.8518518518518519,
61
+ "predict_PERSON_number": 55,
62
+ "predict_PERSON_precision": 0.8679245283018868,
63
+ "predict_PERSON_recall": 0.8363636363636363,
64
+ "predict_QUANTITY_f1": 0.8620689655172413,
65
+ "predict_QUANTITY_number": 28,
66
+ "predict_QUANTITY_precision": 0.8333333333333334,
67
+ "predict_QUANTITY_recall": 0.8928571428571429,
68
+ "predict_TITLE_f1": 0.11764705882352941,
69
+ "predict_TITLE_number": 10,
70
+ "predict_TITLE_precision": 0.14285714285714285,
71
+ "predict_TITLE_recall": 0.1,
72
+ "predict_loss": 0.1715492457151413,
73
+ "predict_overall_accuracy": 0.968184311574328,
74
+ "predict_overall_f1": 0.8280542986425339,
75
+ "predict_overall_precision": 0.8318181818181818,
76
+ "predict_overall_recall": 0.8243243243243243,
77
+ "predict_runtime": 19.7294,
78
+ "predict_samples_per_second": 1.825,
79
+ "predict_steps_per_second": 0.912,
80
+ "train_COMMERCIAL_ITEM_f1": 0.0,
81
+ "train_COMMERCIAL_ITEM_number": 7,
82
+ "train_COMMERCIAL_ITEM_precision": 0.0,
83
+ "train_COMMERCIAL_ITEM_recall": 0.0,
84
+ "train_DATE_f1": 0.9695682944090589,
85
+ "train_DATE_number": 701,
86
+ "train_DATE_precision": 0.9620786516853933,
87
+ "train_DATE_recall": 0.9771754636233951,
88
+ "train_EVENT_f1": 0.0,
89
+ "train_EVENT_number": 22,
90
+ "train_EVENT_precision": 0.0,
91
+ "train_EVENT_recall": 0.0,
92
+ "train_LOCATION_f1": 0.8890041493775933,
93
+ "train_LOCATION_number": 961,
94
+ "train_LOCATION_precision": 0.8862461220268872,
95
+ "train_LOCATION_recall": 0.8917793964620188,
96
+ "train_ORGANIZATION_f1": 0.9276817999196465,
97
+ "train_ORGANIZATION_number": 2468,
98
+ "train_ORGANIZATION_precision": 0.9199203187250996,
99
+ "train_ORGANIZATION_recall": 0.9355753646677472,
100
+ "train_OTHER_f1": 0.5381083562901744,
101
+ "train_OTHER_number": 1083,
102
+ "train_OTHER_precision": 0.5351598173515981,
103
+ "train_OTHER_recall": 0.541089566020314,
104
+ "train_PERSON_f1": 0.9191374663072777,
105
+ "train_PERSON_number": 1449,
106
+ "train_PERSON_precision": 0.8979591836734694,
107
+ "train_PERSON_recall": 0.9413388543823327,
108
+ "train_QUANTITY_f1": 0.9049150206706478,
109
+ "train_QUANTITY_number": 1067,
110
+ "train_QUANTITY_precision": 0.8873873873873874,
111
+ "train_QUANTITY_recall": 0.9231490159325211,
112
+ "train_TITLE_f1": 0.19230769230769232,
113
+ "train_TITLE_number": 264,
114
+ "train_TITLE_precision": 0.22058823529411764,
115
+ "train_TITLE_recall": 0.17045454545454544,
116
+ "train_loss": 0.14113017916679382,
117
+ "train_overall_accuracy": 0.9713395552737148,
118
+ "train_overall_f1": 0.8465208501146292,
119
+ "train_overall_precision": 0.8415670814340274,
120
+ "train_overall_recall": 0.8515332834704562,
121
+ "train_runtime": 701.7841,
122
+ "train_samples": 1416,
123
+ "train_samples_per_second": 2.018,
124
+ "train_steps_per_second": 1.009
125
+ }
config.json ADDED
@@ -0,0 +1,103 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "t5-large",
3
+ "architectures": [
4
+ "EncT5ForTokenClassification"
5
+ ],
6
+ "d_ff": 4096,
7
+ "d_kv": 64,
8
+ "d_model": 1024,
9
+ "decoder_start_token_id": 0,
10
+ "dense_act_fn": "relu",
11
+ "dropout_rate": 0.1,
12
+ "eos_token_id": 1,
13
+ "feed_forward_proj": "relu",
14
+ "finetuning_task": "ner",
15
+ "id2label": {
16
+ "0": "O",
17
+ "1": "B-PERSON",
18
+ "2": "I-PERSON",
19
+ "3": "B-ORGANIZATION",
20
+ "4": "I-ORGANIZATION",
21
+ "5": "B-LOCATION",
22
+ "6": "I-LOCATION",
23
+ "7": "B-TITLE",
24
+ "8": "I-TITLE",
25
+ "9": "B-EVENT",
26
+ "10": "I-EVENT",
27
+ "11": "B-QUANTITY",
28
+ "12": "I-QUANTITY",
29
+ "13": "B-DATE",
30
+ "14": "I-DATE",
31
+ "15": "B-COMMERCIAL_ITEM",
32
+ "16": "I-COMMERCIAL_ITEM",
33
+ "17": "B-OTHER",
34
+ "18": "I-OTHER"
35
+ },
36
+ "initializer_factor": 1.0,
37
+ "is_encoder_decoder": true,
38
+ "is_gated_act": false,
39
+ "label2id": {
40
+ "B-COMMERCIAL_ITEM": 15,
41
+ "B-DATE": 13,
42
+ "B-EVENT": 9,
43
+ "B-LOCATION": 5,
44
+ "B-ORGANIZATION": 3,
45
+ "B-OTHER": 17,
46
+ "B-PERSON": 1,
47
+ "B-QUANTITY": 11,
48
+ "B-TITLE": 7,
49
+ "I-COMMERCIAL_ITEM": 16,
50
+ "I-DATE": 14,
51
+ "I-EVENT": 10,
52
+ "I-LOCATION": 6,
53
+ "I-ORGANIZATION": 4,
54
+ "I-OTHER": 18,
55
+ "I-PERSON": 2,
56
+ "I-QUANTITY": 12,
57
+ "I-TITLE": 8,
58
+ "O": 0
59
+ },
60
+ "layer_norm_epsilon": 1e-06,
61
+ "model_type": "t5",
62
+ "n_positions": 512,
63
+ "num_decoder_layers": 24,
64
+ "num_heads": 16,
65
+ "num_layers": 24,
66
+ "output_past": true,
67
+ "pad_token_id": 0,
68
+ "relative_attention_max_distance": 128,
69
+ "relative_attention_num_buckets": 32,
70
+ "task_specific_params": {
71
+ "summarization": {
72
+ "early_stopping": true,
73
+ "length_penalty": 2.0,
74
+ "max_length": 200,
75
+ "min_length": 30,
76
+ "no_repeat_ngram_size": 3,
77
+ "num_beams": 4,
78
+ "prefix": "summarize: "
79
+ },
80
+ "translation_en_to_de": {
81
+ "early_stopping": true,
82
+ "max_length": 300,
83
+ "num_beams": 4,
84
+ "prefix": "translate English to German: "
85
+ },
86
+ "translation_en_to_fr": {
87
+ "early_stopping": true,
88
+ "max_length": 300,
89
+ "num_beams": 4,
90
+ "prefix": "translate English to French: "
91
+ },
92
+ "translation_en_to_ro": {
93
+ "early_stopping": true,
94
+ "max_length": 300,
95
+ "num_beams": 4,
96
+ "prefix": "translate English to Romanian: "
97
+ }
98
+ },
99
+ "torch_dtype": "float32",
100
+ "transformers_version": "4.21.1",
101
+ "use_cache": true,
102
+ "vocab_size": 32128
103
+ }
eval_classification_report.csv ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ,tag,precision,recall,f1-score,support
2
+ 0,COMMERCIAL_ITEM,0.0,0.0,0.0,1.0
3
+ 1,DATE,0.8125,0.9285714285714286,0.8666666666666666,14.0
4
+ 2,LOCATION,0.7142857142857143,0.75,0.7317073170731706,20.0
5
+ 3,ORGANIZATION,0.8588235294117647,0.9125,0.8848484848484848,80.0
6
+ 4,OTHER,0.3684210526315789,0.3333333333333333,0.35,21.0
7
+ 5,PERSON,0.8181818181818182,0.9310344827586207,0.8709677419354839,29.0
8
+ 6,QUANTITY,0.8,0.8571428571428571,0.8275862068965518,14.0
9
+ 7,TITLE,0.0,0.0,0.0,7.0
10
+ 8,micro avg,0.75,0.7903225806451613,0.7696335078534032,186.0
11
+ 9,macro avg,0.5465265143138595,0.5890727627257799,0.5664720521775447,186.0
12
+ 10,weighted avg,0.7367242552214589,0.7903225806451613,0.762094784276109,186.0
eval_classification_report.json ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "COMMERCIAL_ITEM": {
3
+ "precision": 0.0,
4
+ "recall": 0.0,
5
+ "f1-score": 0.0,
6
+ "support": 1.0
7
+ },
8
+ "DATE": {
9
+ "precision": 0.8125,
10
+ "recall": 0.9285714285714286,
11
+ "f1-score": 0.8666666666666666,
12
+ "support": 14.0
13
+ },
14
+ "LOCATION": {
15
+ "precision": 0.7142857142857143,
16
+ "recall": 0.75,
17
+ "f1-score": 0.7317073170731706,
18
+ "support": 20.0
19
+ },
20
+ "ORGANIZATION": {
21
+ "precision": 0.8588235294117647,
22
+ "recall": 0.9125,
23
+ "f1-score": 0.8848484848484848,
24
+ "support": 80.0
25
+ },
26
+ "OTHER": {
27
+ "precision": 0.3684210526315789,
28
+ "recall": 0.3333333333333333,
29
+ "f1-score": 0.35,
30
+ "support": 21.0
31
+ },
32
+ "PERSON": {
33
+ "precision": 0.8181818181818182,
34
+ "recall": 0.9310344827586207,
35
+ "f1-score": 0.8709677419354839,
36
+ "support": 29.0
37
+ },
38
+ "QUANTITY": {
39
+ "precision": 0.8,
40
+ "recall": 0.8571428571428571,
41
+ "f1-score": 0.8275862068965518,
42
+ "support": 14.0
43
+ },
44
+ "TITLE": {
45
+ "precision": 0.0,
46
+ "recall": 0.0,
47
+ "f1-score": 0.0,
48
+ "support": 7.0
49
+ },
50
+ "micro avg": {
51
+ "precision": 0.75,
52
+ "recall": 0.7903225806451613,
53
+ "f1-score": 0.7696335078534032,
54
+ "support": 186.0
55
+ },
56
+ "macro avg": {
57
+ "precision": 0.5465265143138595,
58
+ "recall": 0.5890727627257799,
59
+ "f1-score": 0.5664720521775447,
60
+ "support": 186.0
61
+ },
62
+ "weighted avg": {
63
+ "precision": 0.7367242552214589,
64
+ "recall": 0.7903225806451613,
65
+ "f1-score": 0.762094784276109,
66
+ "support": 186.0
67
+ }
68
+ }
eval_confusion_matrix.csv ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ,Actual\Predicted,DATE,LOCATION,O,ORGANIZATION,OTHER,PERSON,QUANTITY,TITLE,All
2
+ 0,COMMERCIAL_ITEM,0,0,0,0,1,0,0,2,3
3
+ 1,DATE,30,0,1,0,0,0,0,0,31
4
+ 2,LOCATION,0,27,3,3,0,0,0,0,33
5
+ 3,O,1,0,1327,3,7,2,4,2,1346
6
+ 4,ORGANIZATION,0,1,4,147,0,1,0,4,157
7
+ 5,OTHER,0,1,18,0,17,0,0,0,36
8
+ 6,PERSON,0,0,0,0,0,42,0,2,44
9
+ 7,QUANTITY,0,0,2,0,0,0,33,0,35
10
+ 8,TITLE,0,0,3,4,0,3,0,2,12
11
+ 9,All,31,29,1358,157,25,48,37,12,1697
eval_results.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_COMMERCIAL_ITEM_f1": 0.0,
4
+ "eval_COMMERCIAL_ITEM_number": 1,
5
+ "eval_COMMERCIAL_ITEM_precision": 0.0,
6
+ "eval_COMMERCIAL_ITEM_recall": 0.0,
7
+ "eval_DATE_f1": 0.8666666666666666,
8
+ "eval_DATE_number": 14,
9
+ "eval_DATE_precision": 0.8125,
10
+ "eval_DATE_recall": 0.9285714285714286,
11
+ "eval_LOCATION_f1": 0.7317073170731706,
12
+ "eval_LOCATION_number": 20,
13
+ "eval_LOCATION_precision": 0.7142857142857143,
14
+ "eval_LOCATION_recall": 0.75,
15
+ "eval_ORGANIZATION_f1": 0.8848484848484848,
16
+ "eval_ORGANIZATION_number": 80,
17
+ "eval_ORGANIZATION_precision": 0.8588235294117647,
18
+ "eval_ORGANIZATION_recall": 0.9125,
19
+ "eval_OTHER_f1": 0.35,
20
+ "eval_OTHER_number": 21,
21
+ "eval_OTHER_precision": 0.3684210526315789,
22
+ "eval_OTHER_recall": 0.3333333333333333,
23
+ "eval_PERSON_f1": 0.8709677419354839,
24
+ "eval_PERSON_number": 29,
25
+ "eval_PERSON_precision": 0.8181818181818182,
26
+ "eval_PERSON_recall": 0.9310344827586207,
27
+ "eval_QUANTITY_f1": 0.8275862068965518,
28
+ "eval_QUANTITY_number": 14,
29
+ "eval_QUANTITY_precision": 0.8,
30
+ "eval_QUANTITY_recall": 0.8571428571428571,
31
+ "eval_TITLE_f1": 0.0,
32
+ "eval_TITLE_number": 7,
33
+ "eval_TITLE_precision": 0.0,
34
+ "eval_TITLE_recall": 0.0,
35
+ "eval_loss": 0.19558680057525635,
36
+ "eval_overall_accuracy": 0.9534472598703595,
37
+ "eval_overall_f1": 0.7696335078534032,
38
+ "eval_overall_precision": 0.75,
39
+ "eval_overall_recall": 0.7903225806451613,
40
+ "eval_runtime": 14.8553,
41
+ "eval_samples": 35,
42
+ "eval_samples_per_second": 2.356,
43
+ "eval_steps_per_second": 1.212
44
+ }
eval_results_groups.json ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "eval_loss": 0.19558680057525635,
3
+ "eval_COMMERCIAL_ITEM_precision": 0.0,
4
+ "eval_COMMERCIAL_ITEM_recall": 0.0,
5
+ "eval_COMMERCIAL_ITEM_f1": 0.0,
6
+ "eval_COMMERCIAL_ITEM_number": 1,
7
+ "DATE": {
8
+ "precision": 0.8125,
9
+ "recall": 0.9285714285714286,
10
+ "f1": 0.8666666666666666,
11
+ "number": 14
12
+ },
13
+ "LOCATION": {
14
+ "precision": 0.7142857142857143,
15
+ "recall": 0.75,
16
+ "f1": 0.7317073170731706,
17
+ "number": 20
18
+ },
19
+ "ORGANIZATION": {
20
+ "precision": 0.8588235294117647,
21
+ "recall": 0.9125,
22
+ "f1": 0.8848484848484848,
23
+ "number": 80
24
+ },
25
+ "OTHER": {
26
+ "precision": 0.3684210526315789,
27
+ "recall": 0.3333333333333333,
28
+ "f1": 0.35,
29
+ "number": 21
30
+ },
31
+ "PERSON": {
32
+ "precision": 0.8181818181818182,
33
+ "recall": 0.9310344827586207,
34
+ "f1": 0.8709677419354839,
35
+ "number": 29
36
+ },
37
+ "QUANTITY": {
38
+ "precision": 0.8,
39
+ "recall": 0.8571428571428571,
40
+ "f1": 0.8275862068965518,
41
+ "number": 14
42
+ },
43
+ "TITLE": {
44
+ "precision": 0.0,
45
+ "recall": 0.0,
46
+ "f1": 0.0,
47
+ "number": 7
48
+ },
49
+ "overall": {
50
+ "precision": 0.75,
51
+ "recall": 0.7903225806451613,
52
+ "f1": 0.7696335078534032,
53
+ "accuracy": 0.9534472598703595
54
+ },
55
+ "eval_runtime": 14.8553,
56
+ "eval_samples_per_second": 2.356,
57
+ "eval_steps_per_second": 1.212,
58
+ "epoch": 3.0,
59
+ "eval_samples": 35,
60
+ "model_name": "t5-large",
61
+ "dataset_name": "imvladikon/ner_news_small",
62
+ "split_instructions": "80:10:10",
63
+ "device": "cpu",
64
+ "label2id": {
65
+ "O": 0,
66
+ "B-PERSON": 1,
67
+ "I-PERSON": 2,
68
+ "B-ORGANIZATION": 3,
69
+ "I-ORGANIZATION": 4,
70
+ "B-LOCATION": 5,
71
+ "I-LOCATION": 6,
72
+ "B-TITLE": 7,
73
+ "I-TITLE": 8,
74
+ "B-EVENT": 9,
75
+ "I-EVENT": 10,
76
+ "B-QUANTITY": 11,
77
+ "I-QUANTITY": 12,
78
+ "B-DATE": 13,
79
+ "I-DATE": 14,
80
+ "B-COMMERCIAL_ITEM": 15,
81
+ "I-COMMERCIAL_ITEM": 16,
82
+ "B-OTHER": 17,
83
+ "I-OTHER": 18
84
+ }
85
+ }
eval_sk_classification_report.csv ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ,tag,precision,recall,f1-score,support,entity
2
+ 0,B-COMMERCIAL_ITEM,0.0,0.0,0.0,1.0,COMMERCIAL_ITEM
3
+ 8,I-COMMERCIAL_ITEM,0.0,0.0,0.0,2.0,COMMERCIAL_ITEM
4
+ 9,I-DATE,1.0,0.8823529411764706,0.9375,17.0,DATE
5
+ 1,B-DATE,0.875,1.0,0.9333333333333333,14.0,DATE
6
+ 2,B-LOCATION,0.85,0.85,0.85,20.0,LOCATION
7
+ 10,I-LOCATION,1.0,0.6923076923076923,0.8181818181818181,13.0,LOCATION
8
+ 16,O,0.9771723122238586,0.9858841010401189,0.981508875739645,1346.0,O
9
+ 3,B-ORGANIZATION,0.9146341463414634,0.9375,0.925925925925926,80.0,ORGANIZATION
10
+ 11,I-ORGANIZATION,0.9333333333333333,0.9090909090909091,0.9210526315789475,77.0,ORGANIZATION
11
+ 12,I-OTHER,0.625,0.3333333333333333,0.43478260869565216,15.0,OTHER
12
+ 4,B-OTHER,0.5882352941176471,0.47619047619047616,0.5263157894736842,21.0,OTHER
13
+ 5,B-PERSON,0.8787878787878788,1.0,0.9354838709677419,29.0,PERSON
14
+ 13,I-PERSON,0.8666666666666667,0.8666666666666667,0.8666666666666667,15.0,PERSON
15
+ 6,B-QUANTITY,0.8,0.8571428571428571,0.8275862068965518,14.0,QUANTITY
16
+ 14,I-QUANTITY,0.9090909090909091,0.9523809523809523,0.9302325581395349,21.0,QUANTITY
17
+ 15,I-TITLE,0.18181818181818182,0.4,0.25000000000000006,5.0,TITLE
18
+ 7,B-TITLE,0.0,0.0,0.0,7.0,TITLE
19
+ 17,accuracy,0.9534472598703595,0.9534472598703595,0.9534472598703595,0.9534472598703595,accuracy
20
+ 18,macro avg,0.6705728660223493,0.6554617605487928,0.6552100167999706,1697.0,macro avg
21
+ 19,weighted avg,0.9493070832578766,0.9534472598703595,0.9503954610130667,1697.0,weighted avg
label2id.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "O": 0,
3
+ "B-PERSON": 1,
4
+ "I-PERSON": 2,
5
+ "B-ORGANIZATION": 3,
6
+ "I-ORGANIZATION": 4,
7
+ "B-LOCATION": 5,
8
+ "I-LOCATION": 6,
9
+ "B-TITLE": 7,
10
+ "I-TITLE": 8,
11
+ "B-EVENT": 9,
12
+ "I-EVENT": 10,
13
+ "B-QUANTITY": 11,
14
+ "I-QUANTITY": 12,
15
+ "B-DATE": 13,
16
+ "I-DATE": 14,
17
+ "B-COMMERCIAL_ITEM": 15,
18
+ "I-COMMERCIAL_ITEM": 16,
19
+ "B-OTHER": 17,
20
+ "I-OTHER": 18
21
+ }
predict_classification_report.csv ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ,tag,precision,recall,f1-score,support
2
+ 0,DATE,1.0,0.9444444444444444,0.9714285714285714,18.0
3
+ 1,LOCATION,0.6666666666666666,0.6666666666666666,0.6666666666666666,9.0
4
+ 2,ORGANIZATION,0.961038961038961,0.9487179487179487,0.9548387096774195,78.0
5
+ 3,OTHER,0.5185185185185185,0.5833333333333334,0.5490196078431373,24.0
6
+ 4,PERSON,0.8679245283018868,0.8363636363636363,0.8518518518518519,55.0
7
+ 5,QUANTITY,0.8333333333333334,0.8928571428571429,0.8620689655172413,28.0
8
+ 6,TITLE,0.14285714285714285,0.1,0.11764705882352941,10.0
9
+ 7,micro avg,0.8318181818181818,0.8243243243243243,0.8280542986425339,222.0
10
+ 8,macro avg,0.7129055929595013,0.7103404531975961,0.7105030616869168,222.0
11
+ 9,weighted avg,0.8283929604684321,0.8243243243243243,0.8257020617268388,222.0
predict_classification_report.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "DATE": {
3
+ "precision": 1.0,
4
+ "recall": 0.9444444444444444,
5
+ "f1-score": 0.9714285714285714,
6
+ "support": 18.0
7
+ },
8
+ "LOCATION": {
9
+ "precision": 0.6666666666666666,
10
+ "recall": 0.6666666666666666,
11
+ "f1-score": 0.6666666666666666,
12
+ "support": 9.0
13
+ },
14
+ "ORGANIZATION": {
15
+ "precision": 0.961038961038961,
16
+ "recall": 0.9487179487179487,
17
+ "f1-score": 0.9548387096774195,
18
+ "support": 78.0
19
+ },
20
+ "OTHER": {
21
+ "precision": 0.5185185185185185,
22
+ "recall": 0.5833333333333334,
23
+ "f1-score": 0.5490196078431373,
24
+ "support": 24.0
25
+ },
26
+ "PERSON": {
27
+ "precision": 0.8679245283018868,
28
+ "recall": 0.8363636363636363,
29
+ "f1-score": 0.8518518518518519,
30
+ "support": 55.0
31
+ },
32
+ "QUANTITY": {
33
+ "precision": 0.8333333333333334,
34
+ "recall": 0.8928571428571429,
35
+ "f1-score": 0.8620689655172413,
36
+ "support": 28.0
37
+ },
38
+ "TITLE": {
39
+ "precision": 0.14285714285714285,
40
+ "recall": 0.1,
41
+ "f1-score": 0.11764705882352941,
42
+ "support": 10.0
43
+ },
44
+ "micro avg": {
45
+ "precision": 0.8318181818181818,
46
+ "recall": 0.8243243243243243,
47
+ "f1-score": 0.8280542986425339,
48
+ "support": 222.0
49
+ },
50
+ "macro avg": {
51
+ "precision": 0.7129055929595013,
52
+ "recall": 0.7103404531975961,
53
+ "f1-score": 0.7105030616869168,
54
+ "support": 222.0
55
+ },
56
+ "weighted avg": {
57
+ "precision": 0.8283929604684321,
58
+ "recall": 0.8243243243243243,
59
+ "f1-score": 0.8257020617268388,
60
+ "support": 222.0
61
+ }
62
+ }
predict_confusion_matrix.csv ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ ,Actual\Predicted,DATE,LOCATION,O,ORGANIZATION,OTHER,PERSON,QUANTITY,TITLE,All
2
+ 0,DATE,31,0,1,0,0,0,0,0,32
3
+ 1,LOCATION,0,7,0,1,1,1,0,0,10
4
+ 2,O,0,6,1387,1,3,1,3,0,1401
5
+ 3,ORGANIZATION,0,1,1,164,0,1,0,1,168
6
+ 4,OTHER,0,0,9,0,22,0,0,3,34
7
+ 5,PERSON,0,0,2,0,0,80,0,0,82
8
+ 6,QUANTITY,0,0,4,0,0,0,65,0,69
9
+ 7,TITLE,0,0,6,0,4,3,0,14,27
10
+ 8,All,31,14,1410,166,30,86,68,18,1823
predict_results.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "predict_DATE_f1": 0.9714285714285714,
3
+ "predict_DATE_number": 18,
4
+ "predict_DATE_precision": 1.0,
5
+ "predict_DATE_recall": 0.9444444444444444,
6
+ "predict_LOCATION_f1": 0.6666666666666666,
7
+ "predict_LOCATION_number": 9,
8
+ "predict_LOCATION_precision": 0.6666666666666666,
9
+ "predict_LOCATION_recall": 0.6666666666666666,
10
+ "predict_ORGANIZATION_f1": 0.9548387096774195,
11
+ "predict_ORGANIZATION_number": 78,
12
+ "predict_ORGANIZATION_precision": 0.961038961038961,
13
+ "predict_ORGANIZATION_recall": 0.9487179487179487,
14
+ "predict_OTHER_f1": 0.5490196078431373,
15
+ "predict_OTHER_number": 24,
16
+ "predict_OTHER_precision": 0.5185185185185185,
17
+ "predict_OTHER_recall": 0.5833333333333334,
18
+ "predict_PERSON_f1": 0.8518518518518519,
19
+ "predict_PERSON_number": 55,
20
+ "predict_PERSON_precision": 0.8679245283018868,
21
+ "predict_PERSON_recall": 0.8363636363636363,
22
+ "predict_QUANTITY_f1": 0.8620689655172413,
23
+ "predict_QUANTITY_number": 28,
24
+ "predict_QUANTITY_precision": 0.8333333333333334,
25
+ "predict_QUANTITY_recall": 0.8928571428571429,
26
+ "predict_TITLE_f1": 0.11764705882352941,
27
+ "predict_TITLE_number": 10,
28
+ "predict_TITLE_precision": 0.14285714285714285,
29
+ "predict_TITLE_recall": 0.1,
30
+ "predict_loss": 0.1715492457151413,
31
+ "predict_overall_accuracy": 0.968184311574328,
32
+ "predict_overall_f1": 0.8280542986425339,
33
+ "predict_overall_precision": 0.8318181818181818,
34
+ "predict_overall_recall": 0.8243243243243243,
35
+ "predict_runtime": 19.7294,
36
+ "predict_samples_per_second": 1.825,
37
+ "predict_steps_per_second": 0.912
38
+ }
predict_results_groups.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "predict_loss": 0.1715492457151413,
3
+ "predict_DATE_precision": 1.0,
4
+ "predict_DATE_recall": 0.9444444444444444,
5
+ "predict_DATE_f1": 0.9714285714285714,
6
+ "predict_DATE_number": 18,
7
+ "predict_LOCATION_precision": 0.6666666666666666,
8
+ "predict_LOCATION_recall": 0.6666666666666666,
9
+ "predict_LOCATION_f1": 0.6666666666666666,
10
+ "predict_LOCATION_number": 9,
11
+ "predict_ORGANIZATION_precision": 0.961038961038961,
12
+ "predict_ORGANIZATION_recall": 0.9487179487179487,
13
+ "predict_ORGANIZATION_f1": 0.9548387096774195,
14
+ "predict_ORGANIZATION_number": 78,
15
+ "predict_OTHER_precision": 0.5185185185185185,
16
+ "predict_OTHER_recall": 0.5833333333333334,
17
+ "predict_OTHER_f1": 0.5490196078431373,
18
+ "predict_OTHER_number": 24,
19
+ "predict_PERSON_precision": 0.8679245283018868,
20
+ "predict_PERSON_recall": 0.8363636363636363,
21
+ "predict_PERSON_f1": 0.8518518518518519,
22
+ "predict_PERSON_number": 55,
23
+ "predict_QUANTITY_precision": 0.8333333333333334,
24
+ "predict_QUANTITY_recall": 0.8928571428571429,
25
+ "predict_QUANTITY_f1": 0.8620689655172413,
26
+ "predict_QUANTITY_number": 28,
27
+ "predict_TITLE_precision": 0.14285714285714285,
28
+ "predict_TITLE_recall": 0.1,
29
+ "predict_TITLE_f1": 0.11764705882352941,
30
+ "predict_TITLE_number": 10,
31
+ "predict_overall_precision": 0.8318181818181818,
32
+ "predict_overall_recall": 0.8243243243243243,
33
+ "predict_overall_f1": 0.8280542986425339,
34
+ "predict_overall_accuracy": 0.968184311574328,
35
+ "predict_runtime": 19.7294,
36
+ "predict_samples_per_second": 1.825,
37
+ "predict_steps_per_second": 0.912,
38
+ "model_name": "t5-large",
39
+ "dataset_name": "imvladikon/ner_news_small",
40
+ "split_instructions": "80:10:10",
41
+ "device": "cpu",
42
+ "label2id": {
43
+ "O": 0,
44
+ "B-PERSON": 1,
45
+ "I-PERSON": 2,
46
+ "B-ORGANIZATION": 3,
47
+ "I-ORGANIZATION": 4,
48
+ "B-LOCATION": 5,
49
+ "I-LOCATION": 6,
50
+ "B-TITLE": 7,
51
+ "I-TITLE": 8,
52
+ "B-EVENT": 9,
53
+ "I-EVENT": 10,
54
+ "B-QUANTITY": 11,
55
+ "I-QUANTITY": 12,
56
+ "B-DATE": 13,
57
+ "I-DATE": 14,
58
+ "B-COMMERCIAL_ITEM": 15,
59
+ "I-COMMERCIAL_ITEM": 16,
60
+ "B-OTHER": 17,
61
+ "I-OTHER": 18
62
+ }
63
+ }
predict_sk_classification_report.csv ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ,tag,precision,recall,f1-score,support,entity
2
+ 0,B-DATE,1.0,0.9444444444444444,0.9714285714285714,18.0,DATE
3
+ 7,I-DATE,1.0,1.0,1.0,14.0,DATE
4
+ 8,I-LOCATION,0.2,1.0,0.33333333333333337,1.0,LOCATION
5
+ 1,B-LOCATION,0.6666666666666666,0.6666666666666666,0.6666666666666666,9.0,LOCATION
6
+ 14,O,0.9836879432624114,0.9900071377587437,0.9868374244041266,1401.0,O
7
+ 9,I-ORGANIZATION,1.0,0.9888888888888889,0.9944134078212291,90.0,ORGANIZATION
8
+ 2,B-ORGANIZATION,0.974025974025974,0.9615384615384616,0.967741935483871,78.0,ORGANIZATION
9
+ 3,B-OTHER,0.6296296296296297,0.7083333333333334,0.6666666666666667,24.0,OTHER
10
+ 10,I-OTHER,1.0,0.3,0.4615384615384615,10.0,OTHER
11
+ 4,B-PERSON,0.9433962264150944,0.9090909090909091,0.9259259259259259,55.0,PERSON
12
+ 11,I-PERSON,0.8181818181818182,1.0,0.9,27.0,PERSON
13
+ 5,B-QUANTITY,0.9655172413793104,1.0,0.9824561403508771,28.0,QUANTITY
14
+ 12,I-QUANTITY,0.9487179487179487,0.9024390243902439,0.9249999999999999,41.0,QUANTITY
15
+ 6,B-TITLE,0.5,0.1,0.16666666666666669,10.0,TITLE
16
+ 13,I-TITLE,0.8125,0.7647058823529411,0.787878787878788,17.0,TITLE
17
+ 15,accuracy,0.968184311574328,0.968184311574328,0.968184311574328,0.968184311574328,accuracy
18
+ 16,macro avg,0.829488229885257,0.8157409832309757,0.7824369325443457,1823.0,macro avg
19
+ 17,weighted avg,0.968817370156649,0.968184311574328,0.9663713907913454,1823.0,weighted avg
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dce6c249ae9cfdd142837c7d4e65e4e5868dd78ef7c1be8f3ae6ce013abd4084
3
+ size 1339899703
special_tokens_map.json ADDED
@@ -0,0 +1,108 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<extra_id_0>",
4
+ "<extra_id_1>",
5
+ "<extra_id_2>",
6
+ "<extra_id_3>",
7
+ "<extra_id_4>",
8
+ "<extra_id_5>",
9
+ "<extra_id_6>",
10
+ "<extra_id_7>",
11
+ "<extra_id_8>",
12
+ "<extra_id_9>",
13
+ "<extra_id_10>",
14
+ "<extra_id_11>",
15
+ "<extra_id_12>",
16
+ "<extra_id_13>",
17
+ "<extra_id_14>",
18
+ "<extra_id_15>",
19
+ "<extra_id_16>",
20
+ "<extra_id_17>",
21
+ "<extra_id_18>",
22
+ "<extra_id_19>",
23
+ "<extra_id_20>",
24
+ "<extra_id_21>",
25
+ "<extra_id_22>",
26
+ "<extra_id_23>",
27
+ "<extra_id_24>",
28
+ "<extra_id_25>",
29
+ "<extra_id_26>",
30
+ "<extra_id_27>",
31
+ "<extra_id_28>",
32
+ "<extra_id_29>",
33
+ "<extra_id_30>",
34
+ "<extra_id_31>",
35
+ "<extra_id_32>",
36
+ "<extra_id_33>",
37
+ "<extra_id_34>",
38
+ "<extra_id_35>",
39
+ "<extra_id_36>",
40
+ "<extra_id_37>",
41
+ "<extra_id_38>",
42
+ "<extra_id_39>",
43
+ "<extra_id_40>",
44
+ "<extra_id_41>",
45
+ "<extra_id_42>",
46
+ "<extra_id_43>",
47
+ "<extra_id_44>",
48
+ "<extra_id_45>",
49
+ "<extra_id_46>",
50
+ "<extra_id_47>",
51
+ "<extra_id_48>",
52
+ "<extra_id_49>",
53
+ "<extra_id_50>",
54
+ "<extra_id_51>",
55
+ "<extra_id_52>",
56
+ "<extra_id_53>",
57
+ "<extra_id_54>",
58
+ "<extra_id_55>",
59
+ "<extra_id_56>",
60
+ "<extra_id_57>",
61
+ "<extra_id_58>",
62
+ "<extra_id_59>",
63
+ "<extra_id_60>",
64
+ "<extra_id_61>",
65
+ "<extra_id_62>",
66
+ "<extra_id_63>",
67
+ "<extra_id_64>",
68
+ "<extra_id_65>",
69
+ "<extra_id_66>",
70
+ "<extra_id_67>",
71
+ "<extra_id_68>",
72
+ "<extra_id_69>",
73
+ "<extra_id_70>",
74
+ "<extra_id_71>",
75
+ "<extra_id_72>",
76
+ "<extra_id_73>",
77
+ "<extra_id_74>",
78
+ "<extra_id_75>",
79
+ "<extra_id_76>",
80
+ "<extra_id_77>",
81
+ "<extra_id_78>",
82
+ "<extra_id_79>",
83
+ "<extra_id_80>",
84
+ "<extra_id_81>",
85
+ "<extra_id_82>",
86
+ "<extra_id_83>",
87
+ "<extra_id_84>",
88
+ "<extra_id_85>",
89
+ "<extra_id_86>",
90
+ "<extra_id_87>",
91
+ "<extra_id_88>",
92
+ "<extra_id_89>",
93
+ "<extra_id_90>",
94
+ "<extra_id_91>",
95
+ "<extra_id_92>",
96
+ "<extra_id_93>",
97
+ "<extra_id_94>",
98
+ "<extra_id_95>",
99
+ "<extra_id_96>",
100
+ "<extra_id_97>",
101
+ "<extra_id_98>",
102
+ "<extra_id_99>"
103
+ ],
104
+ "bos_token": "<s>",
105
+ "eos_token": "</s>",
106
+ "pad_token": "<pad>",
107
+ "unk_token": "<unk>"
108
+ }
spiece.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d60acb128cf7b7f2536e8f38a5b18a05535c9e14c7a355904270e15b0945ea86
3
+ size 791656
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,115 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<extra_id_0>",
4
+ "<extra_id_1>",
5
+ "<extra_id_2>",
6
+ "<extra_id_3>",
7
+ "<extra_id_4>",
8
+ "<extra_id_5>",
9
+ "<extra_id_6>",
10
+ "<extra_id_7>",
11
+ "<extra_id_8>",
12
+ "<extra_id_9>",
13
+ "<extra_id_10>",
14
+ "<extra_id_11>",
15
+ "<extra_id_12>",
16
+ "<extra_id_13>",
17
+ "<extra_id_14>",
18
+ "<extra_id_15>",
19
+ "<extra_id_16>",
20
+ "<extra_id_17>",
21
+ "<extra_id_18>",
22
+ "<extra_id_19>",
23
+ "<extra_id_20>",
24
+ "<extra_id_21>",
25
+ "<extra_id_22>",
26
+ "<extra_id_23>",
27
+ "<extra_id_24>",
28
+ "<extra_id_25>",
29
+ "<extra_id_26>",
30
+ "<extra_id_27>",
31
+ "<extra_id_28>",
32
+ "<extra_id_29>",
33
+ "<extra_id_30>",
34
+ "<extra_id_31>",
35
+ "<extra_id_32>",
36
+ "<extra_id_33>",
37
+ "<extra_id_34>",
38
+ "<extra_id_35>",
39
+ "<extra_id_36>",
40
+ "<extra_id_37>",
41
+ "<extra_id_38>",
42
+ "<extra_id_39>",
43
+ "<extra_id_40>",
44
+ "<extra_id_41>",
45
+ "<extra_id_42>",
46
+ "<extra_id_43>",
47
+ "<extra_id_44>",
48
+ "<extra_id_45>",
49
+ "<extra_id_46>",
50
+ "<extra_id_47>",
51
+ "<extra_id_48>",
52
+ "<extra_id_49>",
53
+ "<extra_id_50>",
54
+ "<extra_id_51>",
55
+ "<extra_id_52>",
56
+ "<extra_id_53>",
57
+ "<extra_id_54>",
58
+ "<extra_id_55>",
59
+ "<extra_id_56>",
60
+ "<extra_id_57>",
61
+ "<extra_id_58>",
62
+ "<extra_id_59>",
63
+ "<extra_id_60>",
64
+ "<extra_id_61>",
65
+ "<extra_id_62>",
66
+ "<extra_id_63>",
67
+ "<extra_id_64>",
68
+ "<extra_id_65>",
69
+ "<extra_id_66>",
70
+ "<extra_id_67>",
71
+ "<extra_id_68>",
72
+ "<extra_id_69>",
73
+ "<extra_id_70>",
74
+ "<extra_id_71>",
75
+ "<extra_id_72>",
76
+ "<extra_id_73>",
77
+ "<extra_id_74>",
78
+ "<extra_id_75>",
79
+ "<extra_id_76>",
80
+ "<extra_id_77>",
81
+ "<extra_id_78>",
82
+ "<extra_id_79>",
83
+ "<extra_id_80>",
84
+ "<extra_id_81>",
85
+ "<extra_id_82>",
86
+ "<extra_id_83>",
87
+ "<extra_id_84>",
88
+ "<extra_id_85>",
89
+ "<extra_id_86>",
90
+ "<extra_id_87>",
91
+ "<extra_id_88>",
92
+ "<extra_id_89>",
93
+ "<extra_id_90>",
94
+ "<extra_id_91>",
95
+ "<extra_id_92>",
96
+ "<extra_id_93>",
97
+ "<extra_id_94>",
98
+ "<extra_id_95>",
99
+ "<extra_id_96>",
100
+ "<extra_id_97>",
101
+ "<extra_id_98>",
102
+ "<extra_id_99>"
103
+ ],
104
+ "bos_token": "<s>",
105
+ "eos_token": "</s>",
106
+ "extra_ids": 100,
107
+ "model_max_length": 512,
108
+ "name_or_path": "t5-large",
109
+ "pad_token": "<pad>",
110
+ "sp_model_kwargs": {},
111
+ "special_tokens_map_file": null,
112
+ "tokenizer_class": "EncT5Tokenizer",
113
+ "unk_token": "<unk>",
114
+ "use_fast": true
115
+ }
train_classification_report.csv ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ,tag,precision,recall,f1-score,support
2
+ 0,COMMERCIAL_ITEM,0.0,0.0,0.0,7.0
3
+ 1,DATE,0.9620786516853933,0.9771754636233951,0.9695682944090589,701.0
4
+ 2,EVENT,0.0,0.0,0.0,22.0
5
+ 3,LOCATION,0.8862461220268872,0.8917793964620188,0.8890041493775933,961.0
6
+ 4,ORGANIZATION,0.9199203187250996,0.9355753646677472,0.9276817999196465,2468.0
7
+ 5,OTHER,0.5351598173515981,0.541089566020314,0.5381083562901744,1083.0
8
+ 6,PERSON,0.8979591836734694,0.9413388543823327,0.9191374663072777,1449.0
9
+ 7,QUANTITY,0.8873873873873874,0.9231490159325211,0.9049150206706478,1067.0
10
+ 8,TITLE,0.22058823529411764,0.17045454545454544,0.19230769230769232,264.0
11
+ 9,micro avg,0.8415670814340274,0.8515332834704562,0.8465208501146292,8022.0
12
+ 10,macro avg,0.5899266351271059,0.5978402451714305,0.5934136421424545,8022.0
13
+ 11,weighted avg,0.8329919696469051,0.8515332834704562,0.8419887983670683,8022.0
train_classification_report.json ADDED
@@ -0,0 +1,74 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "COMMERCIAL_ITEM": {
3
+ "precision": 0.0,
4
+ "recall": 0.0,
5
+ "f1-score": 0.0,
6
+ "support": 7.0
7
+ },
8
+ "DATE": {
9
+ "precision": 0.9620786516853933,
10
+ "recall": 0.9771754636233951,
11
+ "f1-score": 0.9695682944090589,
12
+ "support": 701.0
13
+ },
14
+ "EVENT": {
15
+ "precision": 0.0,
16
+ "recall": 0.0,
17
+ "f1-score": 0.0,
18
+ "support": 22.0
19
+ },
20
+ "LOCATION": {
21
+ "precision": 0.8862461220268872,
22
+ "recall": 0.8917793964620188,
23
+ "f1-score": 0.8890041493775933,
24
+ "support": 961.0
25
+ },
26
+ "ORGANIZATION": {
27
+ "precision": 0.9199203187250996,
28
+ "recall": 0.9355753646677472,
29
+ "f1-score": 0.9276817999196465,
30
+ "support": 2468.0
31
+ },
32
+ "OTHER": {
33
+ "precision": 0.5351598173515981,
34
+ "recall": 0.541089566020314,
35
+ "f1-score": 0.5381083562901744,
36
+ "support": 1083.0
37
+ },
38
+ "PERSON": {
39
+ "precision": 0.8979591836734694,
40
+ "recall": 0.9413388543823327,
41
+ "f1-score": 0.9191374663072777,
42
+ "support": 1449.0
43
+ },
44
+ "QUANTITY": {
45
+ "precision": 0.8873873873873874,
46
+ "recall": 0.9231490159325211,
47
+ "f1-score": 0.9049150206706478,
48
+ "support": 1067.0
49
+ },
50
+ "TITLE": {
51
+ "precision": 0.22058823529411764,
52
+ "recall": 0.17045454545454544,
53
+ "f1-score": 0.19230769230769232,
54
+ "support": 264.0
55
+ },
56
+ "micro avg": {
57
+ "precision": 0.8415670814340274,
58
+ "recall": 0.8515332834704562,
59
+ "f1-score": 0.8465208501146292,
60
+ "support": 8022.0
61
+ },
62
+ "macro avg": {
63
+ "precision": 0.5899266351271059,
64
+ "recall": 0.5978402451714305,
65
+ "f1-score": 0.5934136421424545,
66
+ "support": 8022.0
67
+ },
68
+ "weighted avg": {
69
+ "precision": 0.8329919696469051,
70
+ "recall": 0.8515332834704562,
71
+ "f1-score": 0.8419887983670683,
72
+ "support": 8022.0
73
+ }
74
+ }
train_confusion_matrix.csv ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ,Actual\Predicted,DATE,LOCATION,O,ORGANIZATION,OTHER,PERSON,QUANTITY,TITLE,All
2
+ 0,COMMERCIAL_ITEM,0,0,2,1,5,0,0,6,14
3
+ 1,DATE,1290,0,15,0,0,0,3,1,1309
4
+ 2,EVENT,20,4,2,10,3,0,0,4,43
5
+ 3,LOCATION,0,1403,23,34,24,9,2,1,1496
6
+ 4,O,7,29,56717,40,204,20,81,22,57120
7
+ 5,ORGANIZATION,0,48,48,4871,7,30,0,6,5010
8
+ 6,OTHER,0,20,570,9,1142,1,6,68,1816
9
+ 7,PERSON,0,2,17,30,5,2320,0,11,2385
10
+ 8,QUANTITY,2,0,55,0,2,0,2523,0,2582
11
+ 9,TITLE,0,4,102,70,53,105,0,430,764
12
+ 10,All,1319,1510,57551,5065,1445,2485,2615,549,72539
train_results.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train_COMMERCIAL_ITEM_f1": 0.0,
3
+ "train_COMMERCIAL_ITEM_number": 7,
4
+ "train_COMMERCIAL_ITEM_precision": 0.0,
5
+ "train_COMMERCIAL_ITEM_recall": 0.0,
6
+ "train_DATE_f1": 0.9695682944090589,
7
+ "train_DATE_number": 701,
8
+ "train_DATE_precision": 0.9620786516853933,
9
+ "train_DATE_recall": 0.9771754636233951,
10
+ "train_EVENT_f1": 0.0,
11
+ "train_EVENT_number": 22,
12
+ "train_EVENT_precision": 0.0,
13
+ "train_EVENT_recall": 0.0,
14
+ "train_LOCATION_f1": 0.8890041493775933,
15
+ "train_LOCATION_number": 961,
16
+ "train_LOCATION_precision": 0.8862461220268872,
17
+ "train_LOCATION_recall": 0.8917793964620188,
18
+ "train_ORGANIZATION_f1": 0.9276817999196465,
19
+ "train_ORGANIZATION_number": 2468,
20
+ "train_ORGANIZATION_precision": 0.9199203187250996,
21
+ "train_ORGANIZATION_recall": 0.9355753646677472,
22
+ "train_OTHER_f1": 0.5381083562901744,
23
+ "train_OTHER_number": 1083,
24
+ "train_OTHER_precision": 0.5351598173515981,
25
+ "train_OTHER_recall": 0.541089566020314,
26
+ "train_PERSON_f1": 0.9191374663072777,
27
+ "train_PERSON_number": 1449,
28
+ "train_PERSON_precision": 0.8979591836734694,
29
+ "train_PERSON_recall": 0.9413388543823327,
30
+ "train_QUANTITY_f1": 0.9049150206706478,
31
+ "train_QUANTITY_number": 1067,
32
+ "train_QUANTITY_precision": 0.8873873873873874,
33
+ "train_QUANTITY_recall": 0.9231490159325211,
34
+ "train_TITLE_f1": 0.19230769230769232,
35
+ "train_TITLE_number": 264,
36
+ "train_TITLE_precision": 0.22058823529411764,
37
+ "train_TITLE_recall": 0.17045454545454544,
38
+ "train_loss": 0.14113017916679382,
39
+ "train_overall_accuracy": 0.9713395552737148,
40
+ "train_overall_f1": 0.8465208501146292,
41
+ "train_overall_precision": 0.8415670814340274,
42
+ "train_overall_recall": 0.8515332834704562,
43
+ "train_runtime": 701.7841,
44
+ "train_samples_per_second": 2.018,
45
+ "train_steps_per_second": 1.009
46
+ }
train_results_groups.json ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train_loss": 0.14113017916679382,
3
+ "train_COMMERCIAL_ITEM_precision": 0.0,
4
+ "train_COMMERCIAL_ITEM_recall": 0.0,
5
+ "train_COMMERCIAL_ITEM_f1": 0.0,
6
+ "train_COMMERCIAL_ITEM_number": 7,
7
+ "train_DATE_precision": 0.9620786516853933,
8
+ "train_DATE_recall": 0.9771754636233951,
9
+ "train_DATE_f1": 0.9695682944090589,
10
+ "train_DATE_number": 701,
11
+ "train_EVENT_precision": 0.0,
12
+ "train_EVENT_recall": 0.0,
13
+ "train_EVENT_f1": 0.0,
14
+ "train_EVENT_number": 22,
15
+ "train_LOCATION_precision": 0.8862461220268872,
16
+ "train_LOCATION_recall": 0.8917793964620188,
17
+ "train_LOCATION_f1": 0.8890041493775933,
18
+ "train_LOCATION_number": 961,
19
+ "train_ORGANIZATION_precision": 0.9199203187250996,
20
+ "train_ORGANIZATION_recall": 0.9355753646677472,
21
+ "train_ORGANIZATION_f1": 0.9276817999196465,
22
+ "train_ORGANIZATION_number": 2468,
23
+ "train_OTHER_precision": 0.5351598173515981,
24
+ "train_OTHER_recall": 0.541089566020314,
25
+ "train_OTHER_f1": 0.5381083562901744,
26
+ "train_OTHER_number": 1083,
27
+ "train_PERSON_precision": 0.8979591836734694,
28
+ "train_PERSON_recall": 0.9413388543823327,
29
+ "train_PERSON_f1": 0.9191374663072777,
30
+ "train_PERSON_number": 1449,
31
+ "train_QUANTITY_precision": 0.8873873873873874,
32
+ "train_QUANTITY_recall": 0.9231490159325211,
33
+ "train_QUANTITY_f1": 0.9049150206706478,
34
+ "train_QUANTITY_number": 1067,
35
+ "train_TITLE_precision": 0.22058823529411764,
36
+ "train_TITLE_recall": 0.17045454545454544,
37
+ "train_TITLE_f1": 0.19230769230769232,
38
+ "train_TITLE_number": 264,
39
+ "train_overall_precision": 0.8415670814340274,
40
+ "train_overall_recall": 0.8515332834704562,
41
+ "train_overall_f1": 0.8465208501146292,
42
+ "train_overall_accuracy": 0.9713395552737148,
43
+ "train_runtime": 701.7841,
44
+ "train_samples_per_second": 2.018,
45
+ "train_steps_per_second": 1.009,
46
+ "model_name": "t5-large",
47
+ "dataset_name": "imvladikon/ner_news_small",
48
+ "split_instructions": "80:10:10",
49
+ "device": "cpu",
50
+ "label2id": {
51
+ "O": 0,
52
+ "B-PERSON": 1,
53
+ "I-PERSON": 2,
54
+ "B-ORGANIZATION": 3,
55
+ "I-ORGANIZATION": 4,
56
+ "B-LOCATION": 5,
57
+ "I-LOCATION": 6,
58
+ "B-TITLE": 7,
59
+ "I-TITLE": 8,
60
+ "B-EVENT": 9,
61
+ "I-EVENT": 10,
62
+ "B-QUANTITY": 11,
63
+ "I-QUANTITY": 12,
64
+ "B-DATE": 13,
65
+ "I-DATE": 14,
66
+ "B-COMMERCIAL_ITEM": 15,
67
+ "I-COMMERCIAL_ITEM": 16,
68
+ "B-OTHER": 17,
69
+ "I-OTHER": 18
70
+ }
71
+ }
train_sk_classification_report.csv ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ,tag,precision,recall,f1-score,support,entity
2
+ 0,B-COMMERCIAL_ITEM,0.0,0.0,0.0,7.0,COMMERCIAL_ITEM
3
+ 9,I-COMMERCIAL_ITEM,0.0,0.0,0.0,7.0,COMMERCIAL_ITEM
4
+ 10,I-DATE,0.9722675367047309,0.9802631578947368,0.9762489762489763,608.0,DATE
5
+ 1,B-DATE,0.9745042492917847,0.9814550641940085,0.9779673063255152,701.0,DATE
6
+ 2,B-EVENT,0.0,0.0,0.0,22.0,EVENT
7
+ 11,I-EVENT,0.0,0.0,0.0,21.0,EVENT
8
+ 3,B-LOCATION,0.9235602094240838,0.9177939646201873,0.9206680584551149,961.0,LOCATION
9
+ 12,I-LOCATION,0.8936936936936937,0.9271028037383178,0.9100917431192661,535.0,LOCATION
10
+ 18,O,0.98550850549947,0.9929446778711485,0.9892126169650566,57120.0,O
11
+ 4,B-ORGANIZATION,0.9539765845781187,0.9574554294975689,0.9557128412537916,2468.0,ORGANIZATION
12
+ 13,I-ORGANIZATION,0.9582689335394127,0.975609756097561,0.9668615984405458,2542.0,ORGANIZATION
13
+ 5,B-OTHER,0.6866666666666666,0.6657433056325023,0.6760431317393343,1083.0,OTHER
14
+ 14,I-OTHER,0.7848101265822784,0.4229195088676671,0.549645390070922,733.0,OTHER
15
+ 6,B-PERSON,0.919973544973545,0.9599723947550034,0.939547450185748,1449.0,PERSON
16
+ 15,I-PERSON,0.9301130524152107,0.9668803418803419,0.9481403876375065,936.0,PERSON
17
+ 16,I-QUANTITY,0.9615635179153095,0.9742574257425742,0.9678688524590163,1515.0,QUANTITY
18
+ 7,B-QUANTITY,0.9472222222222222,0.9587628865979382,0.952957615277131,1067.0,QUANTITY
19
+ 8,B-TITLE,1.0,0.1553030303030303,0.26885245901639343,264.0,TITLE
20
+ 17,I-TITLE,0.7303149606299213,0.742,0.7361111111111112,500.0,TITLE
21
+ 19,accuracy,0.9713395552737148,0.9713395552737148,0.9713395552737148,0.9713395552737148,accuracy
22
+ 20,macro avg,0.7169707265334972,0.6620244077732942,0.6703120809634436,72539.0,macro avg
23
+ 21,weighted avg,0.9697084179539044,0.9713395552737148,0.9688185642893826,72539.0,weighted avg
trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:277a580730d648f625454730fe2ef2131b5df97408b5035b4a8eca0ebc6a4865
3
+ size 3311