haryoaw commited on
Commit
8a722d3
1 Parent(s): dc62731

Initial Commit

Browse files
Files changed (5) hide show
  1. README.md +32 -60
  2. config.json +14 -21
  3. eval_result_ner.json +1 -1
  4. model.safetensors +2 -2
  5. training_args.bin +1 -1
README.md CHANGED
@@ -1,14 +1,14 @@
1
  ---
2
- base_model: microsoft/mdeberta-v3-base
3
  library_name: transformers
4
  license: mit
 
 
 
5
  metrics:
6
  - precision
7
  - recall
8
  - f1
9
  - accuracy
10
- tags:
11
- - generated_from_trainer
12
  model-index:
13
  - name: scenario-kd-pre-ner-full_data-univner_full66
14
  results: []
@@ -19,13 +19,13 @@ should probably proofread and complete it, then remove this comment. -->
19
 
20
  # scenario-kd-pre-ner-full_data-univner_full66
21
 
22
- This model is a fine-tuned version of [microsoft/mdeberta-v3-base](https://huggingface.co/microsoft/mdeberta-v3-base) on the None dataset.
23
  It achieves the following results on the evaluation set:
24
- - Loss: 0.3678
25
- - Precision: 0.7980
26
- - Recall: 0.7563
27
- - F1: 0.7766
28
- - Accuracy: 0.9766
29
 
30
  ## Model description
31
 
@@ -56,57 +56,29 @@ The following hyperparameters were used during training:
56
 
57
  | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
58
  |:-------------:|:-------:|:-----:|:---------------:|:---------:|:------:|:------:|:--------:|
59
- | 1.5996 | 0.5828 | 500 | 0.9922 | 0.3920 | 0.2627 | 0.3146 | 0.9378 |
60
- | 0.8451 | 1.1655 | 1000 | 0.6919 | 0.5664 | 0.5963 | 0.5810 | 0.9599 |
61
- | 0.5868 | 1.7483 | 1500 | 0.5945 | 0.6779 | 0.6454 | 0.6612 | 0.9663 |
62
- | 0.4643 | 2.3310 | 2000 | 0.5350 | 0.7058 | 0.6705 | 0.6877 | 0.9687 |
63
- | 0.4026 | 2.9138 | 2500 | 0.5062 | 0.7062 | 0.7003 | 0.7033 | 0.9700 |
64
- | 0.3341 | 3.4965 | 3000 | 0.4982 | 0.6936 | 0.7339 | 0.7132 | 0.9705 |
65
- | 0.3079 | 4.0793 | 3500 | 0.4864 | 0.7396 | 0.7003 | 0.7194 | 0.9715 |
66
- | 0.266 | 4.6620 | 4000 | 0.4833 | 0.7660 | 0.6885 | 0.7252 | 0.9722 |
67
- | 0.2438 | 5.2448 | 4500 | 0.4747 | 0.7459 | 0.7099 | 0.7274 | 0.9724 |
68
- | 0.2207 | 5.8275 | 5000 | 0.4751 | 0.7579 | 0.6881 | 0.7213 | 0.9716 |
69
- | 0.2098 | 6.4103 | 5500 | 0.4510 | 0.7454 | 0.7238 | 0.7344 | 0.9728 |
70
- | 0.1974 | 6.9930 | 6000 | 0.4512 | 0.7473 | 0.7309 | 0.7390 | 0.9729 |
71
- | 0.1812 | 7.5758 | 6500 | 0.4305 | 0.7440 | 0.7383 | 0.7411 | 0.9736 |
72
- | 0.1726 | 8.1585 | 7000 | 0.4477 | 0.7544 | 0.7296 | 0.7418 | 0.9731 |
73
- | 0.1637 | 8.7413 | 7500 | 0.4394 | 0.7449 | 0.7501 | 0.7475 | 0.9743 |
74
- | 0.1561 | 9.3240 | 8000 | 0.4380 | 0.7669 | 0.7254 | 0.7456 | 0.9739 |
75
- | 0.1524 | 9.9068 | 8500 | 0.4229 | 0.7694 | 0.7448 | 0.7569 | 0.9751 |
76
- | 0.1453 | 10.4895 | 9000 | 0.4319 | 0.7654 | 0.7387 | 0.7518 | 0.9746 |
77
- | 0.1423 | 11.0723 | 9500 | 0.4307 | 0.7766 | 0.7301 | 0.7526 | 0.9743 |
78
- | 0.1361 | 11.6550 | 10000 | 0.4322 | 0.7788 | 0.7236 | 0.7502 | 0.9741 |
79
- | 0.1346 | 12.2378 | 10500 | 0.4324 | 0.7841 | 0.7184 | 0.7498 | 0.9740 |
80
- | 0.128 | 12.8205 | 11000 | 0.4089 | 0.7685 | 0.7599 | 0.7642 | 0.9753 |
81
- | 0.1271 | 13.4033 | 11500 | 0.4213 | 0.7856 | 0.7143 | 0.7483 | 0.9741 |
82
- | 0.1247 | 13.9860 | 12000 | 0.4140 | 0.7799 | 0.7332 | 0.7559 | 0.9748 |
83
- | 0.1213 | 14.5688 | 12500 | 0.4017 | 0.7700 | 0.7549 | 0.7623 | 0.9754 |
84
- | 0.1172 | 15.1515 | 13000 | 0.4140 | 0.7800 | 0.7399 | 0.7594 | 0.9748 |
85
- | 0.1178 | 15.7343 | 13500 | 0.3935 | 0.7822 | 0.7490 | 0.7652 | 0.9755 |
86
- | 0.1154 | 16.3170 | 14000 | 0.4041 | 0.7915 | 0.7244 | 0.7565 | 0.9750 |
87
- | 0.1137 | 16.8998 | 14500 | 0.3943 | 0.7823 | 0.7498 | 0.7657 | 0.9759 |
88
- | 0.1115 | 17.4825 | 15000 | 0.3853 | 0.7832 | 0.7537 | 0.7682 | 0.9759 |
89
- | 0.1089 | 18.0653 | 15500 | 0.3902 | 0.7816 | 0.7539 | 0.7675 | 0.9756 |
90
- | 0.1068 | 18.6480 | 16000 | 0.3936 | 0.7766 | 0.7605 | 0.7685 | 0.9760 |
91
- | 0.1074 | 19.2308 | 16500 | 0.3786 | 0.7837 | 0.7660 | 0.7748 | 0.9765 |
92
- | 0.1036 | 19.8135 | 17000 | 0.3892 | 0.7869 | 0.7331 | 0.7590 | 0.9755 |
93
- | 0.1058 | 20.3963 | 17500 | 0.3897 | 0.7845 | 0.7513 | 0.7675 | 0.9757 |
94
- | 0.1026 | 20.9790 | 18000 | 0.3869 | 0.7803 | 0.7553 | 0.7676 | 0.9758 |
95
- | 0.1021 | 21.5618 | 18500 | 0.3855 | 0.7866 | 0.7478 | 0.7667 | 0.9759 |
96
- | 0.1007 | 22.1445 | 19000 | 0.3866 | 0.7921 | 0.7266 | 0.7579 | 0.9752 |
97
- | 0.0999 | 22.7273 | 19500 | 0.3811 | 0.7832 | 0.7552 | 0.7689 | 0.9758 |
98
- | 0.0994 | 23.3100 | 20000 | 0.3806 | 0.7896 | 0.7485 | 0.7685 | 0.9761 |
99
- | 0.0985 | 23.8928 | 20500 | 0.3839 | 0.7909 | 0.7511 | 0.7705 | 0.9762 |
100
- | 0.0972 | 24.4755 | 21000 | 0.3742 | 0.7881 | 0.7513 | 0.7692 | 0.9761 |
101
- | 0.0974 | 25.0583 | 21500 | 0.3763 | 0.7942 | 0.7400 | 0.7662 | 0.9756 |
102
- | 0.0957 | 25.6410 | 22000 | 0.3766 | 0.7956 | 0.7534 | 0.7739 | 0.9764 |
103
- | 0.0961 | 26.2238 | 22500 | 0.3769 | 0.7970 | 0.7439 | 0.7696 | 0.9757 |
104
- | 0.0958 | 26.8065 | 23000 | 0.3752 | 0.7977 | 0.7449 | 0.7704 | 0.9759 |
105
- | 0.0955 | 27.3893 | 23500 | 0.3708 | 0.7887 | 0.7576 | 0.7728 | 0.9765 |
106
- | 0.0942 | 27.9720 | 24000 | 0.3709 | 0.7929 | 0.7503 | 0.7710 | 0.9760 |
107
- | 0.0941 | 28.5548 | 24500 | 0.3742 | 0.7915 | 0.7526 | 0.7715 | 0.9761 |
108
- | 0.0946 | 29.1375 | 25000 | 0.3722 | 0.7970 | 0.7553 | 0.7756 | 0.9766 |
109
- | 0.0939 | 29.7203 | 25500 | 0.3678 | 0.7980 | 0.7563 | 0.7766 | 0.9766 |
110
 
111
 
112
  ### Framework versions
 
1
  ---
 
2
  library_name: transformers
3
  license: mit
4
+ base_model: FacebookAI/xlm-roberta-base
5
+ tags:
6
+ - generated_from_trainer
7
  metrics:
8
  - precision
9
  - recall
10
  - f1
11
  - accuracy
 
 
12
  model-index:
13
  - name: scenario-kd-pre-ner-full_data-univner_full66
14
  results: []
 
19
 
20
  # scenario-kd-pre-ner-full_data-univner_full66
21
 
22
+ This model is a fine-tuned version of [FacebookAI/xlm-roberta-base](https://huggingface.co/FacebookAI/xlm-roberta-base) on the None dataset.
23
  It achieves the following results on the evaluation set:
24
+ - Loss: 0.5549
25
+ - Precision: 0.7660
26
+ - Recall: 0.7319
27
+ - F1: 0.7485
28
+ - Accuracy: 0.9802
29
 
30
  ## Model description
31
 
 
56
 
57
  | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
58
  |:-------------:|:-------:|:-----:|:---------------:|:---------:|:------:|:------:|:--------:|
59
+ | 1.2484 | 1.2755 | 500 | 0.8737 | 0.6792 | 0.5631 | 0.6157 | 0.9709 |
60
+ | 0.6459 | 2.5510 | 1000 | 0.7190 | 0.6926 | 0.6739 | 0.6831 | 0.9771 |
61
+ | 0.5071 | 3.8265 | 1500 | 0.6650 | 0.7076 | 0.6863 | 0.6968 | 0.9773 |
62
+ | 0.4233 | 5.1020 | 2000 | 0.6513 | 0.6933 | 0.7019 | 0.6975 | 0.9775 |
63
+ | 0.3655 | 6.3776 | 2500 | 0.6252 | 0.7421 | 0.6822 | 0.7109 | 0.9778 |
64
+ | 0.3251 | 7.6531 | 3000 | 0.6172 | 0.7412 | 0.7174 | 0.7291 | 0.9791 |
65
+ | 0.2963 | 8.9286 | 3500 | 0.6204 | 0.7143 | 0.6677 | 0.6902 | 0.9773 |
66
+ | 0.2699 | 10.2041 | 4000 | 0.5919 | 0.7310 | 0.7288 | 0.7299 | 0.9792 |
67
+ | 0.2469 | 11.4796 | 4500 | 0.6168 | 0.7560 | 0.6863 | 0.7195 | 0.9788 |
68
+ | 0.2313 | 12.7551 | 5000 | 0.5871 | 0.7353 | 0.7133 | 0.7241 | 0.9792 |
69
+ | 0.2148 | 14.0306 | 5500 | 0.5947 | 0.7358 | 0.7122 | 0.7238 | 0.9794 |
70
+ | 0.2022 | 15.3061 | 6000 | 0.5830 | 0.7298 | 0.7019 | 0.7156 | 0.9790 |
71
+ | 0.1933 | 16.5816 | 6500 | 0.5734 | 0.7427 | 0.7143 | 0.7282 | 0.9794 |
72
+ | 0.185 | 17.8571 | 7000 | 0.5814 | 0.7352 | 0.6957 | 0.7149 | 0.9792 |
73
+ | 0.1767 | 19.1327 | 7500 | 0.5670 | 0.7516 | 0.7236 | 0.7373 | 0.9797 |
74
+ | 0.1688 | 20.4082 | 8000 | 0.5770 | 0.7551 | 0.6957 | 0.7241 | 0.9791 |
75
+ | 0.1634 | 21.6837 | 8500 | 0.5621 | 0.7443 | 0.7143 | 0.7290 | 0.9792 |
76
+ | 0.1592 | 22.9592 | 9000 | 0.5691 | 0.7495 | 0.7091 | 0.7287 | 0.9790 |
77
+ | 0.1538 | 24.2347 | 9500 | 0.5557 | 0.7481 | 0.7195 | 0.7335 | 0.9802 |
78
+ | 0.1513 | 25.5102 | 10000 | 0.5687 | 0.7446 | 0.7091 | 0.7264 | 0.9791 |
79
+ | 0.1489 | 26.7857 | 10500 | 0.5554 | 0.7623 | 0.7236 | 0.7424 | 0.9801 |
80
+ | 0.145 | 28.0612 | 11000 | 0.5488 | 0.7564 | 0.7329 | 0.7445 | 0.9804 |
81
+ | 0.144 | 29.3367 | 11500 | 0.5549 | 0.7660 | 0.7319 | 0.7485 | 0.9802 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
82
 
83
 
84
  ### Framework versions
config.json CHANGED
@@ -1,9 +1,12 @@
1
  {
2
- "_name_or_path": "microsoft/mdeberta-v3-base",
3
  "architectures": [
4
- "DebertaForTokenClassificationKD"
5
  ],
6
  "attention_probs_dropout_prob": 0.1,
 
 
 
7
  "hidden_act": "gelu",
8
  "hidden_dropout_prob": 0.1,
9
  "hidden_size": 768,
@@ -27,27 +30,17 @@
27
  "LABEL_5": 5,
28
  "LABEL_6": 6
29
  },
30
- "layer_norm_eps": 1e-07,
31
- "max_position_embeddings": 512,
32
- "max_relative_positions": -1,
33
- "model_type": "deberta-v2",
34
- "norm_rel_ebd": "layer_norm",
35
  "num_attention_heads": 12,
36
  "num_hidden_layers": 6,
37
- "pad_token_id": 0,
38
- "pooler_dropout": 0,
39
- "pooler_hidden_act": "gelu",
40
- "pooler_hidden_size": 768,
41
- "pos_att_type": [
42
- "p2c",
43
- "c2p"
44
- ],
45
- "position_biased_input": false,
46
- "position_buckets": 256,
47
- "relative_attention": true,
48
- "share_att_key": true,
49
  "torch_dtype": "float32",
50
  "transformers_version": "4.44.2",
51
- "type_vocab_size": 0,
52
- "vocab_size": 251000
 
53
  }
 
1
  {
2
+ "_name_or_path": "FacebookAI/xlm-roberta-base",
3
  "architectures": [
4
+ "XLMRobertaForTokenClassificationKD"
5
  ],
6
  "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
  "hidden_act": "gelu",
11
  "hidden_dropout_prob": 0.1,
12
  "hidden_size": 768,
 
30
  "LABEL_5": 5,
31
  "LABEL_6": 6
32
  },
33
+ "layer_norm_eps": 1e-05,
34
+ "max_position_embeddings": 514,
35
+ "model_type": "xlm-roberta",
 
 
36
  "num_attention_heads": 12,
37
  "num_hidden_layers": 6,
38
+ "output_past": true,
39
+ "pad_token_id": 1,
40
+ "position_embedding_type": "absolute",
 
 
 
 
 
 
 
 
 
41
  "torch_dtype": "float32",
42
  "transformers_version": "4.44.2",
43
+ "type_vocab_size": 1,
44
+ "use_cache": true,
45
+ "vocab_size": 250002
46
  }
eval_result_ner.json CHANGED
@@ -1 +1 @@
1
- {"ceb_gja": {"precision": 0.6206896551724138, "recall": 0.7346938775510204, "f1": 0.6728971962616822, "accuracy": 0.9698841698841699}, "en_pud": {"precision": 0.7862903225806451, "recall": 0.7255813953488373, "f1": 0.7547169811320755, "accuracy": 0.9762939176426143}, "de_pud": {"precision": 0.72579001019368, "recall": 0.685274302213667, "f1": 0.704950495049505, "accuracy": 0.9687309549481975}, "pt_pud": {"precision": 0.7682317682317682, "recall": 0.6997270245677889, "f1": 0.7323809523809524, "accuracy": 0.9732985858931089}, "ru_pud": {"precision": 0.6291585127201565, "recall": 0.6206563706563707, "f1": 0.6248785228377066, "accuracy": 0.9628519762335314}, "sv_pud": {"precision": 0.8158179848320694, "recall": 0.7317784256559767, "f1": 0.7715163934426229, "accuracy": 0.9771440553575172}, "tl_trg": {"precision": 0.5517241379310345, "recall": 0.6956521739130435, "f1": 0.6153846153846154, "accuracy": 0.9741144414168937}, "tl_ugnayan": {"precision": 0.575, "recall": 0.696969696969697, "f1": 0.6301369863013698, "accuracy": 0.9699179580674567}, "zh_gsd": {"precision": 0.79702300405954, "recall": 0.7679269882659713, "f1": 0.7822045152722444, "accuracy": 0.972027972027972}, "zh_gsdsimp": {"precision": 0.8299039780521262, "recall": 0.7929226736566186, "f1": 0.8109919571045576, "accuracy": 0.9717782217782218}, "hr_set": {"precision": 0.872624912033779, "recall": 0.8838203848895224, "f1": 0.878186968838527, "accuracy": 0.984830997526793}, "da_ddt": {"precision": 0.793010752688172, "recall": 0.6599552572706935, "f1": 0.7203907203907204, "accuracy": 0.9788486481093485}, "en_ewt": {"precision": 0.7899901864573111, "recall": 0.7398897058823529, "f1": 0.7641196013289036, "accuracy": 0.9761326054906961}, "pt_bosque": {"precision": 0.7790927021696252, "recall": 0.6502057613168725, "f1": 0.708838043965904, "accuracy": 0.9713809592812636}, "sr_set": {"precision": 0.8906064209274673, "recall": 0.8842975206611571, "f1": 0.8874407582938388, "accuracy": 0.9832764206286665}, "sk_snk": {"precision": 0.7134146341463414, "recall": 0.639344262295082, "f1": 0.6743515850144092, "accuracy": 0.9583071608040201}, "sv_talbanken": {"precision": 0.8382352941176471, "recall": 0.8724489795918368, "f1": 0.855, "accuracy": 0.9972518035039505}}
 
1
+ {"ceb_gja": {"precision": 0.391304347826087, "recall": 0.5510204081632653, "f1": 0.4576271186440678, "accuracy": 0.9467181467181467}, "en_pud": {"precision": 0.7514734774066798, "recall": 0.7116279069767442, "f1": 0.73100812231247, "accuracy": 0.9743577635058557}, "de_pud": {"precision": 0.7108307045215563, "recall": 0.6506256015399422, "f1": 0.6793969849246232, "accuracy": 0.9660119075523885}, "pt_pud": {"precision": 0.7792746113989637, "recall": 0.6842584167424932, "f1": 0.7286821705426357, "accuracy": 0.9727431964796855}, "ru_pud": {"precision": 0.630939226519337, "recall": 0.5511583011583011, "f1": 0.5883565172591447, "accuracy": 0.9569103590803409}, "sv_pud": {"precision": 0.7871878393051032, "recall": 0.7045675413022352, "f1": 0.7435897435897436, "accuracy": 0.9741560075487523}, "tl_trg": {"precision": 0.5833333333333334, "recall": 0.6086956521739131, "f1": 0.5957446808510638, "accuracy": 0.9768392370572208}, "tl_ugnayan": {"precision": 0.5, "recall": 0.5757575757575758, "f1": 0.5352112676056339, "accuracy": 0.9626253418413856}, "zh_gsd": {"precision": 0.47767857142857145, "recall": 0.1395045632333768, "f1": 0.21594349142280525, "accuracy": 0.90001665001665}, "zh_gsdsimp": {"precision": 0.43315508021390375, "recall": 0.10615989515072084, "f1": 0.1705263157894737, "accuracy": 0.8986013986013986}, "hr_set": {"precision": 0.7674223341729639, "recall": 0.6514611546685674, "f1": 0.7047031611410948, "accuracy": 0.9631492168178071}, "da_ddt": {"precision": 0.7638888888888888, "recall": 0.6152125279642058, "f1": 0.6815365551425031, "accuracy": 0.9760550733313379}, "en_ewt": {"precision": 0.7879396984924623, "recall": 0.7205882352941176, "f1": 0.7527604416706672, "accuracy": 0.9752958520938758}, "pt_bosque": {"precision": 0.733402489626556, "recall": 0.5818930041152264, "f1": 0.6489215236346948, "accuracy": 0.9633024199391392}, "sr_set": {"precision": 0.7739251040221914, "recall": 0.6587957497048406, "f1": 0.711734693877551, "accuracy": 0.9558707643814027}, "sk_snk": {"precision": 0.6234817813765182, "recall": 0.5049180327868853, "f1": 0.5579710144927537, "accuracy": 0.9392273869346733}, "sv_talbanken": {"precision": 0.8009950248756219, "recall": 0.8214285714285714, "f1": 0.8110831234256928, "accuracy": 0.9961721548805025}}
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8d48a44e848c805d0e5386c482cd38378cac22e8aac8634f20ffcf93023ba43d
3
- size 944366708
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:524137227bc0f44fa866806452b581fbb218f41a5416f7c31180ebda4bc467b7
3
+ size 939737140
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a286fa5c4471b76d12b9555b09649216889aa5b66460a6692ed205046d18cae6
3
  size 5304
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0788c5ff239352ce6a821267905d965046999856d55e0b38e1d269163fb3cca4
3
  size 5304