Hugging Face's logo

orpe42
/
deberta_MP_dynamic

Text Classification
Transformers
TensorBoard
Safetensors
deberta-v2
Generated from Trainer
text-embeddings-inference
Model card Files
xet
 Metrics Community
orpe42 commited on
Commit
bd974cc
·
verified ·
1 Parent(s): 41304a6

Model save

Browse files
Files changed (10) hide show
  1. README.md +97 -0
  2. best_model/config.json +156 -0
  3. best_model/model.safetensors +3 -0
  4. best_model/tokenizer.json +0 -0
  5. best_model/tokenizer_config.json +24 -0
  6. best_model/training_args.bin +3 -0
  7. model.safetensors +1 -1
  8. runs/Aug06_12-10-47_erc-hpc-vm042/events.out.tfevents.1786031039.erc-hpc-vm042.307913.1 +3 -0
  9. tokenizer.json +0 -0
  10. tokenizer_config.json +24 -0
README.md ADDED
@@ -0,0 +1,97 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: mit
4
+ base_model: microsoft/deberta-v3-large
5
+ tags:
6
+ - generated_from_trainer
7
+ model-index:
8
+ - name: deberta_MP_dynamic
9
+ results: []
10
+ ---
11
+
12
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
+ should probably proofread and complete it, then remove this comment. -->
14
+
15
+ # deberta_MP_dynamic
16
+
17
+ This model is a fine-tuned version of [microsoft/deberta-v3-large](https://huggingface.co/microsoft/deberta-v3-large) on an unknown dataset.
18
+ It achieves the following results on the evaluation set:
19
+ - Loss: 0.0151
20
+ - Macro F1: 0.4028
21
+ - Micro F1: 0.5675
22
+ - Macro Precision: 0.7364
23
+ - Macro Recall: 0.3091
24
+ - Micro Precision: 0.8374
25
+ - Micro Recall: 0.4291
26
+ - Exact Match Ratio: 0.1092
27
+ - Macro Roc Auc: 0.9269
28
+
29
+ ## Model description
30
+
31
+ More information needed
32
+
33
+ ## Intended uses & limitations
34
+
35
+ More information needed
36
+
37
+ ## Training and evaluation data
38
+
39
+ More information needed
40
+
41
+ ## Training procedure
42
+
43
+ ### Training hyperparameters
44
+
45
+ The following hyperparameters were used during training:
46
+ - learning_rate: 2e-05
47
+ - train_batch_size: 16
48
+ - eval_batch_size: 16
49
+ - seed: 42
50
+ - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
51
+ - lr_scheduler_type: cosine
52
+ - lr_scheduler_warmup_steps: 0.1
53
+ - num_epochs: 200
54
+
55
+ ### Training results
56
+
57
+ | Training Loss | Epoch | Step | Validation Loss | Macro F1 | Micro F1 | Macro Precision | Macro Recall | Micro Precision | Micro Recall | Exact Match Ratio | Macro Roc Auc |
58
+ |:-------------:|:--------:|:-----:|:---------------:|:--------:|:--------:|:---------------:|:------------:|:---------------:|:------------:|:-----------------:|:-------------:|
59
+ | 0.0331 | 6.5789 | 500 | 0.0255 | 0.0000 | 0.0001 | 0.0061 | 0.0000 | 0.0260 | 0.0000 | 0.0058 | 0.5815 |
60
+ | 0.0286 | 13.1579 | 1000 | 0.0219 | 0.0321 | 0.0954 | 0.1715 | 0.0200 | 0.7864 | 0.0508 | 0.0115 | 0.7792 |
61
+ | 0.0255 | 19.7368 | 1500 | 0.0207 | 0.0498 | 0.1326 | 0.3072 | 0.0317 | 0.8760 | 0.0717 | 0.0210 | 0.8132 |
62
+ | 0.0242 | 26.3158 | 2000 | 0.0189 | 0.1174 | 0.2706 | 0.3581 | 0.0803 | 0.8779 | 0.1600 | 0.0560 | 0.8457 |
63
+ | 0.0224 | 32.8947 | 2500 | 0.0182 | 0.1439 | 0.3002 | 0.4783 | 0.0974 | 0.8678 | 0.1815 | 0.0611 | 0.8658 |
64
+ | 0.0218 | 39.4737 | 3000 | 0.0176 | 0.1759 | 0.3506 | 0.5072 | 0.1212 | 0.8672 | 0.2197 | 0.0660 | 0.8763 |
65
+ | 0.0208 | 46.0526 | 3500 | 0.0175 | 0.1726 | 0.3552 | 0.5286 | 0.1159 | 0.8733 | 0.2230 | 0.0668 | 0.8799 |
66
+ | 0.0195 | 52.6316 | 4000 | 0.0175 | 0.2054 | 0.3555 | 0.5903 | 0.1395 | 0.8757 | 0.2230 | 0.0667 | 0.8911 |
67
+ | 0.0190 | 59.2105 | 4500 | 0.0170 | 0.2605 | 0.4670 | 0.5664 | 0.1944 | 0.8264 | 0.3254 | 0.0844 | 0.8934 |
68
+ | 0.0181 | 65.7895 | 5000 | 0.0167 | 0.2875 | 0.4961 | 0.6030 | 0.2183 | 0.8140 | 0.3568 | 0.0828 | 0.8988 |
69
+ | 0.0176 | 72.3684 | 5500 | 0.0163 | 0.2886 | 0.4953 | 0.6718 | 0.2148 | 0.8333 | 0.3524 | 0.0879 | 0.9027 |
70
+ | 0.0175 | 78.9474 | 6000 | 0.0165 | 0.3284 | 0.5237 | 0.6525 | 0.2505 | 0.8133 | 0.3862 | 0.0918 | 0.9040 |
71
+ | 0.0165 | 85.5263 | 6500 | 0.0164 | 0.2734 | 0.4334 | 0.7028 | 0.1933 | 0.8704 | 0.2885 | 0.0836 | 0.9068 |
72
+ | 0.0160 | 92.1053 | 7000 | 0.0159 | 0.3092 | 0.5050 | 0.6930 | 0.2299 | 0.8381 | 0.3613 | 0.0891 | 0.9104 |
73
+ | 0.0153 | 98.6842 | 7500 | 0.0161 | 0.3255 | 0.5075 | 0.6818 | 0.2428 | 0.8315 | 0.3652 | 0.0853 | 0.9109 |
74
+ | 0.0152 | 105.2632 | 8000 | 0.0159 | 0.3157 | 0.5115 | 0.6832 | 0.2352 | 0.8368 | 0.3683 | 0.0903 | 0.9131 |
75
+ | 0.0144 | 111.8421 | 8500 | 0.0159 | 0.3502 | 0.5287 | 0.6968 | 0.2615 | 0.8258 | 0.3888 | 0.0919 | 0.9136 |
76
+ | 0.0142 | 118.4211 | 9000 | 0.0159 | 0.3545 | 0.5422 | 0.6768 | 0.2674 | 0.8219 | 0.4046 | 0.0987 | 0.9144 |
77
+ | 0.0138 | 125.0 | 9500 | 0.0158 | 0.3707 | 0.5545 | 0.7193 | 0.2826 | 0.8147 | 0.4202 | 0.0960 | 0.9167 |
78
+ | 0.0133 | 131.5789 | 10000 | 0.0158 | 0.3631 | 0.5532 | 0.6678 | 0.2773 | 0.8129 | 0.4193 | 0.0936 | 0.9169 |
79
+ | 0.0131 | 138.1579 | 10500 | 0.0158 | 0.3819 | 0.5506 | 0.6592 | 0.2915 | 0.8177 | 0.4150 | 0.0933 | 0.9172 |
80
+ | 0.0129 | 144.7368 | 11000 | 0.0158 | 0.3766 | 0.5443 | 0.7043 | 0.2859 | 0.8208 | 0.4071 | 0.0966 | 0.9179 |
81
+ | 0.0126 | 151.3158 | 11500 | 0.0157 | 0.3770 | 0.5439 | 0.6897 | 0.2843 | 0.8313 | 0.4042 | 0.0960 | 0.9183 |
82
+ | 0.0124 | 157.8947 | 12000 | 0.0158 | 0.3872 | 0.5572 | 0.6658 | 0.2982 | 0.8159 | 0.4230 | 0.0965 | 0.9181 |
83
+ | 0.0129 | 164.4737 | 12500 | 0.0157 | 0.3924 | 0.5575 | 0.6914 | 0.3008 | 0.8196 | 0.4224 | 0.0966 | 0.9184 |
84
+ | 0.0118 | 171.0526 | 13000 | 0.0157 | 0.3919 | 0.5616 | 0.6757 | 0.2997 | 0.8184 | 0.4275 | 0.0982 | 0.9197 |
85
+ | 0.0121 | 177.6316 | 13500 | 0.0157 | 0.3898 | 0.5625 | 0.7064 | 0.2987 | 0.8177 | 0.4286 | 0.0975 | 0.9197 |
86
+ | 0.0120 | 184.2105 | 14000 | 0.0157 | 0.3953 | 0.5648 | 0.6810 | 0.3049 | 0.8182 | 0.4313 | 0.0976 | 0.9196 |
87
+ | 0.0120 | 190.7895 | 14500 | 0.0157 | 0.3955 | 0.5674 | 0.7004 | 0.3052 | 0.8165 | 0.4348 | 0.0976 | 0.9198 |
88
+ | 0.0117 | 197.3684 | 15000 | 0.0157 | 0.3965 | 0.5679 | 0.6998 | 0.3059 | 0.8160 | 0.4355 | 0.0973 | 0.9198 |
89
+ | 0.0121 | 200.0 | 15200 | 0.0157 | 0.3961 | 0.5678 | 0.7001 | 0.3057 | 0.8162 | 0.4354 | 0.0976 | 0.9198 |
90
+
91
+
92
+ ### Framework versions
93
+
94
+ - Transformers 5.12.1
95
+ - Pytorch 2.5.1+cu121
96
+ - Datasets 5.0.1
97
+ - Tokenizers 0.22.2
best_model/config.json ADDED
@@ -0,0 +1,156 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaV2ForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": null,
7
+ "dtype": "float32",
8
+ "eos_token_id": null,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 1024,
12
+ "id2label": {
13
+ "101": "FOREIGN_SPECIAL_RELATIONS_POSITIVE",
14
+ "102": "FOREIGN_SPECIAL_RELATIONS_NEGATIVE",
15
+ "103": "ANTI_IMPERIALISM",
16
+ "104": "MILITARY_POSITIVE",
17
+ "105": "MILITARY_NEGATIVE",
18
+ "106": "PEACE",
19
+ "107": "INTERNATIONALISM_POSITIVE",
20
+ "108": "EUROPEAN_INTEGRATION_POSITIVE",
21
+ "109": "INTERNATIONALISM_NEGATIVE",
22
+ "110": "EUROPEAN_INTEGRATION_NEGATIVE",
23
+ "201": "FREEDOM_HUMAN_RIGHTS",
24
+ "202": "DEMOCRACY_GENERAL",
25
+ "203": "CONSTITUTIONALISM_POSITIVE",
26
+ "204": "CONSTITUTIONALISM_NEGATIVE",
27
+ "301": "DECENTRALISATION_POSITIVE",
28
+ "302": "CENTRALISATION_POSITIVE",
29
+ "303": "GOVERNMENTAL_EFFICIENCY_POSITIVE",
30
+ "304": "POLITICAL_CORRUPTION_NEGATIVE",
31
+ "305": "POLITICAL_AUTHORITY",
32
+ "401": "FREE_ENTERPRISE_POSITIVE",
33
+ "402": "INCENTIVES_POSITIVE",
34
+ "403": "MARKET_REGULATION",
35
+ "404": "ECONOMIC_PLANNING",
36
+ "405": "CORPORATISM_MIXED_ECONOMY",
37
+ "406": "PROTECTIONISM_POSITIVE",
38
+ "407": "PROTECTIONISM_NEGATIVE",
39
+ "408": "ECONOMIC_GOALS",
40
+ "409": "KEYNESIAN_DEMAND_MANAGEMENT",
41
+ "410": "PRODUCTIVITY",
42
+ "411": "TECHNOLOGY_AND_INFRASTRUCTURE",
43
+ "412": "CONTROLLED_ECONOMY",
44
+ "413": "NATIONALISATION",
45
+ "414": "ECONOMIC_ORTHODOXY",
46
+ "415": "MARXIST_ANALYSIS",
47
+ "416": "ANTI_GROWTH_ECONOMY_POSITIVE",
48
+ "501": "ENVIRONMENTAL_PROTECTION",
49
+ "502": "CULTURE",
50
+ "503": "SOCIAL_JUSTICE",
51
+ "504": "WELFARE_STATE_EXPANSION",
52
+ "505": "WELFARE_STATE_LIMITATION",
53
+ "506": "EDUCATION_EXPANSION",
54
+ "507": "EDUCATION_LIMITATION",
55
+ "601": "NATIONAL_WAY_OF_LIFE_POSITIVE",
56
+ "602": "NATIONAL_WAY_OF_LIFE_NEGATIVE",
57
+ "603": "TRADITIONAL_MORALITY_POSITIVE",
58
+ "604": "TRADITIONAL_MORALITY_NEGATIVE",
59
+ "605": "LAW_AND_ORDER",
60
+ "606": "CIVIC_MINDEDNESS",
61
+ "607": "MULTICULTURALISM_GENERAL_POSITIVE",
62
+ "608": "MULTICULTURALISM_GENERAL_NEGATIVE",
63
+ "701": "LABOUR_GROUPS_POSITIVE",
64
+ "702": "LABOUR_GROUPS_NEGATIVE",
65
+ "703": "AGRICULTURE_AND_FARMERS",
66
+ "704": "MIDDLE_CLASS_AND_PROFESSIONALS",
67
+ "705": "UNDERPRIVILEGED_MINORITY_GROUPS",
68
+ "706": "NON_ECONOMIC_DEMOGRAPHIC_GROUPS"
69
+ },
70
+ "initializer_range": 0.02,
71
+ "intermediate_size": 4096,
72
+ "label2id": {
73
+ "AGRICULTURE_AND_FARMERS": 703,
74
+ "ANTI_GROWTH_ECONOMY_POSITIVE": 416,
75
+ "ANTI_IMPERIALISM": 103,
76
+ "CENTRALISATION_POSITIVE": 302,
77
+ "CIVIC_MINDEDNESS": 606,
78
+ "CONSTITUTIONALISM_NEGATIVE": 204,
79
+ "CONSTITUTIONALISM_POSITIVE": 203,
80
+ "CONTROLLED_ECONOMY": 412,
81
+ "CORPORATISM_MIXED_ECONOMY": 405,
82
+ "CULTURE": 502,
83
+ "DECENTRALISATION_POSITIVE": 301,
84
+ "DEMOCRACY_GENERAL": 202,
85
+ "ECONOMIC_GOALS": 408,
86
+ "ECONOMIC_ORTHODOXY": 414,
87
+ "ECONOMIC_PLANNING": 404,
88
+ "EDUCATION_EXPANSION": 506,
89
+ "EDUCATION_LIMITATION": 507,
90
+ "ENVIRONMENTAL_PROTECTION": 501,
91
+ "EUROPEAN_INTEGRATION_NEGATIVE": 110,
92
+ "EUROPEAN_INTEGRATION_POSITIVE": 108,
93
+ "FOREIGN_SPECIAL_RELATIONS_NEGATIVE": 102,
94
+ "FOREIGN_SPECIAL_RELATIONS_POSITIVE": 101,
95
+ "FREEDOM_HUMAN_RIGHTS": 201,
96
+ "FREE_ENTERPRISE_POSITIVE": 401,
97
+ "GOVERNMENTAL_EFFICIENCY_POSITIVE": 303,
98
+ "INCENTIVES_POSITIVE": 402,
99
+ "INTERNATIONALISM_NEGATIVE": 109,
100
+ "INTERNATIONALISM_POSITIVE": 107,
101
+ "KEYNESIAN_DEMAND_MANAGEMENT": 409,
102
+ "LABOUR_GROUPS_NEGATIVE": 702,
103
+ "LABOUR_GROUPS_POSITIVE": 701,
104
+ "LAW_AND_ORDER": 605,
105
+ "MARKET_REGULATION": 403,
106
+ "MARXIST_ANALYSIS": 415,
107
+ "MIDDLE_CLASS_AND_PROFESSIONALS": 704,
108
+ "MILITARY_NEGATIVE": 105,
109
+ "MILITARY_POSITIVE": 104,
110
+ "MULTICULTURALISM_GENERAL_NEGATIVE": 608,
111
+ "MULTICULTURALISM_GENERAL_POSITIVE": 607,
112
+ "NATIONALISATION": 413,
113
+ "NATIONAL_WAY_OF_LIFE_NEGATIVE": 602,
114
+ "NATIONAL_WAY_OF_LIFE_POSITIVE": 601,
115
+ "NON_ECONOMIC_DEMOGRAPHIC_GROUPS": 706,
116
+ "PEACE": 106,
117
+ "POLITICAL_AUTHORITY": 305,
118
+ "POLITICAL_CORRUPTION_NEGATIVE": 304,
119
+ "PRODUCTIVITY": 410,
120
+ "PROTECTIONISM_NEGATIVE": 407,
121
+ "PROTECTIONISM_POSITIVE": 406,
122
+ "SOCIAL_JUSTICE": 503,
123
+ "TECHNOLOGY_AND_INFRASTRUCTURE": 411,
124
+ "TRADITIONAL_MORALITY_NEGATIVE": 604,
125
+ "TRADITIONAL_MORALITY_POSITIVE": 603,
126
+ "UNDERPRIVILEGED_MINORITY_GROUPS": 705,
127
+ "WELFARE_STATE_EXPANSION": 504,
128
+ "WELFARE_STATE_LIMITATION": 505
129
+ },
130
+ "layer_norm_eps": 1e-07,
131
+ "legacy": true,
132
+ "max_position_embeddings": 512,
133
+ "max_relative_positions": -1,
134
+ "model_type": "deberta-v2",
135
+ "norm_rel_ebd": "layer_norm",
136
+ "num_attention_heads": 16,
137
+ "num_hidden_layers": 24,
138
+ "pad_token_id": 0,
139
+ "pooler_dropout": 0.0,
140
+ "pooler_hidden_act": "gelu",
141
+ "pooler_hidden_size": 1024,
142
+ "pos_att_type": [
143
+ "p2c",
144
+ "c2p"
145
+ ],
146
+ "position_biased_input": false,
147
+ "position_buckets": 256,
148
+ "problem_type": "multi_label_classification",
149
+ "relative_attention": true,
150
+ "share_att_key": true,
151
+ "tie_word_embeddings": true,
152
+ "transformers_version": "5.12.1",
153
+ "type_vocab_size": 0,
154
+ "use_cache": false,
155
+ "vocab_size": 128100
156
+ }
best_model/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a23a8790cbd1fe0affd71331f5577f81cb6a9276aa77d1f52d86c4193bb9615
3
+ size 1740525856
best_model/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
best_model/tokenizer_config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "backend": "tokenizers",
4
+ "bos_token": "[CLS]",
5
+ "cls_token": "[CLS]",
6
+ "do_lower_case": false,
7
+ "eos_token": "[SEP]",
8
+ "extra_special_tokens": [
9
+ "[PAD]",
10
+ "[CLS]",
11
+ "[SEP]"
12
+ ],
13
+ "is_local": false,
14
+ "local_files_only": false,
15
+ "mask_token": "[MASK]",
16
+ "model_max_length": 1000000000000000019884624838656,
17
+ "pad_token": "[PAD]",
18
+ "sep_token": "[SEP]",
19
+ "split_by_punct": false,
20
+ "tokenizer_class": "DebertaV2Tokenizer",
21
+ "unk_id": 3,
22
+ "unk_token": "[UNK]",
23
+ "vocab_type": "spm"
24
+ }
best_model/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ee5d9253044ddabcb049f6a6fc7e0e46b6c22266ed774746db7562f1c6aa68b
3
+ size 4856
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:920465422ab89507a06ca9e22e436b97114bae5b8535426173f7e67fbdc0485c
3
  size 1740525856
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a23a8790cbd1fe0affd71331f5577f81cb6a9276aa77d1f52d86c4193bb9615
3
  size 1740525856
runs/Aug06_12-10-47_erc-hpc-vm042/events.out.tfevents.1786031039.erc-hpc-vm042.307913.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65be6cb94ebc610a985c7541e3598c5ac106f1e81ee29ad08520c0885687b910
3
+ size 811
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "backend": "tokenizers",
4
+ "bos_token": "[CLS]",
5
+ "cls_token": "[CLS]",
6
+ "do_lower_case": false,
7
+ "eos_token": "[SEP]",
8
+ "extra_special_tokens": [
9
+ "[PAD]",
10
+ "[CLS]",
11
+ "[SEP]"
12
+ ],
13
+ "is_local": false,
14
+ "local_files_only": false,
15
+ "mask_token": "[MASK]",
16
+ "model_max_length": 1000000000000000019884624838656,
17
+ "pad_token": "[PAD]",
18
+ "sep_token": "[SEP]",
19
+ "split_by_punct": false,
20
+ "tokenizer_class": "DebertaV2Tokenizer",
21
+ "unk_id": 3,
22
+ "unk_token": "[UNK]",
23
+ "vocab_type": "spm"
24
+ }