Hugging Face's logo

Jefferson0211
/
bertimbau-ghg-ner

Token Classification
Transformers
Safetensors
Portuguese
bert
bertimbau
named-entity-recognition
ner
portuguese
ghg-protocol
Model card Files
xet
 Community
Jefferson0211 commited on
Commit
af51f8e
·
verified ·
1 Parent(s): 93ce765

commit

Browse files
Files changed (30) hide show
  1. checkpoint-36/config.json +209 -0
  2. checkpoint-36/model.safetensors +3 -0
  3. checkpoint-36/optimizer.pt +3 -0
  4. checkpoint-36/rng_state.pth +3 -0
  5. checkpoint-36/scheduler.pt +3 -0
  6. checkpoint-36/special_tokens_map.json +7 -0
  7. checkpoint-36/tokenizer.json +0 -0
  8. checkpoint-36/tokenizer_config.json +58 -0
  9. checkpoint-36/trainer_state.json +82 -0
  10. checkpoint-36/training_args.bin +3 -0
  11. checkpoint-36/vocab.txt +0 -0
  12. checkpoint-45/config.json +209 -0
  13. checkpoint-45/model.safetensors +3 -0
  14. checkpoint-45/optimizer.pt +3 -0
  15. checkpoint-45/rng_state.pth +3 -0
  16. checkpoint-45/scheduler.pt +3 -0
  17. checkpoint-45/special_tokens_map.json +7 -0
  18. checkpoint-45/tokenizer.json +0 -0
  19. checkpoint-45/tokenizer_config.json +58 -0
  20. checkpoint-45/trainer_state.json +94 -0
  21. checkpoint-45/training_args.bin +3 -0
  22. checkpoint-45/vocab.txt +0 -0
  23. config.json +209 -0
  24. label_mapping.json +180 -0
  25. model.safetensors +3 -0
  26. special_tokens_map.json +7 -0
  27. tokenizer.json +0 -0
  28. tokenizer_config.json +58 -0
  29. training_args.bin +3 -0
  30. vocab.txt +0 -0
checkpoint-36/config.json ADDED
@@ -0,0 +1,209 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForTokenClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "directionality": "bidi",
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "O",
13
+ "1": "B-Aterro",
14
+ "2": "I-Aterro",
15
+ "3": "B-Biodiesel (B100)",
16
+ "4": "I-Biodiesel (B100)",
17
+ "5": "B-Biometano",
18
+ "6": "I-Biometano",
19
+ "7": "B-Biometanol",
20
+ "8": "I-Biometanol",
21
+ "9": "B-Bioquerosene (SAF)",
22
+ "10": "I-Bioquerosene (SAF)",
23
+ "11": "B-Carro",
24
+ "12": "I-Carro",
25
+ "13": "B-Carv\u00e3o",
26
+ "14": "I-Carv\u00e3o",
27
+ "15": "B-Chegada",
28
+ "16": "I-Chegada",
29
+ "17": "B-Compostagem",
30
+ "18": "I-Compostagem",
31
+ "19": "B-Desmatamento",
32
+ "20": "I-Desmatamento",
33
+ "21": "B-Dist\u00e2ncia (km)",
34
+ "22": "I-Dist\u00e2ncia (km)",
35
+ "23": "B-Di\u00f3xido de Carbono (CO2)",
36
+ "24": "I-Di\u00f3xido de Carbono (CO2)",
37
+ "25": "B-Eletricidade",
38
+ "26": "I-Eletricidade",
39
+ "27": "B-Etanol Anidro",
40
+ "28": "I-Etanol Anidro",
41
+ "29": "B-Etanol Hidratado",
42
+ "30": "I-Etanol Hidratado",
43
+ "31": "B-Gasolina Automotiva",
44
+ "32": "I-Gasolina Automotiva",
45
+ "33": "B-Gasolina de Avia\u00e7\u00e3o",
46
+ "34": "I-Gasolina de Avia\u00e7\u00e3o",
47
+ "35": "B-G\u00e1s Liquefeito de Petr\u00f3leo (GLP)",
48
+ "36": "I-G\u00e1s Liquefeito de Petr\u00f3leo (GLP)",
49
+ "37": "B-G\u00e1s Natural",
50
+ "38": "I-G\u00e1s Natural",
51
+ "39": "B-G\u00e1s Natural Liquefeito (GNL)",
52
+ "40": "I-G\u00e1s Natural Liquefeito (GNL)",
53
+ "41": "B-G\u00e1s Natural Veicular (GNV)",
54
+ "42": "I-G\u00e1s Natural Veicular (GNV)",
55
+ "43": "B-HVO (diesel verde)",
56
+ "44": "I-HVO (diesel verde)",
57
+ "45": "B-Incinera\u00e7\u00e3o",
58
+ "46": "I-Incinera\u00e7\u00e3o",
59
+ "47": "B-Lan\u00e7amento em Corpos d'\u00c1gua",
60
+ "48": "I-Lan\u00e7amento em Corpos d'\u00c1gua",
61
+ "49": "B-Lubrificantes",
62
+ "50": "I-Lubrificantes",
63
+ "51": "B-Metano (CH4)",
64
+ "52": "I-Metano (CH4)",
65
+ "53": "B-Metanol",
66
+ "54": "I-Metanol",
67
+ "55": "B-Partida",
68
+ "56": "I-Partida",
69
+ "57": "B-Quantidade (Litros)",
70
+ "58": "I-Quantidade (Litros)",
71
+ "59": "B-Quantidade (kWh)",
72
+ "60": "I-Quantidade (kWh)",
73
+ "61": "B-Quantidade (kg)",
74
+ "62": "I-Quantidade (kg)",
75
+ "63": "B-Quantidade (m\u00b3)",
76
+ "64": "I-Quantidade (m\u00b3)",
77
+ "65": "B-Querosene de Avia\u00e7\u00e3o",
78
+ "66": "I-Querosene de Avia\u00e7\u00e3o",
79
+ "67": "B-Reciclagem",
80
+ "68": "I-Reciclagem",
81
+ "69": "B-Reflorestamento",
82
+ "70": "I-Reflorestamento",
83
+ "71": "B-Refrigerantes (HFCs)",
84
+ "72": "I-Refrigerantes (HFCs)",
85
+ "73": "B-Tratamento de Efluentes",
86
+ "74": "I-Tratamento de Efluentes",
87
+ "75": "B-Trem",
88
+ "76": "I-Trem",
89
+ "77": "B-kWh",
90
+ "78": "I-kWh",
91
+ "79": "B-\u00c1rea (hectares)",
92
+ "80": "I-\u00c1rea (hectares)",
93
+ "81": "B-\u00d3leo Combust\u00edvel",
94
+ "82": "I-\u00d3leo Combust\u00edvel",
95
+ "83": "B-\u00d3leo Diesel",
96
+ "84": "I-\u00d3leo Diesel",
97
+ "85": "B-\u00d4nibus",
98
+ "86": "I-\u00d4nibus"
99
+ },
100
+ "initializer_range": 0.02,
101
+ "intermediate_size": 3072,
102
+ "label2id": {
103
+ "B-Aterro": 1,
104
+ "B-Biodiesel (B100)": 3,
105
+ "B-Biometano": 5,
106
+ "B-Biometanol": 7,
107
+ "B-Bioquerosene (SAF)": 9,
108
+ "B-Carro": 11,
109
+ "B-Carv\u00e3o": 13,
110
+ "B-Chegada": 15,
111
+ "B-Compostagem": 17,
112
+ "B-Desmatamento": 19,
113
+ "B-Dist\u00e2ncia (km)": 21,
114
+ "B-Di\u00f3xido de Carbono (CO2)": 23,
115
+ "B-Eletricidade": 25,
116
+ "B-Etanol Anidro": 27,
117
+ "B-Etanol Hidratado": 29,
118
+ "B-Gasolina Automotiva": 31,
119
+ "B-Gasolina de Avia\u00e7\u00e3o": 33,
120
+ "B-G\u00e1s Liquefeito de Petr\u00f3leo (GLP)": 35,
121
+ "B-G\u00e1s Natural": 37,
122
+ "B-G\u00e1s Natural Liquefeito (GNL)": 39,
123
+ "B-G\u00e1s Natural Veicular (GNV)": 41,
124
+ "B-HVO (diesel verde)": 43,
125
+ "B-Incinera\u00e7\u00e3o": 45,
126
+ "B-Lan\u00e7amento em Corpos d'\u00c1gua": 47,
127
+ "B-Lubrificantes": 49,
128
+ "B-Metano (CH4)": 51,
129
+ "B-Metanol": 53,
130
+ "B-Partida": 55,
131
+ "B-Quantidade (Litros)": 57,
132
+ "B-Quantidade (kWh)": 59,
133
+ "B-Quantidade (kg)": 61,
134
+ "B-Quantidade (m\u00b3)": 63,
135
+ "B-Querosene de Avia\u00e7\u00e3o": 65,
136
+ "B-Reciclagem": 67,
137
+ "B-Reflorestamento": 69,
138
+ "B-Refrigerantes (HFCs)": 71,
139
+ "B-Tratamento de Efluentes": 73,
140
+ "B-Trem": 75,
141
+ "B-kWh": 77,
142
+ "B-\u00c1rea (hectares)": 79,
143
+ "B-\u00d3leo Combust\u00edvel": 81,
144
+ "B-\u00d3leo Diesel": 83,
145
+ "B-\u00d4nibus": 85,
146
+ "I-Aterro": 2,
147
+ "I-Biodiesel (B100)": 4,
148
+ "I-Biometano": 6,
149
+ "I-Biometanol": 8,
150
+ "I-Bioquerosene (SAF)": 10,
151
+ "I-Carro": 12,
152
+ "I-Carv\u00e3o": 14,
153
+ "I-Chegada": 16,
154
+ "I-Compostagem": 18,
155
+ "I-Desmatamento": 20,
156
+ "I-Dist\u00e2ncia (km)": 22,
157
+ "I-Di\u00f3xido de Carbono (CO2)": 24,
158
+ "I-Eletricidade": 26,
159
+ "I-Etanol Anidro": 28,
160
+ "I-Etanol Hidratado": 30,
161
+ "I-Gasolina Automotiva": 32,
162
+ "I-Gasolina de Avia\u00e7\u00e3o": 34,
163
+ "I-G\u00e1s Liquefeito de Petr\u00f3leo (GLP)": 36,
164
+ "I-G\u00e1s Natural": 38,
165
+ "I-G\u00e1s Natural Liquefeito (GNL)": 40,
166
+ "I-G\u00e1s Natural Veicular (GNV)": 42,
167
+ "I-HVO (diesel verde)": 44,
168
+ "I-Incinera\u00e7\u00e3o": 46,
169
+ "I-Lan\u00e7amento em Corpos d'\u00c1gua": 48,
170
+ "I-Lubrificantes": 50,
171
+ "I-Metano (CH4)": 52,
172
+ "I-Metanol": 54,
173
+ "I-Partida": 56,
174
+ "I-Quantidade (Litros)": 58,
175
+ "I-Quantidade (kWh)": 60,
176
+ "I-Quantidade (kg)": 62,
177
+ "I-Quantidade (m\u00b3)": 64,
178
+ "I-Querosene de Avia\u00e7\u00e3o": 66,
179
+ "I-Reciclagem": 68,
180
+ "I-Reflorestamento": 70,
181
+ "I-Refrigerantes (HFCs)": 72,
182
+ "I-Tratamento de Efluentes": 74,
183
+ "I-Trem": 76,
184
+ "I-kWh": 78,
185
+ "I-\u00c1rea (hectares)": 80,
186
+ "I-\u00d3leo Combust\u00edvel": 82,
187
+ "I-\u00d3leo Diesel": 84,
188
+ "I-\u00d4nibus": 86,
189
+ "O": 0
190
+ },
191
+ "layer_norm_eps": 1e-12,
192
+ "max_position_embeddings": 512,
193
+ "model_type": "bert",
194
+ "num_attention_heads": 12,
195
+ "num_hidden_layers": 12,
196
+ "output_past": true,
197
+ "pad_token_id": 0,
198
+ "pooler_fc_size": 768,
199
+ "pooler_num_attention_heads": 12,
200
+ "pooler_num_fc_layers": 3,
201
+ "pooler_size_per_head": 128,
202
+ "pooler_type": "first_token_transform",
203
+ "position_embedding_type": "absolute",
204
+ "torch_dtype": "float32",
205
+ "transformers_version": "4.54.1",
206
+ "type_vocab_size": 2,
207
+ "use_cache": true,
208
+ "vocab_size": 29794
209
+ }
checkpoint-36/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aef9635c63e028b7cae65270be9f9f3fcc5291827a828173ae1fac581bda2fe4
3
+ size 433621132
checkpoint-36/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:55324da7ef8d009f936f13bead7ff2ce31451c3b313e3950c3bd227fd86b93ff
3
+ size 867356875
checkpoint-36/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bafc777607443a0b1dd392aafd4f69526dc11c371941eb37a8fe415a396740cf
3
+ size 14455
checkpoint-36/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c511b55e3e72b1a938625cff8cbbef782727bea1a8848a153fd4f6477b22173
3
+ size 1465
checkpoint-36/special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "mask_token": "[MASK]",
4
+ "pad_token": "[PAD]",
5
+ "sep_token": "[SEP]",
6
+ "unk_token": "[UNK]"
7
+ }
checkpoint-36/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-36/tokenizer_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": false,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 1000000000000000019884624838656,
51
+ "never_split": null,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "strip_accents": null,
55
+ "tokenize_chinese_chars": true,
56
+ "tokenizer_class": "BertTokenizer",
57
+ "unk_token": "[UNK]"
58
+ }
checkpoint-36/trainer_state.json ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 36,
3
+ "best_metric": 0.3309560716152191,
4
+ "best_model_checkpoint": "./modelo_bertimbau_ghg_ner\\checkpoint-36",
5
+ "epoch": 4.0,
6
+ "eval_steps": 500,
7
+ "global_step": 36,
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": 0.9594190516873131,
15
+ "eval_f1": 0.0,
16
+ "eval_loss": 1.2153310775756836,
17
+ "eval_precision": 0.0,
18
+ "eval_recall": 0.0,
19
+ "eval_runtime": 14.1172,
20
+ "eval_samples_per_second": 2.408,
21
+ "eval_steps_per_second": 0.213,
22
+ "step": 9
23
+ },
24
+ {
25
+ "epoch": 2.0,
26
+ "eval_accuracy": 0.9594190516873131,
27
+ "eval_f1": 0.0,
28
+ "eval_loss": 0.40973424911499023,
29
+ "eval_precision": 0.0,
30
+ "eval_recall": 0.0,
31
+ "eval_runtime": 10.7293,
32
+ "eval_samples_per_second": 3.169,
33
+ "eval_steps_per_second": 0.28,
34
+ "step": 18
35
+ },
36
+ {
37
+ "epoch": 3.0,
38
+ "eval_accuracy": 0.9594190516873131,
39
+ "eval_f1": 0.0,
40
+ "eval_loss": 0.3423544466495514,
41
+ "eval_precision": 0.0,
42
+ "eval_recall": 0.0,
43
+ "eval_runtime": 17.2156,
44
+ "eval_samples_per_second": 1.975,
45
+ "eval_steps_per_second": 0.174,
46
+ "step": 27
47
+ },
48
+ {
49
+ "epoch": 4.0,
50
+ "eval_accuracy": 0.9594190516873131,
51
+ "eval_f1": 0.0,
52
+ "eval_loss": 0.3309560716152191,
53
+ "eval_precision": 0.0,
54
+ "eval_recall": 0.0,
55
+ "eval_runtime": 10.9382,
56
+ "eval_samples_per_second": 3.108,
57
+ "eval_steps_per_second": 0.274,
58
+ "step": 36
59
+ }
60
+ ],
61
+ "logging_steps": 100,
62
+ "max_steps": 45,
63
+ "num_input_tokens_seen": 0,
64
+ "num_train_epochs": 5,
65
+ "save_steps": 500,
66
+ "stateful_callbacks": {
67
+ "TrainerControl": {
68
+ "args": {
69
+ "should_epoch_stop": false,
70
+ "should_evaluate": false,
71
+ "should_log": false,
72
+ "should_save": true,
73
+ "should_training_stop": false
74
+ },
75
+ "attributes": {}
76
+ }
77
+ },
78
+ "total_flos": 139116700864512.0,
79
+ "train_batch_size": 16,
80
+ "trial_name": null,
81
+ "trial_params": null
82
+ }
checkpoint-36/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f08bbb5a047eb222170b0cd23aa687c92d3b3536131d24cb029f7fdd7d8ef695
3
+ size 5713
checkpoint-36/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-45/config.json ADDED
@@ -0,0 +1,209 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForTokenClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "directionality": "bidi",
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "O",
13
+ "1": "B-Aterro",
14
+ "2": "I-Aterro",
15
+ "3": "B-Biodiesel (B100)",
16
+ "4": "I-Biodiesel (B100)",
17
+ "5": "B-Biometano",
18
+ "6": "I-Biometano",
19
+ "7": "B-Biometanol",
20
+ "8": "I-Biometanol",
21
+ "9": "B-Bioquerosene (SAF)",
22
+ "10": "I-Bioquerosene (SAF)",
23
+ "11": "B-Carro",
24
+ "12": "I-Carro",
25
+ "13": "B-Carv\u00e3o",
26
+ "14": "I-Carv\u00e3o",
27
+ "15": "B-Chegada",
28
+ "16": "I-Chegada",
29
+ "17": "B-Compostagem",
30
+ "18": "I-Compostagem",
31
+ "19": "B-Desmatamento",
32
+ "20": "I-Desmatamento",
33
+ "21": "B-Dist\u00e2ncia (km)",
34
+ "22": "I-Dist\u00e2ncia (km)",
35
+ "23": "B-Di\u00f3xido de Carbono (CO2)",
36
+ "24": "I-Di\u00f3xido de Carbono (CO2)",
37
+ "25": "B-Eletricidade",
38
+ "26": "I-Eletricidade",
39
+ "27": "B-Etanol Anidro",
40
+ "28": "I-Etanol Anidro",
41
+ "29": "B-Etanol Hidratado",
42
+ "30": "I-Etanol Hidratado",
43
+ "31": "B-Gasolina Automotiva",
44
+ "32": "I-Gasolina Automotiva",
45
+ "33": "B-Gasolina de Avia\u00e7\u00e3o",
46
+ "34": "I-Gasolina de Avia\u00e7\u00e3o",
47
+ "35": "B-G\u00e1s Liquefeito de Petr\u00f3leo (GLP)",
48
+ "36": "I-G\u00e1s Liquefeito de Petr\u00f3leo (GLP)",
49
+ "37": "B-G\u00e1s Natural",
50
+ "38": "I-G\u00e1s Natural",
51
+ "39": "B-G\u00e1s Natural Liquefeito (GNL)",
52
+ "40": "I-G\u00e1s Natural Liquefeito (GNL)",
53
+ "41": "B-G\u00e1s Natural Veicular (GNV)",
54
+ "42": "I-G\u00e1s Natural Veicular (GNV)",
55
+ "43": "B-HVO (diesel verde)",
56
+ "44": "I-HVO (diesel verde)",
57
+ "45": "B-Incinera\u00e7\u00e3o",
58
+ "46": "I-Incinera\u00e7\u00e3o",
59
+ "47": "B-Lan\u00e7amento em Corpos d'\u00c1gua",
60
+ "48": "I-Lan\u00e7amento em Corpos d'\u00c1gua",
61
+ "49": "B-Lubrificantes",
62
+ "50": "I-Lubrificantes",
63
+ "51": "B-Metano (CH4)",
64
+ "52": "I-Metano (CH4)",
65
+ "53": "B-Metanol",
66
+ "54": "I-Metanol",
67
+ "55": "B-Partida",
68
+ "56": "I-Partida",
69
+ "57": "B-Quantidade (Litros)",
70
+ "58": "I-Quantidade (Litros)",
71
+ "59": "B-Quantidade (kWh)",
72
+ "60": "I-Quantidade (kWh)",
73
+ "61": "B-Quantidade (kg)",
74
+ "62": "I-Quantidade (kg)",
75
+ "63": "B-Quantidade (m\u00b3)",
76
+ "64": "I-Quantidade (m\u00b3)",
77
+ "65": "B-Querosene de Avia\u00e7\u00e3o",
78
+ "66": "I-Querosene de Avia\u00e7\u00e3o",
79
+ "67": "B-Reciclagem",
80
+ "68": "I-Reciclagem",
81
+ "69": "B-Reflorestamento",
82
+ "70": "I-Reflorestamento",
83
+ "71": "B-Refrigerantes (HFCs)",
84
+ "72": "I-Refrigerantes (HFCs)",
85
+ "73": "B-Tratamento de Efluentes",
86
+ "74": "I-Tratamento de Efluentes",
87
+ "75": "B-Trem",
88
+ "76": "I-Trem",
89
+ "77": "B-kWh",
90
+ "78": "I-kWh",
91
+ "79": "B-\u00c1rea (hectares)",
92
+ "80": "I-\u00c1rea (hectares)",
93
+ "81": "B-\u00d3leo Combust\u00edvel",
94
+ "82": "I-\u00d3leo Combust\u00edvel",
95
+ "83": "B-\u00d3leo Diesel",
96
+ "84": "I-\u00d3leo Diesel",
97
+ "85": "B-\u00d4nibus",
98
+ "86": "I-\u00d4nibus"
99
+ },
100
+ "initializer_range": 0.02,
101
+ "intermediate_size": 3072,
102
+ "label2id": {
103
+ "B-Aterro": 1,
104
+ "B-Biodiesel (B100)": 3,
105
+ "B-Biometano": 5,
106
+ "B-Biometanol": 7,
107
+ "B-Bioquerosene (SAF)": 9,
108
+ "B-Carro": 11,
109
+ "B-Carv\u00e3o": 13,
110
+ "B-Chegada": 15,
111
+ "B-Compostagem": 17,
112
+ "B-Desmatamento": 19,
113
+ "B-Dist\u00e2ncia (km)": 21,
114
+ "B-Di\u00f3xido de Carbono (CO2)": 23,
115
+ "B-Eletricidade": 25,
116
+ "B-Etanol Anidro": 27,
117
+ "B-Etanol Hidratado": 29,
118
+ "B-Gasolina Automotiva": 31,
119
+ "B-Gasolina de Avia\u00e7\u00e3o": 33,
120
+ "B-G\u00e1s Liquefeito de Petr\u00f3leo (GLP)": 35,
121
+ "B-G\u00e1s Natural": 37,
122
+ "B-G\u00e1s Natural Liquefeito (GNL)": 39,
123
+ "B-G\u00e1s Natural Veicular (GNV)": 41,
124
+ "B-HVO (diesel verde)": 43,
125
+ "B-Incinera\u00e7\u00e3o": 45,
126
+ "B-Lan\u00e7amento em Corpos d'\u00c1gua": 47,
127
+ "B-Lubrificantes": 49,
128
+ "B-Metano (CH4)": 51,
129
+ "B-Metanol": 53,
130
+ "B-Partida": 55,
131
+ "B-Quantidade (Litros)": 57,
132
+ "B-Quantidade (kWh)": 59,
133
+ "B-Quantidade (kg)": 61,
134
+ "B-Quantidade (m\u00b3)": 63,
135
+ "B-Querosene de Avia\u00e7\u00e3o": 65,
136
+ "B-Reciclagem": 67,
137
+ "B-Reflorestamento": 69,
138
+ "B-Refrigerantes (HFCs)": 71,
139
+ "B-Tratamento de Efluentes": 73,
140
+ "B-Trem": 75,
141
+ "B-kWh": 77,
142
+ "B-\u00c1rea (hectares)": 79,
143
+ "B-\u00d3leo Combust\u00edvel": 81,
144
+ "B-\u00d3leo Diesel": 83,
145
+ "B-\u00d4nibus": 85,
146
+ "I-Aterro": 2,
147
+ "I-Biodiesel (B100)": 4,
148
+ "I-Biometano": 6,
149
+ "I-Biometanol": 8,
150
+ "I-Bioquerosene (SAF)": 10,
151
+ "I-Carro": 12,
152
+ "I-Carv\u00e3o": 14,
153
+ "I-Chegada": 16,
154
+ "I-Compostagem": 18,
155
+ "I-Desmatamento": 20,
156
+ "I-Dist\u00e2ncia (km)": 22,
157
+ "I-Di\u00f3xido de Carbono (CO2)": 24,
158
+ "I-Eletricidade": 26,
159
+ "I-Etanol Anidro": 28,
160
+ "I-Etanol Hidratado": 30,
161
+ "I-Gasolina Automotiva": 32,
162
+ "I-Gasolina de Avia\u00e7\u00e3o": 34,
163
+ "I-G\u00e1s Liquefeito de Petr\u00f3leo (GLP)": 36,
164
+ "I-G\u00e1s Natural": 38,
165
+ "I-G\u00e1s Natural Liquefeito (GNL)": 40,
166
+ "I-G\u00e1s Natural Veicular (GNV)": 42,
167
+ "I-HVO (diesel verde)": 44,
168
+ "I-Incinera\u00e7\u00e3o": 46,
169
+ "I-Lan\u00e7amento em Corpos d'\u00c1gua": 48,
170
+ "I-Lubrificantes": 50,
171
+ "I-Metano (CH4)": 52,
172
+ "I-Metanol": 54,
173
+ "I-Partida": 56,
174
+ "I-Quantidade (Litros)": 58,
175
+ "I-Quantidade (kWh)": 60,
176
+ "I-Quantidade (kg)": 62,
177
+ "I-Quantidade (m\u00b3)": 64,
178
+ "I-Querosene de Avia\u00e7\u00e3o": 66,
179
+ "I-Reciclagem": 68,
180
+ "I-Reflorestamento": 70,
181
+ "I-Refrigerantes (HFCs)": 72,
182
+ "I-Tratamento de Efluentes": 74,
183
+ "I-Trem": 76,
184
+ "I-kWh": 78,
185
+ "I-\u00c1rea (hectares)": 80,
186
+ "I-\u00d3leo Combust\u00edvel": 82,
187
+ "I-\u00d3leo Diesel": 84,
188
+ "I-\u00d4nibus": 86,
189
+ "O": 0
190
+ },
191
+ "layer_norm_eps": 1e-12,
192
+ "max_position_embeddings": 512,
193
+ "model_type": "bert",
194
+ "num_attention_heads": 12,
195
+ "num_hidden_layers": 12,
196
+ "output_past": true,
197
+ "pad_token_id": 0,
198
+ "pooler_fc_size": 768,
199
+ "pooler_num_attention_heads": 12,
200
+ "pooler_num_fc_layers": 3,
201
+ "pooler_size_per_head": 128,
202
+ "pooler_type": "first_token_transform",
203
+ "position_embedding_type": "absolute",
204
+ "torch_dtype": "float32",
205
+ "transformers_version": "4.54.1",
206
+ "type_vocab_size": 2,
207
+ "use_cache": true,
208
+ "vocab_size": 29794
209
+ }
checkpoint-45/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6d88acb3cc285af0fb51d31951758bea2a04c329284b7834ecce81b84e6f213b
3
+ size 433621132
checkpoint-45/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2f2f507be1a6620c5f27b06551713e372651d0ccea6258dc14140bfa1e6aaaa1
3
+ size 867356875
checkpoint-45/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b9184338ed632bb3b57f8e71bede309e929346896719706f35a57902002456b7
3
+ size 14455
checkpoint-45/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:842ce412250e9e464ea81a62fe102a133f52d70ccc41add812aeb320b0156a9c
3
+ size 1465
checkpoint-45/special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "mask_token": "[MASK]",
4
+ "pad_token": "[PAD]",
5
+ "sep_token": "[SEP]",
6
+ "unk_token": "[UNK]"
7
+ }
checkpoint-45/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-45/tokenizer_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": false,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 1000000000000000019884624838656,
51
+ "never_split": null,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "strip_accents": null,
55
+ "tokenize_chinese_chars": true,
56
+ "tokenizer_class": "BertTokenizer",
57
+ "unk_token": "[UNK]"
58
+ }
checkpoint-45/trainer_state.json ADDED
@@ -0,0 +1,94 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 45,
3
+ "best_metric": 0.3264974355697632,
4
+ "best_model_checkpoint": "./modelo_bertimbau_ghg_ner\\checkpoint-45",
5
+ "epoch": 5.0,
6
+ "eval_steps": 500,
7
+ "global_step": 45,
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": 0.9594190516873131,
15
+ "eval_f1": 0.0,
16
+ "eval_loss": 1.2153310775756836,
17
+ "eval_precision": 0.0,
18
+ "eval_recall": 0.0,
19
+ "eval_runtime": 14.1172,
20
+ "eval_samples_per_second": 2.408,
21
+ "eval_steps_per_second": 0.213,
22
+ "step": 9
23
+ },
24
+ {
25
+ "epoch": 2.0,
26
+ "eval_accuracy": 0.9594190516873131,
27
+ "eval_f1": 0.0,
28
+ "eval_loss": 0.40973424911499023,
29
+ "eval_precision": 0.0,
30
+ "eval_recall": 0.0,
31
+ "eval_runtime": 10.7293,
32
+ "eval_samples_per_second": 3.169,
33
+ "eval_steps_per_second": 0.28,
34
+ "step": 18
35
+ },
36
+ {
37
+ "epoch": 3.0,
38
+ "eval_accuracy": 0.9594190516873131,
39
+ "eval_f1": 0.0,
40
+ "eval_loss": 0.3423544466495514,
41
+ "eval_precision": 0.0,
42
+ "eval_recall": 0.0,
43
+ "eval_runtime": 17.2156,
44
+ "eval_samples_per_second": 1.975,
45
+ "eval_steps_per_second": 0.174,
46
+ "step": 27
47
+ },
48
+ {
49
+ "epoch": 4.0,
50
+ "eval_accuracy": 0.9594190516873131,
51
+ "eval_f1": 0.0,
52
+ "eval_loss": 0.3309560716152191,
53
+ "eval_precision": 0.0,
54
+ "eval_recall": 0.0,
55
+ "eval_runtime": 10.9382,
56
+ "eval_samples_per_second": 3.108,
57
+ "eval_steps_per_second": 0.274,
58
+ "step": 36
59
+ },
60
+ {
61
+ "epoch": 5.0,
62
+ "eval_accuracy": 0.9594190516873131,
63
+ "eval_f1": 0.0,
64
+ "eval_loss": 0.3264974355697632,
65
+ "eval_precision": 0.0,
66
+ "eval_recall": 0.0,
67
+ "eval_runtime": 14.5567,
68
+ "eval_samples_per_second": 2.336,
69
+ "eval_steps_per_second": 0.206,
70
+ "step": 45
71
+ }
72
+ ],
73
+ "logging_steps": 100,
74
+ "max_steps": 45,
75
+ "num_input_tokens_seen": 0,
76
+ "num_train_epochs": 5,
77
+ "save_steps": 500,
78
+ "stateful_callbacks": {
79
+ "TrainerControl": {
80
+ "args": {
81
+ "should_epoch_stop": false,
82
+ "should_evaluate": false,
83
+ "should_log": false,
84
+ "should_save": true,
85
+ "should_training_stop": true
86
+ },
87
+ "attributes": {}
88
+ }
89
+ },
90
+ "total_flos": 173895876080640.0,
91
+ "train_batch_size": 16,
92
+ "trial_name": null,
93
+ "trial_params": null
94
+ }
checkpoint-45/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f08bbb5a047eb222170b0cd23aa687c92d3b3536131d24cb029f7fdd7d8ef695
3
+ size 5713
checkpoint-45/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
config.json ADDED
@@ -0,0 +1,209 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForTokenClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "directionality": "bidi",
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "O",
13
+ "1": "B-Aterro",
14
+ "2": "I-Aterro",
15
+ "3": "B-Biodiesel (B100)",
16
+ "4": "I-Biodiesel (B100)",
17
+ "5": "B-Biometano",
18
+ "6": "I-Biometano",
19
+ "7": "B-Biometanol",
20
+ "8": "I-Biometanol",
21
+ "9": "B-Bioquerosene (SAF)",
22
+ "10": "I-Bioquerosene (SAF)",
23
+ "11": "B-Carro",
24
+ "12": "I-Carro",
25
+ "13": "B-Carv\u00e3o",
26
+ "14": "I-Carv\u00e3o",
27
+ "15": "B-Chegada",
28
+ "16": "I-Chegada",
29
+ "17": "B-Compostagem",
30
+ "18": "I-Compostagem",
31
+ "19": "B-Desmatamento",
32
+ "20": "I-Desmatamento",
33
+ "21": "B-Dist\u00e2ncia (km)",
34
+ "22": "I-Dist\u00e2ncia (km)",
35
+ "23": "B-Di\u00f3xido de Carbono (CO2)",
36
+ "24": "I-Di\u00f3xido de Carbono (CO2)",
37
+ "25": "B-Eletricidade",
38
+ "26": "I-Eletricidade",
39
+ "27": "B-Etanol Anidro",
40
+ "28": "I-Etanol Anidro",
41
+ "29": "B-Etanol Hidratado",
42
+ "30": "I-Etanol Hidratado",
43
+ "31": "B-Gasolina Automotiva",
44
+ "32": "I-Gasolina Automotiva",
45
+ "33": "B-Gasolina de Avia\u00e7\u00e3o",
46
+ "34": "I-Gasolina de Avia\u00e7\u00e3o",
47
+ "35": "B-G\u00e1s Liquefeito de Petr\u00f3leo (GLP)",
48
+ "36": "I-G\u00e1s Liquefeito de Petr\u00f3leo (GLP)",
49
+ "37": "B-G\u00e1s Natural",
50
+ "38": "I-G\u00e1s Natural",
51
+ "39": "B-G\u00e1s Natural Liquefeito (GNL)",
52
+ "40": "I-G\u00e1s Natural Liquefeito (GNL)",
53
+ "41": "B-G\u00e1s Natural Veicular (GNV)",
54
+ "42": "I-G\u00e1s Natural Veicular (GNV)",
55
+ "43": "B-HVO (diesel verde)",
56
+ "44": "I-HVO (diesel verde)",
57
+ "45": "B-Incinera\u00e7\u00e3o",
58
+ "46": "I-Incinera\u00e7\u00e3o",
59
+ "47": "B-Lan\u00e7amento em Corpos d'\u00c1gua",
60
+ "48": "I-Lan\u00e7amento em Corpos d'\u00c1gua",
61
+ "49": "B-Lubrificantes",
62
+ "50": "I-Lubrificantes",
63
+ "51": "B-Metano (CH4)",
64
+ "52": "I-Metano (CH4)",
65
+ "53": "B-Metanol",
66
+ "54": "I-Metanol",
67
+ "55": "B-Partida",
68
+ "56": "I-Partida",
69
+ "57": "B-Quantidade (Litros)",
70
+ "58": "I-Quantidade (Litros)",
71
+ "59": "B-Quantidade (kWh)",
72
+ "60": "I-Quantidade (kWh)",
73
+ "61": "B-Quantidade (kg)",
74
+ "62": "I-Quantidade (kg)",
75
+ "63": "B-Quantidade (m\u00b3)",
76
+ "64": "I-Quantidade (m\u00b3)",
77
+ "65": "B-Querosene de Avia\u00e7\u00e3o",
78
+ "66": "I-Querosene de Avia\u00e7\u00e3o",
79
+ "67": "B-Reciclagem",
80
+ "68": "I-Reciclagem",
81
+ "69": "B-Reflorestamento",
82
+ "70": "I-Reflorestamento",
83
+ "71": "B-Refrigerantes (HFCs)",
84
+ "72": "I-Refrigerantes (HFCs)",
85
+ "73": "B-Tratamento de Efluentes",
86
+ "74": "I-Tratamento de Efluentes",
87
+ "75": "B-Trem",
88
+ "76": "I-Trem",
89
+ "77": "B-kWh",
90
+ "78": "I-kWh",
91
+ "79": "B-\u00c1rea (hectares)",
92
+ "80": "I-\u00c1rea (hectares)",
93
+ "81": "B-\u00d3leo Combust\u00edvel",
94
+ "82": "I-\u00d3leo Combust\u00edvel",
95
+ "83": "B-\u00d3leo Diesel",
96
+ "84": "I-\u00d3leo Diesel",
97
+ "85": "B-\u00d4nibus",
98
+ "86": "I-\u00d4nibus"
99
+ },
100
+ "initializer_range": 0.02,
101
+ "intermediate_size": 3072,
102
+ "label2id": {
103
+ "B-Aterro": 1,
104
+ "B-Biodiesel (B100)": 3,
105
+ "B-Biometano": 5,
106
+ "B-Biometanol": 7,
107
+ "B-Bioquerosene (SAF)": 9,
108
+ "B-Carro": 11,
109
+ "B-Carv\u00e3o": 13,
110
+ "B-Chegada": 15,
111
+ "B-Compostagem": 17,
112
+ "B-Desmatamento": 19,
113
+ "B-Dist\u00e2ncia (km)": 21,
114
+ "B-Di\u00f3xido de Carbono (CO2)": 23,
115
+ "B-Eletricidade": 25,
116
+ "B-Etanol Anidro": 27,
117
+ "B-Etanol Hidratado": 29,
118
+ "B-Gasolina Automotiva": 31,
119
+ "B-Gasolina de Avia\u00e7\u00e3o": 33,
120
+ "B-G\u00e1s Liquefeito de Petr\u00f3leo (GLP)": 35,
121
+ "B-G\u00e1s Natural": 37,
122
+ "B-G\u00e1s Natural Liquefeito (GNL)": 39,
123
+ "B-G\u00e1s Natural Veicular (GNV)": 41,
124
+ "B-HVO (diesel verde)": 43,
125
+ "B-Incinera\u00e7\u00e3o": 45,
126
+ "B-Lan\u00e7amento em Corpos d'\u00c1gua": 47,
127
+ "B-Lubrificantes": 49,
128
+ "B-Metano (CH4)": 51,
129
+ "B-Metanol": 53,
130
+ "B-Partida": 55,
131
+ "B-Quantidade (Litros)": 57,
132
+ "B-Quantidade (kWh)": 59,
133
+ "B-Quantidade (kg)": 61,
134
+ "B-Quantidade (m\u00b3)": 63,
135
+ "B-Querosene de Avia\u00e7\u00e3o": 65,
136
+ "B-Reciclagem": 67,
137
+ "B-Reflorestamento": 69,
138
+ "B-Refrigerantes (HFCs)": 71,
139
+ "B-Tratamento de Efluentes": 73,
140
+ "B-Trem": 75,
141
+ "B-kWh": 77,
142
+ "B-\u00c1rea (hectares)": 79,
143
+ "B-\u00d3leo Combust\u00edvel": 81,
144
+ "B-\u00d3leo Diesel": 83,
145
+ "B-\u00d4nibus": 85,
146
+ "I-Aterro": 2,
147
+ "I-Biodiesel (B100)": 4,
148
+ "I-Biometano": 6,
149
+ "I-Biometanol": 8,
150
+ "I-Bioquerosene (SAF)": 10,
151
+ "I-Carro": 12,
152
+ "I-Carv\u00e3o": 14,
153
+ "I-Chegada": 16,
154
+ "I-Compostagem": 18,
155
+ "I-Desmatamento": 20,
156
+ "I-Dist\u00e2ncia (km)": 22,
157
+ "I-Di\u00f3xido de Carbono (CO2)": 24,
158
+ "I-Eletricidade": 26,
159
+ "I-Etanol Anidro": 28,
160
+ "I-Etanol Hidratado": 30,
161
+ "I-Gasolina Automotiva": 32,
162
+ "I-Gasolina de Avia\u00e7\u00e3o": 34,
163
+ "I-G\u00e1s Liquefeito de Petr\u00f3leo (GLP)": 36,
164
+ "I-G\u00e1s Natural": 38,
165
+ "I-G\u00e1s Natural Liquefeito (GNL)": 40,
166
+ "I-G\u00e1s Natural Veicular (GNV)": 42,
167
+ "I-HVO (diesel verde)": 44,
168
+ "I-Incinera\u00e7\u00e3o": 46,
169
+ "I-Lan\u00e7amento em Corpos d'\u00c1gua": 48,
170
+ "I-Lubrificantes": 50,
171
+ "I-Metano (CH4)": 52,
172
+ "I-Metanol": 54,
173
+ "I-Partida": 56,
174
+ "I-Quantidade (Litros)": 58,
175
+ "I-Quantidade (kWh)": 60,
176
+ "I-Quantidade (kg)": 62,
177
+ "I-Quantidade (m\u00b3)": 64,
178
+ "I-Querosene de Avia\u00e7\u00e3o": 66,
179
+ "I-Reciclagem": 68,
180
+ "I-Reflorestamento": 70,
181
+ "I-Refrigerantes (HFCs)": 72,
182
+ "I-Tratamento de Efluentes": 74,
183
+ "I-Trem": 76,
184
+ "I-kWh": 78,
185
+ "I-\u00c1rea (hectares)": 80,
186
+ "I-\u00d3leo Combust\u00edvel": 82,
187
+ "I-\u00d3leo Diesel": 84,
188
+ "I-\u00d4nibus": 86,
189
+ "O": 0
190
+ },
191
+ "layer_norm_eps": 1e-12,
192
+ "max_position_embeddings": 512,
193
+ "model_type": "bert",
194
+ "num_attention_heads": 12,
195
+ "num_hidden_layers": 12,
196
+ "output_past": true,
197
+ "pad_token_id": 0,
198
+ "pooler_fc_size": 768,
199
+ "pooler_num_attention_heads": 12,
200
+ "pooler_num_fc_layers": 3,
201
+ "pooler_size_per_head": 128,
202
+ "pooler_type": "first_token_transform",
203
+ "position_embedding_type": "absolute",
204
+ "torch_dtype": "float32",
205
+ "transformers_version": "4.54.1",
206
+ "type_vocab_size": 2,
207
+ "use_cache": true,
208
+ "vocab_size": 29794
209
+ }
label_mapping.json ADDED
@@ -0,0 +1,180 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "label_to_id": {
3
+ "O": 0,
4
+ "B-Aterro": 1,
5
+ "I-Aterro": 2,
6
+ "B-Biodiesel (B100)": 3,
7
+ "I-Biodiesel (B100)": 4,
8
+ "B-Biometano": 5,
9
+ "I-Biometano": 6,
10
+ "B-Biometanol": 7,
11
+ "I-Biometanol": 8,
12
+ "B-Bioquerosene (SAF)": 9,
13
+ "I-Bioquerosene (SAF)": 10,
14
+ "B-Carro": 11,
15
+ "I-Carro": 12,
16
+ "B-Carvão": 13,
17
+ "I-Carvão": 14,
18
+ "B-Chegada": 15,
19
+ "I-Chegada": 16,
20
+ "B-Compostagem": 17,
21
+ "I-Compostagem": 18,
22
+ "B-Desmatamento": 19,
23
+ "I-Desmatamento": 20,
24
+ "B-Distância (km)": 21,
25
+ "I-Distância (km)": 22,
26
+ "B-Dióxido de Carbono (CO2)": 23,
27
+ "I-Dióxido de Carbono (CO2)": 24,
28
+ "B-Eletricidade": 25,
29
+ "I-Eletricidade": 26,
30
+ "B-Etanol Anidro": 27,
31
+ "I-Etanol Anidro": 28,
32
+ "B-Etanol Hidratado": 29,
33
+ "I-Etanol Hidratado": 30,
34
+ "B-Gasolina Automotiva": 31,
35
+ "I-Gasolina Automotiva": 32,
36
+ "B-Gasolina de Aviação": 33,
37
+ "I-Gasolina de Aviação": 34,
38
+ "B-Gás Liquefeito de Petróleo (GLP)": 35,
39
+ "I-Gás Liquefeito de Petróleo (GLP)": 36,
40
+ "B-Gás Natural": 37,
41
+ "I-Gás Natural": 38,
42
+ "B-Gás Natural Liquefeito (GNL)": 39,
43
+ "I-Gás Natural Liquefeito (GNL)": 40,
44
+ "B-Gás Natural Veicular (GNV)": 41,
45
+ "I-Gás Natural Veicular (GNV)": 42,
46
+ "B-HVO (diesel verde)": 43,
47
+ "I-HVO (diesel verde)": 44,
48
+ "B-Incineração": 45,
49
+ "I-Incineração": 46,
50
+ "B-Lançamento em Corpos d'Água": 47,
51
+ "I-Lançamento em Corpos d'Água": 48,
52
+ "B-Lubrificantes": 49,
53
+ "I-Lubrificantes": 50,
54
+ "B-Metano (CH4)": 51,
55
+ "I-Metano (CH4)": 52,
56
+ "B-Metanol": 53,
57
+ "I-Metanol": 54,
58
+ "B-Partida": 55,
59
+ "I-Partida": 56,
60
+ "B-Quantidade (Litros)": 57,
61
+ "I-Quantidade (Litros)": 58,
62
+ "B-Quantidade (kWh)": 59,
63
+ "I-Quantidade (kWh)": 60,
64
+ "B-Quantidade (kg)": 61,
65
+ "I-Quantidade (kg)": 62,
66
+ "B-Quantidade (m³)": 63,
67
+ "I-Quantidade (m³)": 64,
68
+ "B-Querosene de Aviação": 65,
69
+ "I-Querosene de Aviação": 66,
70
+ "B-Reciclagem": 67,
71
+ "I-Reciclagem": 68,
72
+ "B-Reflorestamento": 69,
73
+ "I-Reflorestamento": 70,
74
+ "B-Refrigerantes (HFCs)": 71,
75
+ "I-Refrigerantes (HFCs)": 72,
76
+ "B-Tratamento de Efluentes": 73,
77
+ "I-Tratamento de Efluentes": 74,
78
+ "B-Trem": 75,
79
+ "I-Trem": 76,
80
+ "B-kWh": 77,
81
+ "I-kWh": 78,
82
+ "B-Área (hectares)": 79,
83
+ "I-Área (hectares)": 80,
84
+ "B-Óleo Combustível": 81,
85
+ "I-Óleo Combustível": 82,
86
+ "B-Óleo Diesel": 83,
87
+ "I-Óleo Diesel": 84,
88
+ "B-Ônibus": 85,
89
+ "I-Ônibus": 86
90
+ },
91
+ "id_to_label": {
92
+ "0": "O",
93
+ "1": "B-Aterro",
94
+ "2": "I-Aterro",
95
+ "3": "B-Biodiesel (B100)",
96
+ "4": "I-Biodiesel (B100)",
97
+ "5": "B-Biometano",
98
+ "6": "I-Biometano",
99
+ "7": "B-Biometanol",
100
+ "8": "I-Biometanol",
101
+ "9": "B-Bioquerosene (SAF)",
102
+ "10": "I-Bioquerosene (SAF)",
103
+ "11": "B-Carro",
104
+ "12": "I-Carro",
105
+ "13": "B-Carvão",
106
+ "14": "I-Carvão",
107
+ "15": "B-Chegada",
108
+ "16": "I-Chegada",
109
+ "17": "B-Compostagem",
110
+ "18": "I-Compostagem",
111
+ "19": "B-Desmatamento",
112
+ "20": "I-Desmatamento",
113
+ "21": "B-Distância (km)",
114
+ "22": "I-Distância (km)",
115
+ "23": "B-Dióxido de Carbono (CO2)",
116
+ "24": "I-Dióxido de Carbono (CO2)",
117
+ "25": "B-Eletricidade",
118
+ "26": "I-Eletricidade",
119
+ "27": "B-Etanol Anidro",
120
+ "28": "I-Etanol Anidro",
121
+ "29": "B-Etanol Hidratado",
122
+ "30": "I-Etanol Hidratado",
123
+ "31": "B-Gasolina Automotiva",
124
+ "32": "I-Gasolina Automotiva",
125
+ "33": "B-Gasolina de Aviação",
126
+ "34": "I-Gasolina de Aviação",
127
+ "35": "B-Gás Liquefeito de Petróleo (GLP)",
128
+ "36": "I-Gás Liquefeito de Petróleo (GLP)",
129
+ "37": "B-Gás Natural",
130
+ "38": "I-Gás Natural",
131
+ "39": "B-Gás Natural Liquefeito (GNL)",
132
+ "40": "I-Gás Natural Liquefeito (GNL)",
133
+ "41": "B-Gás Natural Veicular (GNV)",
134
+ "42": "I-Gás Natural Veicular (GNV)",
135
+ "43": "B-HVO (diesel verde)",
136
+ "44": "I-HVO (diesel verde)",
137
+ "45": "B-Incineração",
138
+ "46": "I-Incineração",
139
+ "47": "B-Lançamento em Corpos d'Água",
140
+ "48": "I-Lançamento em Corpos d'Água",
141
+ "49": "B-Lubrificantes",
142
+ "50": "I-Lubrificantes",
143
+ "51": "B-Metano (CH4)",
144
+ "52": "I-Metano (CH4)",
145
+ "53": "B-Metanol",
146
+ "54": "I-Metanol",
147
+ "55": "B-Partida",
148
+ "56": "I-Partida",
149
+ "57": "B-Quantidade (Litros)",
150
+ "58": "I-Quantidade (Litros)",
151
+ "59": "B-Quantidade (kWh)",
152
+ "60": "I-Quantidade (kWh)",
153
+ "61": "B-Quantidade (kg)",
154
+ "62": "I-Quantidade (kg)",
155
+ "63": "B-Quantidade (m³)",
156
+ "64": "I-Quantidade (m³)",
157
+ "65": "B-Querosene de Aviação",
158
+ "66": "I-Querosene de Aviação",
159
+ "67": "B-Reciclagem",
160
+ "68": "I-Reciclagem",
161
+ "69": "B-Reflorestamento",
162
+ "70": "I-Reflorestamento",
163
+ "71": "B-Refrigerantes (HFCs)",
164
+ "72": "I-Refrigerantes (HFCs)",
165
+ "73": "B-Tratamento de Efluentes",
166
+ "74": "I-Tratamento de Efluentes",
167
+ "75": "B-Trem",
168
+ "76": "I-Trem",
169
+ "77": "B-kWh",
170
+ "78": "I-kWh",
171
+ "79": "B-Área (hectares)",
172
+ "80": "I-Área (hectares)",
173
+ "81": "B-Óleo Combustível",
174
+ "82": "I-Óleo Combustível",
175
+ "83": "B-Óleo Diesel",
176
+ "84": "I-Óleo Diesel",
177
+ "85": "B-Ônibus",
178
+ "86": "I-Ônibus"
179
+ }
180
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6d88acb3cc285af0fb51d31951758bea2a04c329284b7834ecce81b84e6f213b
3
+ size 433621132
special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "mask_token": "[MASK]",
4
+ "pad_token": "[PAD]",
5
+ "sep_token": "[SEP]",
6
+ "unk_token": "[UNK]"
7
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": false,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 1000000000000000019884624838656,
51
+ "never_split": null,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "strip_accents": null,
55
+ "tokenize_chinese_chars": true,
56
+ "tokenizer_class": "BertTokenizer",
57
+ "unk_token": "[UNK]"
58
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f08bbb5a047eb222170b0cd23aa687c92d3b3536131d24cb029f7fdd7d8ef695
3
+ size 5713
vocab.txt ADDED
The diff for this file is too large to render. See raw diff