File size: 4,860 Bytes
839f981 8519bc1 839f981 5612cee a7cc914 b945833 a7cc914 cc5e011 a7cc914 5612cee a7cc914 5612cee a7cc914 5612cee a7cc914 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 |
---
language:
- pt
license: cc-by-4.0
datasets:
- wiki_lingua
thumbnail: null
tags:
- named-entity-recognition
- Transformer
- pytorch
- bert
metrics:
- f1
- precision
- recall
model-index:
- name: rpunct-ptbr
results:
- task:
type: named-entity-recognition
dataset:
type: wiki_lingua
name: wiki_lingua
metrics:
- type: f1
value: 55.70
name: F1 Score
- type: precision
value: 57.72
name: Precision
- type: recall
value: 53.83
name: Recall
widget:
- text: "henrique foi no lago pescar com o pedro mais tarde foram para a casa do pedro fritar os peixes"
- text: "cinco trabalhadores da construção civil em capacetes e coletes amarelos estão ocupados no trabalho"
- text: "na quinta feira em visita a belo horizonte pedro sobrevoa a cidade atingida pelas chuvas"
- text: "coube ao representante de classe contar que na avaliação de língua portuguesa alguns alunos se mantiveram concentrados e outros dispersos"
---
# 🤗 bert-restore-punctuation-ptbr
* 🪄 [W&B Dashboard](https://wandb.ai/dominguesm/RestorePunctuationPTBR)
**Coming soon python package for simpler use.**
This is a [bert-base-portuguese-cased](https://huggingface.co/neuralmind/bert-base-portuguese-cased) model finetuned for punctuation restoration on [WikiLingua](https://github.com/esdurmus/Wikilingua).
This model is intended for direct use as a punctuation restoration model for the general Portuguese language. Alternatively, you can use this for further fine-tuning on domain-specific texts for punctuation restoration tasks.
Model restores the following punctuations -- **[! ? . , - : ; ' ]**
The model also restores the upper-casing of words.
-----------------------------------------------
## 🎯 Accuracy
| label | precision | recall | f1-score | support|
| ------------------------- | -------------|-------- | ----------|--------|
| **Upper - OU** | 0.89 | 0.91 | 0.90 | 69376
| **None - OO** | 0.99 | 0.98 | 0.98 | 857659
| **Full stop/period - .O** | 0.86 | 0.93 | 0.89 | 60410
| **Comma - ,O** | 0.85 | 0.83 | 0.84 | 48608
| **Upper + Comma - ,U** | 0.73 | 0.76 | 0.75 | 3521
| **Question - ?O** | 0.68 | 0.78 | 0.73 | 1168
| **Upper + period - .U** | 0.66 | 0.72 | 0.69 | 1884
| **Upper + colon - :U** | 0.59 | 0.63 | 0.61 | 352
| **Colon - :O** | 0.70 | 0.53 | 0.60 | 2420
| **Question Mark - ?U** | 0.50 | 0.56 | 0.53 | 36
| **Upper + Exclam. - !U** | 0.38 | 0.32 | 0.34 | 38
| **Exclamation Mark - !O** | 0.30 | 0.05 | 0.08 | 783
| **Semicolon - ;O** | 0.35 | 0.04 | 0.08 | 1557
| **Apostrophe - 'O** | 0.00 | 0.00 | 0.00 | 3
| **Hyphen - -O** | 0.00 | 0.00 | 0.00 | 3
| | | | |
| **accuracy** | | | 0.96 | 1047818
| **macro avg** | 0.57 | 0.54 | 0.54 | 1047818
| **weighted avg** | 0.96 | 0.96 | 0.96 | 1047818
-----------------------------------------------
## 🤷 Output
Example:
```json
[
{
"entity_group": "OU",
"score": 0.8026431202888489,
"word": "henrique",
"start": 0,
"end": 8
},
{
"entity_group": "OO",
"score": 0.9925149083137512,
"word": "foi no lago pescar com o",
"start": 9,
"end": 33
},
{
"entity_group": ".U",
"score": 0.8426014184951782,
"word": "pedro",
"start": 34,
"end": 39
},
{
"entity_group": "OU",
"score": 0.9519776105880737,
"word": "mais",
"start": 40,
"end": 44
},
{
"entity_group": ",O",
"score": 0.8551820516586304,
"word": "tarde",
"start": 45,
"end": 50
},
{
"entity_group": "OO",
"score": 0.9902807474136353,
"word": "foram para a casa do",
"start": 51,
"end": 71
},
{
"entity_group": "OU",
"score": 0.9227372407913208,
"word": "pedro",
"start": 72,
"end": 77
},
{
"entity_group": "OO",
"score": 0.9997054934501648,
"word": "fritar os",
"start": 78,
"end": 87
},
{
"entity_group": ".O",
"score": 0.9813661575317383,
"word": "peixes",
"start": 88,
"end": 94
}
]
```
This output refers to:
```
Henrique foi no lago pescar com o Pedro. Mais tarde, foram para a casa do Pedro fritar os peixes.
```
-----------------------------------------------
## 🤙 Contact
[Maicon Domingues]([email protected]) for questions, feedback and/or requests for similar models.
|