oliverguhr
commited on
Commit
·
ac9a78c
1
Parent(s):
0b2070a
updated
Browse files
README.md
ADDED
@@ -0,0 +1,52 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
language:
|
3 |
+
- nl
|
4 |
+
tags:
|
5 |
+
- punctuation prediction
|
6 |
+
- punctuation
|
7 |
+
datasets: sonar
|
8 |
+
license: mit
|
9 |
+
widget:
|
10 |
+
- text: "hervatting van de zitting ik verklaar de zitting van het europees parlement die op vrijdag 17 december werd onderbroken te zijn hervat"
|
11 |
+
example_title: "Euro Parl Sample"
|
12 |
+
metrics:
|
13 |
+
- f1
|
14 |
+
---
|
15 |
+
|
16 |
+
## Model
|
17 |
+
|
18 |
+
Trained on Sonar corpus
|
19 |
+
|
20 |
+
## Performance
|
21 |
+
|
22 |
+
Evaluated on dutch Euro Parl
|
23 |
+
```
|
24 |
+
precision recall f1-score support
|
25 |
+
|
26 |
+
0 0.990421 0.994986 0.992698 9627605
|
27 |
+
. 0.942931 0.948408 0.945662 433554
|
28 |
+
, 0.813030 0.773804 0.792932 379759
|
29 |
+
? 0.806700 0.790499 0.798518 13494
|
30 |
+
- 0.606461 0.045317 0.084332 27341
|
31 |
+
: 0.599856 0.501284 0.546158 18305
|
32 |
+
|
33 |
+
accuracy 0.981467 10500058
|
34 |
+
macro avg 0.793233 0.675716 0.693383 10500058
|
35 |
+
weighted avg 0.980127 0.981467 0.980138 10500058
|
36 |
+
|
37 |
+
```
|
38 |
+
|
39 |
+
Usage:
|
40 |
+
|
41 |
+
```bash
|
42 |
+
pip install deepmultilingualpunctuation
|
43 |
+
```
|
44 |
+
|
45 |
+
```python
|
46 |
+
from deepmultilingualpunctuation import PunctuationModel
|
47 |
+
|
48 |
+
model = PunctuationModel(model="oliverguhr/fullstop-dutch-punctuation-prediction")
|
49 |
+
text = "hervatting van de zitting ik verklaar de zitting van het europees parlement die op vrijdag 17 december werd onderbroken te zijn hervat"
|
50 |
+
result = model.restore_punctuation(text)
|
51 |
+
print(result)
|
52 |
+
```
|