Datasets:
kardosdrur
commited on
Commit
•
14fbdc9
1
Parent(s):
3dcfa84
Added dataset creation script
Browse files- build_dataset.py +20 -0
build_dataset.py
ADDED
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import xml.etree.ElementTree as ET
|
2 |
+
|
3 |
+
from datasets import Dataset
|
4 |
+
|
5 |
+
print("Parsing alignment file.")
|
6 |
+
tree = ET.parse("./nb-nn.tmx")
|
7 |
+
|
8 |
+
|
9 |
+
records = []
|
10 |
+
pairs = tree.findall(".//tu")
|
11 |
+
for pair in pairs:
|
12 |
+
nb = pair.find("./tuv[@{http://www.w3.org/XML/1998/namespace}lang = 'nb']/seg").text
|
13 |
+
nn = pair.find("./tuv[@{http://www.w3.org/XML/1998/namespace}lang = 'nn']/seg").text
|
14 |
+
records.append(dict(nb=nb, nn=nn))
|
15 |
+
|
16 |
+
ds = Dataset.from_list(records)
|
17 |
+
ds = ds.shuffle(seed=42)
|
18 |
+
ds = ds.train_test_split(seed=42, test_size=0.2)
|
19 |
+
|
20 |
+
ds.push_to_hub("kardosdrur/norwegian-courts")
|