Added nordjylland news

data/nordjyllandnews/create.py

@@ -4,11 +4,9 @@ This scripts download nordjylland news and converts it to the format of danish d
 
 import random
 from pathlib import Path
+from typing import cast
 
-from datasets import load_dataset
-
-source = "nordjyllandnews"
-ds = load_dataset("alexandrainst/nordjylland-news-summarization", split="train")
+from datasets import Dataset, load_dataset
 
 schemas = [
     "{summary}\n\n{text}",
@@ -16,13 +14,13 @@ schemas = [
     "{text}\n\nReferat:\n{summary}",
     "Lav et referat af nedenstående tekst:\n\nTekst:\n{text}\n\nReferat:\n{summary}",
 ]
+source = "nordjyllandnews"
 
 
 def convert_sample(example):
     schema = random.sample(schemas, k=1)[0]
-
     new_example = dict(
-        text=schema.format(text=example["text"], summary=example["summary"]),
+        text_new=schema.format(text=example["text"], summary=example["summary"]),
         source=source,
         domain="News",
         license="Creative Commons Legal Code\n\nCC0 1.0 Universal",
@@ -34,7 +32,20 @@ def convert_sample(example):
     return new_example
 
 
-ds = ds.map(convert_sample)
+def main():
+    ds = load_dataset("alexandrainst/nordjylland-news-summarization", split="train")
+    ds = cast(Dataset, ds)
+
+    ds = ds.map(convert_sample, remove_columns=ds.column_names)
+    ds = ds.rename_columns({"text_new": "text"})
+    ds = ds.add_column("id", [f"{source}_{i}" for i in range(len(ds))])  # type: ignore
+    ds = ds.select_columns(
+        ["text", "source", "id", "domain", "license", "added", "created", "metadata"]
+    )
+
+    save_path = Path(__file__).parent / f"{source}.parquet"
+    ds.to_parquet(save_path)
+
 
-save_path = Path(__file__).parent / f"{source}.parquet"
-ds.to_parquet(save_path)
+if __name__ == "__main__":
+    main()

data/nordjyllandnews/nordjyllandnews.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:438950abf3c8b922fabf84caa08f4b9df170b0e1e6606c071a721dd21323cfe3
-size 78048869
+oid sha256:c577afba60df86561471214e53122783bff296b440e743926948c4fe01cd7b97
+size 71246640

uv.lock

@@ -199,7 +199,7 @@ wheels = [
 
 [[package]]
 name = "danish-dynaword"
-version = "1.0.2"
+version = "1.0.3"
 source = { virtual = "." }
 dependencies = [
     { name = "datasets" },