Fix encoding in reader

nkjp-pos.py

@@ -124,7 +124,7 @@ class NKJPPOS(datasets.GeneratorBasedBuilder):
     def _generate_examples(
             self, filepath: str
     ) -> Generator[Tuple[int, Dict[str, str]], None, None]:
-        with open(filepath, 'r') as f:
+        with open(filepath, 'r', encoding="utf-8") as f:
             reader = csv.reader(f, delimiter='\t', quoting=csv.QUOTE_NONE)
 
             tokens = []