set the default encoding to utf-8

aio_from_tohoku.py

@@ -60,7 +60,7 @@ class Aio(datasets.GeneratorBasedBuilder):
         return split_generators
 
     def _generate_examples(self, filepath: str) -> Iterator[Tuple[int, Dict[str, Any]]]:
-        with open(filepath) as f:
+        with open(filepath, encoding="utf-8") as f:
             for i, line in enumerate(f):
                 item = json.loads(line)
                 yield i, item