Update openwebtext_split.py

openwebtext_split.py

@@ -55,8 +55,8 @@ def custom_iter_archive(path_or_buf, _filter=lambda x: True):
                 continue
             if not file_path.endswith('xz'):
                 continue
-            file_obj = stream.extract(tarinfo)
-            print(type(file_obj))
+            file_obj = stream.extractfile(tarinfo)
+            print(file_obj.getmembers())
             for j, xzinfo in enumerate(file_obj):
                 f = io.BytesIO(xzinfo)
                 yield f