Update openwebtext_split.py

openwebtext_split.py

@@ -55,11 +55,9 @@ def custom_iter_archive(path_or_buf, _filter=lambda x: True):
             if not file_path.endswith('xz'):
                 continue
             file_obj = stream.extractfile(tarinfo)
-            print(dir(file_obj))
-            with tarfile.open(file_obj, 'rb') as inner_stream:
-                for j, xzinfo in enumerate(inner_stream):
-                    print(type(xzinfo))
-                    yield xzinfo.name, xzinfo
+            for j, xzinfo in enumerate(file_obj):
+                print(type(xzinfo))
+                yield xzinfo.name, xzinfo
             stream.members = []
         del stream
     if hasattr(path_or_buf, "read"):