当我尝试使用saver = tf.train.import_meta_graph(meta_graph_path, clear_devices=True)导入元图时，我从导入器中获取KeyError：'MaxBytesInUse'。
Tensorflow版本：1.7-gpu-python3 操作系统：Ubuntu 16.04
以下是错误的堆栈跟踪：
/usr/local/lib/python3.5/dist-packages/tensorflow/python/training/saver.py in import_meta_graph(meta_graph_or_file, clear_devices, import_scope, **kwargs)
1953 clear_devices=clear_devices,
1954 import_scope=import_scope,
-> 1955 **kwargs)
1956
1957 if meta_graph_def.HasField("saver_def"):
/usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/meta_graph.py in import_scoped_meta_graph(meta_graph_or_file, clear_devices, graph, import_scope, input_map, unbound_inputs_col_name, restore_collections_predicate)
741 name=(import_scope or scope_to_prepend_to_names),
742 input_map=input_map,
--> 743 producer_op_list=producer_op_list)
744
745 # Restores all the other collections.
/usr/local/lib/python3.5/dist-packages/tensorflow/python/util/deprecation.py in new_func(*args, **kwargs)
430 'in a future version' if date is None else ('after %s' % date),
431 instructions)
--> 432 return func(*args, **kwargs)
433 return tf_decorator.make_decorator(func, new_func, 'deprecated',
434 _add_deprecated_arg_notice_to_docstring(
/usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/importer.py in import_graph_def(graph_def, input_map, return_elements, name, op_dict, producer_op_list)
458 if producer_op_list is not None:
459 # TODO(skyewm): make a copy of graph_def so we're not mutating the argument?
--> 460 _RemoveDefaultAttrs(op_dict, producer_op_list, graph_def)
461
462 graph = ops.get_default_graph()
/usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/importer.py in _RemoveDefaultAttrs(op_dict, producer_op_list, graph_def)
225 # Remove any default attr values that aren't in op_def.
226 if node.op in producer_op_dict:
--> 227 op_def = op_dict[node.op]
228 producer_op_def = producer_op_dict[node.op]
229 # We make a copy of node.attr to iterate through since we may modify
KeyError: 'MaxBytesInUse'