How to get the main graph index of each batch_data

[streetcorner]

Hello, after I use clusterloader to separate my graph data, I want to add an attribute to each batch_Data for model training. How can I know the node index of each node in the entire graph in each batch_data?
my torch version is 1.12.1，There is no n_id or node_idx in the batch_data attributes

My code example is as follows：
class CustomClusterLoader(ClusterLoader):
def init(self, data, total_attr, batch_size, shuffle=True):
cluster_data = ClusterData(data, num_parts=100)
super().init(cluster_data, batch_size=batch_size, shuffle=shuffle)
self.total_attr = total_attr
for batch_data in super().iter():
print(f"batch_data keys: {batch_data.keys()}")
print(f"batch_data attributes: {dir(batch_data)}")

def __iter__(self):
    for batch_data in super().__iter__():  
        curent_attr = "xxx"
        batch_data.extra_attr = curent_attr
        yield batch_data  

curent_attr is curent_attr is the information that needs to be extracted from the main graph based on the node index. So I need to know the index of the node of batch_data in the main graph。
Thanks!