How to use the cartography.intel.gcp.compute.load_gcp_instances function in cartography

def _ensure_local_neo4j_has_test_instance_data(neo4j_session):
    cartography.intel.gcp.compute.load_gcp_instances(
        neo4j_session,
        tests.data.gcp.compute.TRANSFORMED_GCP_INSTANCES,
        TEST_UPDATE_TAG,
    )

def test_transform_and_load_gcp_instances_and_nics(neo4j_session):
    """
    Ensure that we can correctly transform and load GCP instances.
    """
    instance_responses = [tests.data.gcp.compute.GCP_LIST_INSTANCES_RESPONSE]
    instance_list = cartography.intel.gcp.compute.transform_gcp_instances(instance_responses)
    cartography.intel.gcp.compute.load_gcp_instances(neo4j_session, instance_list, TEST_UPDATE_TAG)

    instance_id1 = 'projects/project-abc/zones/europe-west2-b/instances/instance-1-test'
    instance_id2 = 'projects/project-abc/zones/europe-west2-b/instances/instance-1'

    nic_query = """
    MATCH(i:GCPInstance)-[r:NETWORK_INTERFACE]->(nic:GCPNetworkInterface)
    OPTIONAL MATCH (i)-[:TAGGED]->(t:GCPNetworkTag)
    RETURN i.id, i.zone_name, i.project_id, i.hostname, t.value, r.lastupdated, nic.nic_id, nic.private_ip
    """
    objects = neo4j_session.run(nic_query)
    actual_nodes = {
        (
            o['i.id'],
            o['i.zone_name'],
            o['i.project_id'],
            o['nic.nic_id'],