@task
def make_nodes(
    nodes_to_make_lst,
    some_arguments
):

    # Step 1 of the we extract the accumulated task loop result from context
    loop_payload = prefect.context.get("task_loop_result", {})

    # This is what gets returned from loop result
    nodes_to_make_lst = loop_payload.get("nodes_to_make_lst", nodes_to_make_lst)
    #context_df = loop_payload.get("context_df", args.context_df)
    
    # Make nodes in parallel  
	####!!! This is the only thing I care about parallelizing !!!#####
    with Flow("Prallelize Make node") as parflow:

        make_node_response = make_node.map(
            node_key_dict=nodes_to_make_lst,
            some_arguments
        )
    get_defs_state = parflow.run(
        # executor=executor
    )
    make_nodes_lst = get_defs_state.result[make_node_response].result[0]
    
    pprint(f"make_node_response {type(make_nodes_lst)}")
    terms_to_get = [d['term'] for d in make_nodes_lst if "term" in d]
    
    pprint(f"Making new level")
    pprint(f"{terms_to_get}")

    db_conn = lmap_io.db_connect.run(graph_db_name=graph_db_name)
    num_nodes_completed = lmap_io.count_nodes.run(db_conn["graph_db"])
    pprint(num_nodes_completed)
    if num_nodes_completed>=max_num_nodes:

        logger.info(f"Finished Making Knowledge Graph: {num_nodes_completed}")
        return  num_nodes_completed

    raise LOOP(
                message=f"{len(make_nodes_lst)}", 
                result=dict(
                    nodes_to_make_lst=make_nodes_lst
                )
    )