I have some code that uses asyncio to map over a a large dataset by spawning a few hundred workers. Is there a way to use multiple cores, to make it even faster?

from tqdm import tqdm
import asyncio
import aiohttp
from pathlib import Path
from collections import Counter


WORKER_COUNT = 500  # Number of workers


async def fetch_and_save(input_item):
     # do work
     print("working...")
     return result_code

async def worker(task_queue, session, results, pbar):
    while not task_queue.empty():
        item_id = await task_queue.get()
        result = await fetch_and_save(item_id, session)
        results[result] += 1
        task_queue.task_done()

async def main():
    async with aiohttp.ClientSession() as session:
        task_queue = asyncio.Queue()
        results = Counter()
        
        # Queue all items
        ids = read_items_from_disk()
        import random
        random.seed(42)

        random.shuffle(ids)

        for id in ids:
            await task_queue.put(id)
        pbar = tqdm(total=task_queue.qsize())

        # Create worker coroutines
        workers = [asyncio.create_task(worker(task_queue, session, results, pbar)) for _ in range(WORKER_COUNT)]

        # Wait for all tasks in the queue to be processed
        await task_queue.join()

        # Cancel all remaining workers after queue is empty
        for worker_task in workers:
            worker_task.cancel()

asyncio.run(main())