struct_builder.py并行处理chunks可能导致的负优化问题 #59
Description
graphrag_agent\graph\structure\struct_builder.py
parallel_process_chunks()函数：
with concurrent.futures.ThreadPoolExecutor(max_workers=max_workers) as executor:
future_to_batch = {
executor.submit(process_chunk_batch, batch, i * batch_size): i
for i, batch in enumerate(chunk_batches)
}
process_chunk_batch()应该是纯CPU操作，开的线程池因为GIL有可能是负优化。优化是否可以考虑写入数据库时加线程池？