Hi!
Thanks for the library!
Could you write up some documentation on how those options work for ParallelStream.map?
My intuition says that as long as num_workers: 1, it's effectively not parallel, and worker_work_ratio is for managing the batch size passed to each worker for fine tuning acceptable parallel overhead?