Confusion about worker setup

[tlhr]

I'm trying to use the python API to parametrize a largish set of molecules in the most efficient way possible. I'm a bit confused about how resources are allocated. I'm currently doing something like this:

    with BespokeExecutor(
        n_fragmenter_workers=1,
        n_optimizer_workers=1,
        n_qc_compute_workers=3,
        qc_compute_worker_config=BespokeWorkerConfig(n_cores=2),
    ) as exec:
        tasks = [exec.submit(input_schema=schema) for schema in schemas]
        outputs = [wait_until_complete(task) for task in tasks]

My impression was that in this case, the BespokeExecutor is running with a total of 5 workers, of which 3 (the QC workers) use 2 cores each, for a total of 8 cores (the limit of my machine). However when running this with 4 schemas, I see far more processes running, causing oversubscription and thus high inefficiency (after a few minutes I can see at least 15 processes, but it's a bit hard to keep track with the different stages).

What's the best way of handling larger batches like this? Or is this a non-issue? Thank you :)