hi folks. I am using a dask executor to perform tasks on multiple machines via map (). but it happens that the processing of one task takes a very long time and because I work with a pandas and I cannot parallelize the calculations already inside map (). I would like to use modin for this. has anyone tried something like this?

Hi @bral - I haven't personally used modin so others might want to chime in here but I see no reason you couldn't use it, particularly in conjunction with Prefect resource manager tasks (depending on how you want to manage your dask resources).