Automatically infer the number of available cores and allocate the majority of them to parallel processing. Reserves a small proportion of cores to avoid slowing or crashing.

assign_cores(worker_cores = 0.9, verbose = T)