finalize with recipe

Right now for a random forest model I am doing something like finalize(mtry_long(), train_data). It guesses the number of columns, but I have a recipe that marks one column as the response and one as an ID column. Can finalize use this recipe info? Should I just specify this number to mtry() itself?

If you are using the model and recipe in a workflow together in {tune} it should deal with this automatically