I am running a linear mixed effects model with a dataset consisting of ~500K observations. I have a continuous response (i.e. Y), 24 categorical variables (i.e. X1,..., X24)