In our experiments, the standard DisplayR choice model code takes a long time to run, often over 3 hours with warnings about non-optimal solutions. We need a faster execution to ensure optimal results. Could DisplayR optimize the existing package to execute faster?