So i have a dataset of simulations that have a bunch of parameters that are of 2 groups, lets call them config_params and results_params. For every simulation I also have a time of running. I would like to use keras to create model that would given config_params return me results_params. The problem i have is that I dont know how to use time in a model to get possibly best result_params in a way that smaller time = better. How should I approach this?