I wonder if there is a way to evaluate batches of individuals when running deap ? The classic implementations evaluate individuals one by one but my evaluation function requires me to evaluate the inidviduals with a surrogate model that is only accessible online and connecting to the server where this model is kept is time-consuming so I would prefer directly sending the whole population to the server for evaluation instead of having to connect for each individual.
(It takes approx. 40 seconds to get my results whatever the number of evaluation I am asking and there is no way for me to change this as it is due to the software I am connecting to that can't be changed)
Evaluation function is called with : fitnesses = toolbox.map(toolbox.evaluate, invalid_ind) and I have previously defined toolbox.register("evaluate", evaluation_fct)
