I am using hyperopt in python. I wanted to create some solutions as a list and feed them to hyperopt as trials, So I used generate_trials_to_calculate method to create the trials:
trials=generate_trials_to_calculate(vals_list)
best = fmin(fn=objective, space=space, algo=tpe.suggest, trials=trials, max_evals=iteration_number)
This code runs only on one core of CPU. Now, I want to use "SparkTrials" in order to run it in parallel among the cores of CPU, but I don't know how to generate a "SparkTrials" with some initial solutions. Is there anybody to know how I can do it?