Multiple trials with same set of parameters using bayesianoptimization and parallelTrialCount > 1

[fabianvdW]

/kind bug

What steps did you take and what happened:
[A clear and concise description of what the bug is.]
Using bayesianoptimization and parallelTrialCount > 1 will result in multiple trials having suggested the same set of parameters. This is due to the way

katib/pkg/suggestion/v1beta1/skopt/base_service.py

Lines 121 to 125 in 2db65b2

	for i in range(request_number):
	skopt_suggested = self.skopt_optimizer.ask()
	logger.info("New suggested parameters for Trial: {}".format(skopt_suggested))
	return_trial_list.append(
	BaseSkoptService.convert(self.search_space, skopt_suggested))

are structured

[fabianvdW]

For a quick demonstration of the issue run:

import skopt
import random
opt = skopt.Optimizer(dimensions = [(0.,1.)])
for i in range(10):
    v = opt.ask()
    opt.tell(v, random.random())
for _ in range(10):
    print(opt.ask())