Is there a way to specify the initial population in optuna's NSGA-II?

Question

I created a neural network model that predicts certain properties from coordinates.

Using that model, I want to find the coordinates that minimize the properties in optuna's NSGA-II sampler.

Normally, we would generate a random initial population by specifying a range of coordinates.

However, I would like to include the coordinates used to construct the neural network as part of the initial population.

Is there any way to do it?

The following is a sample code. I want to include a part of the value specified by myself in the "#" part like x, y = [3, 2], [4.2, 1.4]

import optuna
import matplotlib.pyplot as plt
%matplotlib inline

import warnings
warnings.simplefilter('ignore')

def objective(trial):
    x = trial.suggest_uniform("x", 0, 5)   #This is the normal way
    y = trial.suggest_uniform("y", 0, 3)   #This is the normal way
    v0 = 4 * x ** 2 + 4 * y ** 2
    v1 = (x - 5) ** 2 + (y - 5) ** 2
    return v0, v1

study = optuna.multi_objective.create_study(
    directions=["minimize", "minimize"], 
    sampler=optuna.multi_objective.samplers.NSGAIIMultiObjectiveSampler()
)

study.optimize(objective, n_trials=100)