public void UpdateParams(int p_e, int p_epochs)
{
_exploration.UpdateParams((float)p_e / (p_epochs / 2));
Console.WriteLine("Exploration param >> " + _exploration.ToString());
_agent.GetOptimizer().UpdateAlpha((float)p_e / p_epochs);
Console.WriteLine("Learning param >> " + _agent.GetOptimizer().Alpha.ToString());
}