Resetting the Optimizer in Deep RL: An Empirical Study | Read Paper on Bytez