Walker 2D Pybullet Environment

Question

I am trying to solve the Walker2DBulletEnv-v0 by implementing the SAC algorithm. Around the first 700 episodes the robot keeps its balance returning a score about ~600. Are these promising results and should I just keep running the program in order for networks to improve or should I try adjusting the hyperparameters. Moreover if these results aren't promising which results are so considered "good";