Hamed Masoudi

Hi, I think you're right. I trained the decision transformer in maze2d-medium-dense-v1 environment and calculated the normalized score with this command: `env.get_normalized_score(average return of 100 episodes)`. However, I obtained a...