Nico Gürtler

Hello and thank you for your answer! The episode transitions are tuples containing the last observation, action, reward, new observation and the done boolean. https://github.com/hill-a/stable-baselines/blob/4fada47f1b71b7548c935b1f01c6fb04199b3d54/stable_baselines/her/replay_buffer.py#L75 The goal selection strategy final...