IsaacGym AllegroHand throws an error when returning the mean rewards in enjoy:
File "/home/andrewzhang/sample-factory/sample_factory/enjoy.py", line 269, in enjoy
return ExperimentStatus.SUCCESS, float(np.mean(episode_rewards))
TypeError: only size-1 arrays can be converted to Python scalars
Seems like this happens if some agents do a different number of episodes than other agents.
I changed it to do the same thing as logging.
IsaacGym AllegroHand throws an error when returning the mean rewards in enjoy:
Seems like this happens if some agents do a different number of episodes than other agents. I changed it to do the same thing as logging.