RL-lander / results.json

It seems that DQN performs the worst if trained for 1e6 timesteps. But it did train quicker, taking about 17 min, as opposed to 20-22.

3f14e63 almost 4 years ago

Raw History Blame Contribute Delete

165 Bytes

{"mean_reward": -29.892793940205593, "std_reward": 28.12063664298142, "is_deterministic": true, "n_eval_episodes": 10, "eval_datetime": "2022-12-15T20:38:34.501158"}