2023-05-16 11:51:26 - SimpleLog - INFO: - General Configs: 2023-05-16 11:51:26 - SimpleLog - INFO: - ================================================================================ 2023-05-16 11:51:26 - SimpleLog - INFO: - Name Value Type 2023-05-16 11:51:26 - SimpleLog - INFO: - env_name gym 2023-05-16 11:51:26 - SimpleLog - INFO: - algo_name DoubleDQN 2023-05-16 11:51:26 - SimpleLog - INFO: - mode train 2023-05-16 11:51:26 - SimpleLog - INFO: - device cpu 2023-05-16 11:51:26 - SimpleLog - INFO: - seed 1 2023-05-16 11:51:26 - SimpleLog - INFO: - max_episode 100 2023-05-16 11:51:26 - SimpleLog - INFO: - max_step 200 2023-05-16 11:51:26 - SimpleLog - INFO: - collect_traj 0 2023-05-16 11:51:26 - SimpleLog - INFO: - mp_backend ray 2023-05-16 11:51:26 - SimpleLog - INFO: - n_workers 2 2023-05-16 11:51:26 - SimpleLog - INFO: - online_eval 1 2023-05-16 11:51:26 - SimpleLog - INFO: - online_eval_episode 10 2023-05-16 11:51:26 - SimpleLog - INFO: - model_save_fre 500 2023-05-16 11:51:26 - SimpleLog - INFO: - load_checkpoint 0 2023-05-16 11:51:26 - SimpleLog - INFO: - load_path Train_CartPole-v1_DQN_20221026-054757 2023-05-16 11:51:26 - SimpleLog - INFO: - load_model_step best 2023-05-16 11:51:26 - SimpleLog - INFO: - ================================================================================ 2023-05-16 11:51:26 - SimpleLog - INFO: - Algo Configs: 2023-05-16 11:51:26 - SimpleLog - INFO: - ================================================================================ 2023-05-16 11:51:26 - SimpleLog - INFO: - Name Value Type 2023-05-16 11:51:26 - SimpleLog - INFO: - epsilon_start 0.95 2023-05-16 11:51:26 - SimpleLog - INFO: - epsilon_end 0.01 2023-05-16 11:51:26 - SimpleLog - INFO: - epsilon_decay 500 2023-05-16 11:51:26 - SimpleLog - INFO: - gamma 0.99 2023-05-16 11:51:26 - SimpleLog - INFO: - lr 0.0001 2023-05-16 11:51:26 - SimpleLog - INFO: - buffer_size 100000 2023-05-16 11:51:26 - SimpleLog - INFO: - batch_size 64 2023-05-16 11:51:26 - SimpleLog - INFO: - target_update 4 2023-05-16 11:51:26 - SimpleLog - INFO: - value_layers [{'layer_type': 'linear', 'layer_dim': [256], 'activation': 'relu'}, {'layer_type': 'linear', 'layer_dim': [256], 'activation': 'relu'}] 2023-05-16 11:51:26 - SimpleLog - INFO: - buffer_type REPLAY_QUE 2023-05-16 11:51:26 - SimpleLog - INFO: - ================================================================================ 2023-05-16 11:51:26 - SimpleLog - INFO: - Env Configs: 2023-05-16 11:51:26 - SimpleLog - INFO: - ================================================================================ 2023-05-16 11:51:26 - SimpleLog - INFO: - Name Value Type 2023-05-16 11:51:26 - SimpleLog - INFO: - id CartPole-v1 2023-05-16 11:51:26 - SimpleLog - INFO: - render_mode None 2023-05-16 11:51:26 - SimpleLog - INFO: - wrapper None 2023-05-16 11:51:26 - SimpleLog - INFO: - ignore_params ['wrapper', 'ignore_params'] 2023-05-16 11:51:26 - SimpleLog - INFO: - ================================================================================ 2023-05-16 11:51:32 - SimpleLog - INFO: - obs_space: Box([-4.8000002e+00 -3.4028235e+38 -4.1887903e-01 -3.4028235e+38], [4.8000002e+00 3.4028235e+38 4.1887903e-01 3.4028235e+38], (4,), float32), n_actions: Discrete(2) 2023-05-16 11:51:37 - RayLog - INFO: - Worker 0 finished episode 0 with reward 16.0 in 16 steps 2023-05-16 11:51:37 - RayLog - INFO: - Worker 1 finished episode 0 with reward 20.0 in 20 steps 2023-05-16 11:51:39 - RayLog - INFO: - Worker 0 finished episode 2 with reward 11.0 in 11 steps 2023-05-16 11:51:39 - RayLog - INFO: - Worker 1 finished episode 2 with reward 15.0 in 15 steps 2023-05-16 11:51:39 - RayLog - INFO: - Worker 1 finished episode 4 with reward 13.0 in 13 steps 2023-05-16 11:51:39 - RayLog - INFO: - Worker 0 finished episode 3 with reward 22.0 in 22 steps 2023-05-16 11:51:39 - RayLog - INFO: - Worker 1 finished episode 5 with reward 9.0 in 9 steps 2023-05-16 11:51:40 - RayLog - INFO: - Worker 0 finished episode 6 with reward 14.0 in 14 steps 2023-05-16 11:51:40 - RayLog - INFO: - Worker 1 finished episode 7 with reward 12.0 in 12 steps 2023-05-16 11:51:40 - RayLog - INFO: - Worker 1 finished episode 9 with reward 13.0 in 13 steps 2023-05-16 11:51:40 - RayLog - INFO: - Worker 1 finished episode 10 with reward 14.0 in 14 steps 2023-05-16 11:51:40 - RayLog - INFO: - Worker 0 finished episode 8 with reward 35.0 in 35 steps 2023-05-16 11:51:40 - RayLog - INFO: - Worker 1 finished episode 11 with reward 13.0 in 13 steps 2023-05-16 11:51:40 - RayLog - INFO: - Worker 0 finished episode 12 with reward 15.0 in 15 steps 2023-05-16 11:51:40 - RayLog - INFO: - Worker 1 finished episode 13 with reward 12.0 in 12 steps 2023-05-16 11:51:41 - RayLog - INFO: - Worker 0 finished episode 14 with reward 10.0 in 10 steps 2023-05-16 11:51:41 - RayLog - INFO: - Worker 1 finished episode 15 with reward 11.0 in 11 steps 2023-05-16 11:51:41 - RayLog - INFO: - Worker 0 finished episode 16 with reward 11.0 in 11 steps 2023-05-16 11:51:41 - RayLog - INFO: - Worker 0 finished episode 18 with reward 12.0 in 12 steps 2023-05-16 11:51:41 - RayLog - INFO: - Worker 1 finished episode 17 with reward 21.0 in 21 steps 2023-05-16 11:51:41 - RayLog - INFO: - Worker 0 finished episode 19 with reward 15.0 in 15 steps 2023-05-16 11:51:41 - RayLog - INFO: - Worker 1 finished episode 20 with reward 18.0 in 18 steps 2023-05-16 11:51:41 - RayLog - INFO: - Worker 0 finished episode 21 with reward 13.0 in 13 steps 2023-05-16 11:51:42 - RayLog - INFO: - Worker 1 finished episode 22 with reward 10.0 in 10 steps 2023-05-16 11:51:42 - RayLog - INFO: - Worker 0 finished episode 23 with reward 16.0 in 16 steps 2023-05-16 11:51:42 - RayLog - INFO: - Worker 1 finished episode 24 with reward 19.0 in 19 steps 2023-05-16 11:51:42 - RayLog - INFO: - Worker 0 finished episode 25 with reward 10.0 in 10 steps 2023-05-16 11:51:42 - RayLog - INFO: - Worker 0 finished episode 27 with reward 10.0 in 10 steps 2023-05-16 11:51:42 - RayLog - INFO: - Worker 0 finished episode 28 with reward 10.0 in 10 steps 2023-05-16 11:51:42 - RayLog - INFO: - Worker 1 finished episode 26 with reward 26.0 in 26 steps 2023-05-16 11:51:42 - RayLog - INFO: - Worker 0 finished episode 29 with reward 12.0 in 12 steps 2023-05-16 11:51:43 - RayLog - INFO: - Worker 1 finished episode 30 with reward 15.0 in 15 steps 2023-05-16 11:51:43 - RayLog - INFO: - Worker 1 finished episode 32 with reward 10.0 in 10 steps 2023-05-16 11:51:43 - RayLog - INFO: - Worker 0 finished episode 31 with reward 18.0 in 18 steps 2023-05-16 11:51:43 - RayLog - INFO: - Worker 1 finished episode 33 with reward 13.0 in 13 steps 2023-05-16 11:51:43 - RayLog - INFO: - Worker 0 finished episode 34 with reward 10.0 in 10 steps 2023-05-16 11:51:43 - RayLog - INFO: - Worker 0 finished episode 36 with reward 9.0 in 9 steps 2023-05-16 11:51:43 - RayLog - INFO: - Worker 1 finished episode 35 with reward 12.0 in 12 steps 2023-05-16 11:51:45 - RayLog - INFO: - update_step: 500, online_eval_reward: 9.000 2023-05-16 11:51:45 - RayLog - INFO: - current update step obtain a better online_eval_reward: 9.000, save the best model! 2023-05-16 11:51:45 - RayLog - INFO: - Worker 1 finished episode 38 with reward 10.0 in 10 steps 2023-05-16 11:51:45 - RayLog - INFO: - Worker 0 finished episode 37 with reward 13.0 in 13 steps 2023-05-16 11:51:45 - RayLog - INFO: - Worker 0 finished episode 40 with reward 9.0 in 9 steps 2023-05-16 11:51:45 - RayLog - INFO: - Worker 1 finished episode 39 with reward 14.0 in 14 steps 2023-05-16 11:51:45 - RayLog - INFO: - Worker 0 finished episode 41 with reward 14.0 in 14 steps 2023-05-16 11:51:45 - RayLog - INFO: - Worker 1 finished episode 42 with reward 20.0 in 20 steps 2023-05-16 11:51:45 - RayLog - INFO: - Worker 0 finished episode 43 with reward 15.0 in 15 steps 2023-05-16 11:51:46 - RayLog - INFO: - Worker 1 finished episode 44 with reward 19.0 in 19 steps 2023-05-16 11:51:46 - RayLog - INFO: - Worker 0 finished episode 45 with reward 17.0 in 17 steps 2023-05-16 11:51:46 - RayLog - INFO: - Worker 0 finished episode 47 with reward 12.0 in 12 steps 2023-05-16 11:51:46 - RayLog - INFO: - Worker 1 finished episode 46 with reward 15.0 in 15 steps 2023-05-16 11:51:46 - RayLog - INFO: - Worker 0 finished episode 48 with reward 14.0 in 14 steps 2023-05-16 11:51:46 - RayLog - INFO: - Worker 1 finished episode 49 with reward 16.0 in 16 steps 2023-05-16 11:51:46 - RayLog - INFO: - Worker 0 finished episode 50 with reward 9.0 in 9 steps 2023-05-16 11:51:46 - RayLog - INFO: - Worker 1 finished episode 51 with reward 13.0 in 13 steps 2023-05-16 11:51:46 - RayLog - INFO: - Worker 0 finished episode 52 with reward 10.0 in 10 steps 2023-05-16 11:51:47 - RayLog - INFO: - Worker 1 finished episode 53 with reward 13.0 in 13 steps 2023-05-16 11:51:47 - RayLog - INFO: - Worker 0 finished episode 54 with reward 13.0 in 13 steps 2023-05-16 11:51:47 - RayLog - INFO: - Worker 0 finished episode 56 with reward 14.0 in 14 steps 2023-05-16 11:51:47 - RayLog - INFO: - Worker 1 finished episode 55 with reward 20.0 in 20 steps 2023-05-16 11:51:47 - RayLog - INFO: - Worker 0 finished episode 57 with reward 14.0 in 14 steps 2023-05-16 11:51:47 - RayLog - INFO: - Worker 1 finished episode 58 with reward 16.0 in 16 steps 2023-05-16 11:51:47 - RayLog - INFO: - Worker 0 finished episode 59 with reward 11.0 in 11 steps 2023-05-16 11:51:48 - RayLog - INFO: - Worker 1 finished episode 60 with reward 14.0 in 14 steps 2023-05-16 11:51:48 - RayLog - INFO: - Worker 0 finished episode 61 with reward 9.0 in 9 steps 2023-05-16 11:51:48 - RayLog - INFO: - Worker 0 finished episode 63 with reward 19.0 in 19 steps 2023-05-16 11:51:48 - RayLog - INFO: - Worker 1 finished episode 62 with reward 23.0 in 23 steps 2023-05-16 11:51:48 - RayLog - INFO: - Worker 0 finished episode 64 with reward 18.0 in 18 steps 2023-05-16 11:51:49 - RayLog - INFO: - Worker 1 finished episode 65 with reward 26.0 in 26 steps 2023-05-16 11:51:49 - RayLog - INFO: - update_step: 1000, online_eval_reward: 69.000 2023-05-16 11:51:49 - RayLog - INFO: - current update step obtain a better online_eval_reward: 69.000, save the best model! 2023-05-16 11:51:50 - RayLog - INFO: - Worker 1 finished episode 67 with reward 93.0 in 93 steps 2023-05-16 11:51:51 - RayLog - INFO: - Worker 0 finished episode 66 with reward 127.0 in 127 steps 2023-05-16 11:51:51 - RayLog - INFO: - Worker 1 finished episode 68 with reward 40.0 in 40 steps 2023-05-16 11:51:52 - RayLog - INFO: - Worker 0 finished episode 69 with reward 54.0 in 54 steps 2023-05-16 11:51:52 - RayLog - INFO: - Worker 1 finished episode 70 with reward 48.0 in 48 steps 2023-05-16 11:51:53 - RayLog - INFO: - Worker 0 finished episode 71 with reward 62.0 in 62 steps 2023-05-16 11:51:53 - RayLog - INFO: - Worker 1 finished episode 72 with reward 60.0 in 60 steps 2023-05-16 11:51:54 - RayLog - INFO: - Worker 1 finished episode 74 with reward 35.0 in 35 steps 2023-05-16 11:51:54 - RayLog - INFO: - Worker 0 finished episode 73 with reward 47.0 in 47 steps 2023-05-16 11:51:54 - RayLog - INFO: - update_step: 1500, online_eval_reward: 63.000 2023-05-16 11:51:54 - RayLog - INFO: - Worker 1 finished episode 75 with reward 38.0 in 38 steps 2023-05-16 11:51:54 - RayLog - INFO: - Worker 0 finished episode 76 with reward 46.0 in 46 steps 2023-05-16 11:51:55 - RayLog - INFO: - Worker 1 finished episode 77 with reward 40.0 in 40 steps 2023-05-16 11:51:55 - RayLog - INFO: - Worker 0 finished episode 78 with reward 57.0 in 57 steps 2023-05-16 11:51:56 - RayLog - INFO: - Worker 1 finished episode 79 with reward 38.0 in 38 steps 2023-05-16 11:51:56 - RayLog - INFO: - Worker 1 finished episode 81 with reward 33.0 in 33 steps 2023-05-16 11:51:56 - RayLog - INFO: - Worker 0 finished episode 80 with reward 51.0 in 51 steps 2023-05-16 11:51:57 - RayLog - INFO: - Worker 1 finished episode 82 with reward 44.0 in 44 steps 2023-05-16 11:51:58 - RayLog - INFO: - Worker 0 finished episode 83 with reward 70.0 in 70 steps 2023-05-16 11:51:58 - RayLog - INFO: - Worker 1 finished episode 84 with reward 55.0 in 55 steps 2023-05-16 11:51:58 - RayLog - INFO: - update_step: 2000, online_eval_reward: 82.000 2023-05-16 11:51:58 - RayLog - INFO: - current update step obtain a better online_eval_reward: 82.000, save the best model! 2023-05-16 11:51:59 - RayLog - INFO: - Worker 0 finished episode 85 with reward 66.0 in 66 steps 2023-05-16 11:51:59 - RayLog - INFO: - Worker 1 finished episode 86 with reward 56.0 in 56 steps 2023-05-16 11:52:00 - RayLog - INFO: - Worker 1 finished episode 88 with reward 45.0 in 45 steps 2023-05-16 11:52:00 - RayLog - INFO: - Worker 0 finished episode 87 with reward 68.0 in 68 steps 2023-05-16 11:52:01 - RayLog - INFO: - Worker 1 finished episode 89 with reward 50.0 in 50 steps 2023-05-16 11:52:02 - RayLog - INFO: - Worker 0 finished episode 90 with reward 79.0 in 79 steps 2023-05-16 11:52:02 - RayLog - INFO: - Worker 1 finished episode 91 with reward 57.0 in 57 steps 2023-05-16 11:52:04 - RayLog - INFO: - update_step: 2500, online_eval_reward: 77.000 2023-05-16 11:52:04 - RayLog - INFO: - Worker 1 finished episode 93 with reward 66.0 in 66 steps 2023-05-16 11:52:04 - RayLog - INFO: - Worker 0 finished episode 92 with reward 84.0 in 84 steps 2023-05-16 11:52:05 - RayLog - INFO: - Worker 1 finished episode 94 with reward 56.0 in 56 steps 2023-05-16 11:52:07 - RayLog - INFO: - Worker 0 finished episode 95 with reward 134.0 in 134 steps 2023-05-16 11:52:08 - RayLog - INFO: - Worker 1 finished episode 96 with reward 115.0 in 115 steps 2023-05-16 11:52:10 - RayLog - INFO: - update_step: 3000, online_eval_reward: 200.000 2023-05-16 11:52:10 - RayLog - INFO: - current update step obtain a better online_eval_reward: 200.000, save the best model! 2023-05-16 11:52:12 - RayLog - INFO: - Worker 0 finished episode 97 with reward 200.0 in 200 steps 2023-05-16 11:52:13 - RayLog - INFO: - Worker 1 finished episode 98 with reward 200.0 in 200 steps 2023-05-16 11:52:15 - RayLog - INFO: - update_step: 3500, online_eval_reward: 200.000 2023-05-16 11:52:16 - RayLog - INFO: - Worker 0 finished episode 99 with reward 200.0 in 200 steps 2023-05-16 11:52:17 - RayLog - INFO: - Worker 1 finished episode 100 with reward 200.0 in 200 steps 2023-05-16 11:52:19 - SimpleLog - INFO: - Finish training! total time consumed: 53.70s