george-chen's picture
PPO + RND 策略初始上传
15009bf verified
{
"name": "root",
"gauges": {
"Pyramids.Policy.Entropy.mean": {
"value": 0.3719618320465088,
"min": 0.3411966562271118,
"max": 1.3696941137313843,
"count": 33
},
"Pyramids.Policy.Entropy.sum": {
"value": 11289.78515625,
"min": 10175.8486328125,
"max": 41551.0390625,
"count": 33
},
"Pyramids.Step.mean": {
"value": 989910.0,
"min": 29952.0,
"max": 989910.0,
"count": 33
},
"Pyramids.Step.sum": {
"value": 989910.0,
"min": 29952.0,
"max": 989910.0,
"count": 33
},
"Pyramids.Policy.ExtrinsicValueEstimate.mean": {
"value": 0.557979166507721,
"min": -0.06942194700241089,
"max": 0.6186352968215942,
"count": 33
},
"Pyramids.Policy.ExtrinsicValueEstimate.sum": {
"value": 156.79214477539062,
"min": -16.730690002441406,
"max": 176.9296875,
"count": 33
},
"Pyramids.Policy.RndValueEstimate.mean": {
"value": 0.01715184934437275,
"min": 0.0017983707366511226,
"max": 0.3535535931587219,
"count": 33
},
"Pyramids.Policy.RndValueEstimate.sum": {
"value": 4.819669723510742,
"min": 0.5143340229988098,
"max": 85.20641326904297,
"count": 33
},
"Pyramids.Losses.PolicyLoss.mean": {
"value": 0.07034769644780595,
"min": 0.06546779639646359,
"max": 0.07268598883335167,
"count": 33
},
"Pyramids.Losses.PolicyLoss.sum": {
"value": 0.9848677502692833,
"min": 0.5000903865713198,
"max": 1.068005848811784,
"count": 33
},
"Pyramids.Losses.ValueLoss.mean": {
"value": 0.01634651034830659,
"min": 0.0006953251826898772,
"max": 0.016392956733143605,
"count": 33
},
"Pyramids.Losses.ValueLoss.sum": {
"value": 0.22885114487629227,
"min": 0.009039227374968404,
"max": 0.24589435099715407,
"count": 33
},
"Pyramids.Policy.LearningRate.mean": {
"value": 7.71586171379286e-06,
"min": 7.71586171379286e-06,
"max": 0.00029515063018788575,
"count": 33
},
"Pyramids.Policy.LearningRate.sum": {
"value": 0.00010802206399310005,
"min": 0.00010802206399310005,
"max": 0.0035089730303424,
"count": 33
},
"Pyramids.Policy.Epsilon.mean": {
"value": 0.10257192142857142,
"min": 0.10257192142857142,
"max": 0.19838354285714285,
"count": 33
},
"Pyramids.Policy.Epsilon.sum": {
"value": 1.4360069,
"min": 1.3886848,
"max": 2.569657600000001,
"count": 33
},
"Pyramids.Policy.Beta.mean": {
"value": 0.0002669349507142859,
"min": 0.0002669349507142859,
"max": 0.00983851593142857,
"count": 33
},
"Pyramids.Policy.Beta.sum": {
"value": 0.0037370893100000025,
"min": 0.0037370893100000025,
"max": 0.11698879424000001,
"count": 33
},
"Pyramids.Losses.RNDLoss.mean": {
"value": 0.013100835494697094,
"min": 0.012954888865351677,
"max": 0.552263081073761,
"count": 33
},
"Pyramids.Losses.RNDLoss.sum": {
"value": 0.18341170251369476,
"min": 0.18136844038963318,
"max": 3.8658416271209717,
"count": 33
},
"Pyramids.Environment.EpisodeLength.mean": {
"value": 319.68817204301075,
"min": 288.7708333333333,
"max": 999.0,
"count": 33
},
"Pyramids.Environment.EpisodeLength.sum": {
"value": 29731.0,
"min": 15984.0,
"max": 33727.0,
"count": 33
},
"Pyramids.Environment.CumulativeReward.mean": {
"value": 1.6157720272419274,
"min": -1.0000000521540642,
"max": 1.6487083178168784,
"count": 33
},
"Pyramids.Environment.CumulativeReward.sum": {
"value": 150.26679853349924,
"min": -30.31500169634819,
"max": 158.27599851042032,
"count": 33
},
"Pyramids.Policy.ExtrinsicReward.mean": {
"value": 1.6157720272419274,
"min": -1.0000000521540642,
"max": 1.6487083178168784,
"count": 33
},
"Pyramids.Policy.ExtrinsicReward.sum": {
"value": 150.26679853349924,
"min": -30.31500169634819,
"max": 158.27599851042032,
"count": 33
},
"Pyramids.Policy.RndReward.mean": {
"value": 0.04350975932613183,
"min": 0.04291234110511747,
"max": 10.579602957703173,
"count": 33
},
"Pyramids.Policy.RndReward.sum": {
"value": 4.04640761733026,
"min": 4.04640761733026,
"max": 169.27364732325077,
"count": 33
},
"Pyramids.IsTraining.mean": {
"value": 1.0,
"min": 1.0,
"max": 1.0,
"count": 33
},
"Pyramids.IsTraining.sum": {
"value": 1.0,
"min": 1.0,
"max": 1.0,
"count": 33
}
},
"metadata": {
"timer_format_version": "0.1.0",
"start_time_seconds": "1749106443",
"python_version": "3.10.12 (main, Jul 5 2023, 18:54:27) [GCC 11.2.0]",
"command_line_arguments": "/usr/local/bin/mlagents-learn ./config/ppo/PyramidsRND.yaml --env=./training-envs-executables/linux/Pyramids/Pyramids --run-id=Pyramids Training --no-graphics",
"mlagents_version": "1.2.0.dev0",
"mlagents_envs_version": "1.2.0.dev0",
"communication_protocol_version": "1.5.0",
"pytorch_version": "2.7.1+cu126",
"numpy_version": "1.23.5",
"end_time_seconds": "1749108800"
},
"total": 2357.555942989,
"count": 1,
"self": 0.4793554309994761,
"children": {
"run_training.setup": {
"total": 0.02053319200012993,
"count": 1,
"self": 0.02053319200012993
},
"TrainerController.start_learning": {
"total": 2357.0560543660004,
"count": 1,
"self": 1.542523446983978,
"children": {
"TrainerController._reset_env": {
"total": 2.382902737999757,
"count": 1,
"self": 2.382902737999757
},
"TrainerController.advance": {
"total": 2353.0454447260163,
"count": 64005,
"self": 1.5895731469286147,
"children": {
"env_step": {
"total": 1655.1234488499545,
"count": 64005,
"self": 1495.3045436899179,
"children": {
"SubprocessEnvManager._take_step": {
"total": 158.94342934701854,
"count": 64005,
"self": 4.822890691946668,
"children": {
"TorchPolicy.evaluate": {
"total": 154.12053865507187,
"count": 62571,
"self": 154.12053865507187
}
}
},
"workers": {
"total": 0.8754758130180562,
"count": 64005,
"self": 0.0,
"children": {
"worker_root": {
"total": 2351.6258521539517,
"count": 64005,
"is_parallel": true,
"self": 974.6646188260752,
"children": {
"run_training.setup": {
"total": 0.0,
"count": 0,
"is_parallel": true,
"self": 0.0,
"children": {
"steps_from_proto": {
"total": 0.0019300460003250919,
"count": 1,
"is_parallel": true,
"self": 0.0006351480001285381,
"children": {
"_process_rank_one_or_two_observation": {
"total": 0.0012948980001965538,
"count": 8,
"is_parallel": true,
"self": 0.0012948980001965538
}
}
},
"UnityEnvironment.step": {
"total": 0.048744485000042914,
"count": 1,
"is_parallel": true,
"self": 0.0005470760002026509,
"children": {
"UnityEnvironment._generate_step_input": {
"total": 0.000444459999926039,
"count": 1,
"is_parallel": true,
"self": 0.000444459999926039
},
"communicator.exchange": {
"total": 0.04616155199983041,
"count": 1,
"is_parallel": true,
"self": 0.04616155199983041
},
"steps_from_proto": {
"total": 0.0015913970000838162,
"count": 1,
"is_parallel": true,
"self": 0.0003637900003923278,
"children": {
"_process_rank_one_or_two_observation": {
"total": 0.0012276069996914885,
"count": 8,
"is_parallel": true,
"self": 0.0012276069996914885
}
}
}
}
}
}
},
"UnityEnvironment.step": {
"total": 1376.9612333278765,
"count": 64004,
"is_parallel": true,
"self": 33.146260648868065,
"children": {
"UnityEnvironment._generate_step_input": {
"total": 23.434034766969035,
"count": 64004,
"is_parallel": true,
"self": 23.434034766969035
},
"communicator.exchange": {
"total": 1222.1144927710775,
"count": 64004,
"is_parallel": true,
"self": 1222.1144927710775
},
"steps_from_proto": {
"total": 98.26644514096188,
"count": 64004,
"is_parallel": true,
"self": 20.05921853192922,
"children": {
"_process_rank_one_or_two_observation": {
"total": 78.20722660903266,
"count": 512032,
"is_parallel": true,
"self": 78.20722660903266
}
}
}
}
}
}
}
}
}
}
},
"trainer_advance": {
"total": 696.3324227291332,
"count": 64005,
"self": 2.958680399087825,
"children": {
"process_trajectory": {
"total": 130.90989250404255,
"count": 64005,
"self": 130.6430170740432,
"children": {
"RLTrainer._checkpoint": {
"total": 0.2668754299993452,
"count": 2,
"self": 0.2668754299993452
}
}
},
"_update_policy": {
"total": 562.4638498260028,
"count": 453,
"self": 312.8493838020295,
"children": {
"TorchPPOOptimizer.update": {
"total": 249.61446602397336,
"count": 22791,
"self": 249.61446602397336
}
}
}
}
}
}
},
"trainer_threads": {
"total": 9.69999746303074e-07,
"count": 1,
"self": 9.69999746303074e-07
},
"TrainerController._save_models": {
"total": 0.0851824850005869,
"count": 1,
"self": 0.0014934290002202033,
"children": {
"RLTrainer._checkpoint": {
"total": 0.08368905600036669,
"count": 1,
"self": 0.08368905600036669
}
}
}
}
}
}
}