AkkyMa's picture
mlagent trained on pyramids. 3M steps
58e7aeb
{
"name": "root",
"gauges": {
"Pyramids.Policy.Entropy.mean": {
"value": 0.15857157111167908,
"min": 0.15857157111167908,
"max": 1.543981671333313,
"count": 100
},
"Pyramids.Policy.Entropy.sum": {
"value": 4764.7587890625,
"min": 4764.7587890625,
"max": 46838.2265625,
"count": 100
},
"Pyramids.Step.mean": {
"value": 2999950.0,
"min": 29952.0,
"max": 2999950.0,
"count": 100
},
"Pyramids.Step.sum": {
"value": 2999950.0,
"min": 29952.0,
"max": 2999950.0,
"count": 100
},
"Pyramids.Policy.ExtrinsicValueEstimate.mean": {
"value": 0.588160514831543,
"min": -0.1437324434518814,
"max": 0.8028568625450134,
"count": 100
},
"Pyramids.Policy.ExtrinsicValueEstimate.sum": {
"value": 164.6849365234375,
"min": -34.06459045410156,
"max": 240.8570556640625,
"count": 100
},
"Pyramids.Policy.RndValueEstimate.mean": {
"value": 0.0068190377205610275,
"min": -0.027372222393751144,
"max": 0.13060776889324188,
"count": 100
},
"Pyramids.Policy.RndValueEstimate.sum": {
"value": 1.9093306064605713,
"min": -7.363127708435059,
"max": 31.345863342285156,
"count": 100
},
"Pyramids.Losses.PolicyLoss.mean": {
"value": 0.0643930189516617,
"min": 0.06432597154413261,
"max": 0.07492538075636263,
"count": 100
},
"Pyramids.Losses.PolicyLoss.sum": {
"value": 0.9015022653232638,
"min": 0.48371602765695276,
"max": 1.1238807113454394,
"count": 100
},
"Pyramids.Losses.ValueLoss.mean": {
"value": 0.01468287803886813,
"min": 3.470613956909611e-05,
"max": 0.017370028032261568,
"count": 100
},
"Pyramids.Losses.ValueLoss.sum": {
"value": 0.20556029254415384,
"min": 0.0004858859539673455,
"max": 0.2605504204839235,
"count": 100
},
"Pyramids.Policy.LearningRate.mean": {
"value": 1.5420637717261931e-06,
"min": 1.5420637717261931e-06,
"max": 0.00029838354339596195,
"count": 100
},
"Pyramids.Policy.LearningRate.sum": {
"value": 2.1588892804166704e-05,
"min": 2.1588892804166704e-05,
"max": 0.0038015834328055664,
"count": 100
},
"Pyramids.Policy.Epsilon.mean": {
"value": 0.10051398809523811,
"min": 0.10051398809523811,
"max": 0.19946118095238097,
"count": 100
},
"Pyramids.Policy.Epsilon.sum": {
"value": 1.4071958333333334,
"min": 1.3897045333333333,
"max": 2.667194433333333,
"count": 100
},
"Pyramids.Policy.Beta.mean": {
"value": 6.13474107142858e-05,
"min": 6.13474107142858e-05,
"max": 0.009946171977142856,
"count": 100
},
"Pyramids.Policy.Beta.sum": {
"value": 0.0008588637500000012,
"min": 0.0008588637500000012,
"max": 0.12673272389,
"count": 100
},
"Pyramids.Losses.RNDLoss.mean": {
"value": 0.003701410023495555,
"min": 0.003517433535307646,
"max": 0.2633723020553589,
"count": 100
},
"Pyramids.Losses.RNDLoss.sum": {
"value": 0.05181974172592163,
"min": 0.04924406856298447,
"max": 1.8436061143875122,
"count": 100
},
"Pyramids.Environment.EpisodeLength.mean": {
"value": 312.74444444444447,
"min": 236.55905511811022,
"max": 999.0,
"count": 100
},
"Pyramids.Environment.EpisodeLength.sum": {
"value": 28147.0,
"min": 15984.0,
"max": 32948.0,
"count": 100
},
"Pyramids.Environment.CumulativeReward.mean": {
"value": 1.5316488696469202,
"min": -1.0000000521540642,
"max": 1.7487222082203342,
"count": 100
},
"Pyramids.Environment.CumulativeReward.sum": {
"value": 137.8483982682228,
"min": -32.000001668930054,
"max": 221.7859979942441,
"count": 100
},
"Pyramids.Policy.ExtrinsicReward.mean": {
"value": 1.5316488696469202,
"min": -1.0000000521540642,
"max": 1.7487222082203342,
"count": 100
},
"Pyramids.Policy.ExtrinsicReward.sum": {
"value": 137.8483982682228,
"min": -32.000001668930054,
"max": 221.7859979942441,
"count": 100
},
"Pyramids.Policy.RndReward.mean": {
"value": 0.012229293376973753,
"min": 0.009215955869501432,
"max": 5.251085902564228,
"count": 100
},
"Pyramids.Policy.RndReward.sum": {
"value": 1.1006364039276377,
"min": 1.0676441566174617,
"max": 84.01737444102764,
"count": 100
},
"Pyramids.IsTraining.mean": {
"value": 1.0,
"min": 1.0,
"max": 1.0,
"count": 100
},
"Pyramids.IsTraining.sum": {
"value": 1.0,
"min": 1.0,
"max": 1.0,
"count": 100
}
},
"metadata": {
"timer_format_version": "0.1.0",
"start_time_seconds": "1673196212",
"python_version": "3.8.16 (default, Dec 7 2022, 01:12:13) \n[GCC 7.5.0]",
"command_line_arguments": "/usr/local/bin/mlagents-learn /content/ml-agents/config/ppo/PyramidsRND.yaml --env=/content/ml-agents/trained-envs-executables/linux/Pyramids/Pyramids --run-id=Test training --no-graphics --force",
"mlagents_version": "0.29.0.dev0",
"mlagents_envs_version": "0.29.0.dev0",
"communication_protocol_version": "1.5.0",
"pytorch_version": "1.8.1+cu102",
"numpy_version": "1.21.6",
"end_time_seconds": "1673202835"
},
"total": 6622.441328919999,
"count": 1,
"self": 0.4363996339989171,
"children": {
"run_training.setup": {
"total": 0.10465832500005945,
"count": 1,
"self": 0.10465832500005945
},
"TrainerController.start_learning": {
"total": 6621.900270961,
"count": 1,
"self": 3.788250254980994,
"children": {
"TrainerController._reset_env": {
"total": 8.052275997999914,
"count": 1,
"self": 8.052275997999914
},
"TrainerController.advance": {
"total": 6609.960688211019,
"count": 193754,
"self": 3.9513188350401833,
"children": {
"env_step": {
"total": 4646.2246730619445,
"count": 193754,
"self": 4334.178061711059,
"children": {
"SubprocessEnvManager._take_step": {
"total": 309.71383146188305,
"count": 193754,
"self": 12.832201786917608,
"children": {
"TorchPolicy.evaluate": {
"total": 296.88162967496544,
"count": 187565,
"self": 100.81691595295933,
"children": {
"TorchPolicy.sample_actions": {
"total": 196.0647137220061,
"count": 187565,
"self": 196.0647137220061
}
}
}
}
},
"workers": {
"total": 2.332779889002609,
"count": 193754,
"self": 0.0,
"children": {
"worker_root": {
"total": 6610.578882897,
"count": 193754,
"is_parallel": true,
"self": 2574.2281525200533,
"children": {
"run_training.setup": {
"total": 0.0,
"count": 0,
"is_parallel": true,
"self": 0.0,
"children": {
"steps_from_proto": {
"total": 0.005880848000060723,
"count": 1,
"is_parallel": true,
"self": 0.0038315440001497336,
"children": {
"_process_rank_one_or_two_observation": {
"total": 0.0020493039999109897,
"count": 8,
"is_parallel": true,
"self": 0.0020493039999109897
}
}
},
"UnityEnvironment.step": {
"total": 0.046209895999936634,
"count": 1,
"is_parallel": true,
"self": 0.00048119099983523483,
"children": {
"UnityEnvironment._generate_step_input": {
"total": 0.0004219070000317515,
"count": 1,
"is_parallel": true,
"self": 0.0004219070000317515
},
"communicator.exchange": {
"total": 0.043660666000050696,
"count": 1,
"is_parallel": true,
"self": 0.043660666000050696
},
"steps_from_proto": {
"total": 0.0016461320000189517,
"count": 1,
"is_parallel": true,
"self": 0.0004197490001160986,
"children": {
"_process_rank_one_or_two_observation": {
"total": 0.001226382999902853,
"count": 8,
"is_parallel": true,
"self": 0.001226382999902853
}
}
}
}
}
}
},
"UnityEnvironment.step": {
"total": 4036.3507303769466,
"count": 193753,
"is_parallel": true,
"self": 83.69451287680431,
"children": {
"UnityEnvironment._generate_step_input": {
"total": 69.05359577403499,
"count": 193753,
"is_parallel": true,
"self": 69.05359577403499
},
"communicator.exchange": {
"total": 3598.0965021771544,
"count": 193753,
"is_parallel": true,
"self": 3598.0965021771544
},
"steps_from_proto": {
"total": 285.50611954895294,
"count": 193753,
"is_parallel": true,
"self": 67.861109807866,
"children": {
"_process_rank_one_or_two_observation": {
"total": 217.64500974108694,
"count": 1550024,
"is_parallel": true,
"self": 217.64500974108694
}
}
}
}
}
}
}
}
}
}
},
"trainer_advance": {
"total": 1959.7846963140346,
"count": 193754,
"self": 7.090088829112119,
"children": {
"process_trajectory": {
"total": 446.3787026329247,
"count": 193754,
"self": 445.80569031992434,
"children": {
"RLTrainer._checkpoint": {
"total": 0.5730123130003903,
"count": 6,
"self": 0.5730123130003903
}
}
},
"_update_policy": {
"total": 1506.3159048519976,
"count": 1384,
"self": 573.5196004159192,
"children": {
"TorchPPOOptimizer.update": {
"total": 932.7963044360785,
"count": 68382,
"self": 932.7963044360785
}
}
}
}
}
}
},
"trainer_threads": {
"total": 9.030000001075678e-07,
"count": 1,
"self": 9.030000001075678e-07
},
"TrainerController._save_models": {
"total": 0.09905559399976482,
"count": 1,
"self": 0.0013711919991692412,
"children": {
"RLTrainer._checkpoint": {
"total": 0.09768440200059558,
"count": 1,
"self": 0.09768440200059558
}
}
}
}
}
}
}