poca-SoccerTwos / run_logs /timers.json
nlamot's picture
5M steps, increased learning rate with linear scheduler
5c58095
raw
history blame
15.7 kB
{
"name": "root",
"gauges": {
"SoccerTwos.Policy.Entropy.mean": {
"value": 1.755637288093567,
"min": 1.7151814699172974,
"max": 3.295724630355835,
"count": 499
},
"SoccerTwos.Policy.Entropy.sum": {
"value": 36910.51953125,
"min": 13612.6572265625,
"max": 172905.75,
"count": 499
},
"SoccerTwos.Environment.EpisodeLength.mean": {
"value": 63.6125,
"min": 39.645161290322584,
"max": 999.0,
"count": 499
},
"SoccerTwos.Environment.EpisodeLength.sum": {
"value": 20356.0,
"min": 10692.0,
"max": 31500.0,
"count": 499
},
"SoccerTwos.Self-play.ELO.mean": {
"value": 1540.6404213531655,
"min": 1197.4957713356807,
"max": 1576.4143309371482,
"count": 478
},
"SoccerTwos.Self-play.ELO.sum": {
"value": 246502.4674165065,
"min": 2399.421175774324,
"max": 363079.2013600959,
"count": 478
},
"SoccerTwos.Step.mean": {
"value": 4999972.0,
"min": 9254.0,
"max": 4999972.0,
"count": 500
},
"SoccerTwos.Step.sum": {
"value": 4999972.0,
"min": 9254.0,
"max": 4999972.0,
"count": 500
},
"SoccerTwos.Policy.ExtrinsicBaselineEstimate.mean": {
"value": -0.034482039511203766,
"min": -0.10746785998344421,
"max": 0.16136549413204193,
"count": 500
},
"SoccerTwos.Policy.ExtrinsicBaselineEstimate.sum": {
"value": -5.517126083374023,
"min": -22.675718307495117,
"max": 27.432022094726562,
"count": 500
},
"SoccerTwos.Policy.ExtrinsicValueEstimate.mean": {
"value": -0.034504301846027374,
"min": -0.11580751091241837,
"max": 0.1590351015329361,
"count": 500
},
"SoccerTwos.Policy.ExtrinsicValueEstimate.sum": {
"value": -5.520688533782959,
"min": -24.43538475036621,
"max": 26.952106475830078,
"count": 500
},
"SoccerTwos.Environment.CumulativeReward.mean": {
"value": 0.0,
"min": 0.0,
"max": 0.0,
"count": 500
},
"SoccerTwos.Environment.CumulativeReward.sum": {
"value": 0.0,
"min": 0.0,
"max": 0.0,
"count": 500
},
"SoccerTwos.Policy.ExtrinsicReward.mean": {
"value": -0.08047500103712082,
"min": -0.5921090895479376,
"max": 0.4576799988746643,
"count": 500
},
"SoccerTwos.Policy.ExtrinsicReward.sum": {
"value": -12.876000165939331,
"min": -66.35360014438629,
"max": 60.00919944047928,
"count": 500
},
"SoccerTwos.Environment.GroupCumulativeReward.mean": {
"value": -0.08047500103712082,
"min": -0.5921090895479376,
"max": 0.4576799988746643,
"count": 500
},
"SoccerTwos.Environment.GroupCumulativeReward.sum": {
"value": -12.876000165939331,
"min": -66.35360014438629,
"max": 60.00919944047928,
"count": 500
},
"SoccerTwos.IsTraining.mean": {
"value": 1.0,
"min": 1.0,
"max": 1.0,
"count": 500
},
"SoccerTwos.IsTraining.sum": {
"value": 1.0,
"min": 1.0,
"max": 1.0,
"count": 500
},
"SoccerTwos.Losses.PolicyLoss.mean": {
"value": 0.015034021192695946,
"min": 0.010941779592152064,
"max": 0.024873137117053072,
"count": 239
},
"SoccerTwos.Losses.PolicyLoss.sum": {
"value": 0.015034021192695946,
"min": 0.010941779592152064,
"max": 0.024873137117053072,
"count": 239
},
"SoccerTwos.Losses.ValueLoss.mean": {
"value": 0.11739198019107183,
"min": 3.1021232795562052e-06,
"max": 0.1251884085436662,
"count": 239
},
"SoccerTwos.Losses.ValueLoss.sum": {
"value": 0.11739198019107183,
"min": 3.1021232795562052e-06,
"max": 0.1251884085436662,
"count": 239
},
"SoccerTwos.Losses.BaselineLoss.mean": {
"value": 0.11775108724832535,
"min": 3.304618659664508e-06,
"max": 0.1258082126577695,
"count": 239
},
"SoccerTwos.Losses.BaselineLoss.sum": {
"value": 0.11775108724832535,
"min": 3.304618659664508e-06,
"max": 0.1258082126577695,
"count": 239
},
"SoccerTwos.Policy.LearningRate.mean": {
"value": 1.0132997973599917e-06,
"min": 1.0132997973599917e-06,
"max": 0.0004968980006203998,
"count": 239
},
"SoccerTwos.Policy.LearningRate.sum": {
"value": 1.0132997973599917e-06,
"min": 1.0132997973599917e-06,
"max": 0.0004968980006203998,
"count": 239
},
"SoccerTwos.Policy.Epsilon.mean": {
"value": 0.10020264000000001,
"min": 0.10020264000000001,
"max": 0.19937960000000005,
"count": 239
},
"SoccerTwos.Policy.Epsilon.sum": {
"value": 0.10020264000000001,
"min": 0.10020264000000001,
"max": 0.19937960000000005,
"count": 239
},
"SoccerTwos.Policy.Beta.mean": {
"value": 2.0111735999999927e-05,
"min": 2.0111735999999927e-05,
"max": 0.004969042040000001,
"count": 239
},
"SoccerTwos.Policy.Beta.sum": {
"value": 2.0111735999999927e-05,
"min": 2.0111735999999927e-05,
"max": 0.004969042040000001,
"count": 239
}
},
"metadata": {
"timer_format_version": "0.1.0",
"start_time_seconds": "1703848764",
"python_version": "3.10.12 (main, Jul 5 2023, 15:02:25) [Clang 14.0.6 ]",
"command_line_arguments": "/opt/homebrew/Caskroom/miniconda/base/envs/rl/bin/mlagents-learn ./config/poca/SoccerTwos.yaml --env=./training-envs-executables/SoccerTwos.app --run-id=SoccerTwos --no-graphics --force",
"mlagents_version": "1.1.0.dev0",
"mlagents_envs_version": "1.1.0.dev0",
"communication_protocol_version": "1.5.0",
"pytorch_version": "2.1.2",
"numpy_version": "1.23.5",
"end_time_seconds": "1703853889"
},
"total": 5124.430119083001,
"count": 1,
"self": 0.19001916600973345,
"children": {
"run_training.setup": {
"total": 0.012270416977116838,
"count": 1,
"self": 0.012270416977116838
},
"TrainerController.start_learning": {
"total": 5124.227829500014,
"count": 1,
"self": 2.531355981802335,
"children": {
"TrainerController._reset_env": {
"total": 1.5887895020714495,
"count": 20,
"self": 1.5887895020714495
},
"TrainerController.advance": {
"total": 5119.955040642148,
"count": 341985,
"self": 2.185286336200079,
"children": {
"env_step": {
"total": 2303.891035468987,
"count": 341985,
"self": 1941.7491068352829,
"children": {
"SubprocessEnvManager._take_step": {
"total": 360.6655651805049,
"count": 341985,
"self": 11.17105897743022,
"children": {
"TorchPolicy.evaluate": {
"total": 349.4945062030747,
"count": 631296,
"self": 349.4945062030747
}
}
},
"workers": {
"total": 1.4763634531991556,
"count": 341985,
"self": 0.0,
"children": {
"worker_root": {
"total": 5119.23122701954,
"count": 341985,
"is_parallel": true,
"self": 3492.6540978672565,
"children": {
"steps_from_proto": {
"total": 0.013060167053481564,
"count": 40,
"is_parallel": true,
"self": 0.0021915972174610943,
"children": {
"_process_rank_one_or_two_observation": {
"total": 0.01086856983602047,
"count": 160,
"is_parallel": true,
"self": 0.01086856983602047
}
}
},
"UnityEnvironment.step": {
"total": 1626.5640689852298,
"count": 341985,
"is_parallel": true,
"self": 78.99318619386759,
"children": {
"UnityEnvironment._generate_step_input": {
"total": 37.23996753204847,
"count": 341985,
"is_parallel": true,
"self": 37.23996753204847
},
"communicator.exchange": {
"total": 1310.1512493426271,
"count": 341985,
"is_parallel": true,
"self": 1310.1512493426271
},
"steps_from_proto": {
"total": 200.17966591668664,
"count": 683970,
"is_parallel": true,
"self": 30.950994182756403,
"children": {
"_process_rank_one_or_two_observation": {
"total": 169.22867173393024,
"count": 2735880,
"is_parallel": true,
"self": 169.22867173393024
}
}
}
}
}
}
}
}
}
}
},
"trainer_advance": {
"total": 2813.8787188369606,
"count": 341985,
"self": 21.16630858287681,
"children": {
"process_trajectory": {
"total": 556.1312195862702,
"count": 341985,
"self": 554.5936779602489,
"children": {
"RLTrainer._checkpoint": {
"total": 1.537541626021266,
"count": 10,
"self": 1.537541626021266
}
}
},
"_update_policy": {
"total": 2236.5811906678136,
"count": 239,
"self": 242.41541127837263,
"children": {
"TorchPOCAOptimizer.update": {
"total": 1994.165779389441,
"count": 7185,
"self": 1994.165779389441
}
}
}
}
}
}
},
"trainer_threads": {
"total": 3.3300602808594704e-07,
"count": 1,
"self": 3.3300602808594704e-07
},
"TrainerController._save_models": {
"total": 0.15264304098673165,
"count": 1,
"self": 0.000656957970932126,
"children": {
"RLTrainer._checkpoint": {
"total": 0.15198608301579952,
"count": 1,
"self": 0.15198608301579952
}
}
}
}
}
}
}