Unity 机器学习代理工具包 (ML-Agents) 是一个开源项目,它使游戏和模拟能够作为训练智能代理的环境。
您最多选择25个主题 主题必须以中文或者字母或数字开头,可以包含连字符 (-),并且长度不得超过35个字符
 
 
 
 
 

298 行
16 KiB

{
"name": "root",
"gauges": {
"WalkerStatic.Policy.Entropy.mean": {
"value": 1.2709734439849854,
"min": 1.2709734439849854,
"max": 1.4189627170562744,
"count": 666
},
"WalkerStatic.Environment.EpisodeLength.mean": {
"value": 848.2285714285714,
"min": 24.081035923141187,
"max": 999.0,
"count": 666
},
"WalkerStatic.Policy.ExtrinsicValueEstimate.mean": {
"value": 380.5071716308594,
"min": 0.03258597478270531,
"max": 400.4476318359375,
"count": 666
},
"WalkerStatic.Environment.CumulativeReward.mean": {
"value": 1696.214483823095,
"min": 1.9118849212758517,
"max": 1987.2431925455728,
"count": 666
},
"WalkerStatic.Policy.ExtrinsicReward.mean": {
"value": 1696.214483823095,
"min": 1.9118849212758517,
"max": 1987.2431925455728,
"count": 666
},
"WalkerStatic.Losses.ValueLoss.mean": {
"value": 103.36185455322266,
"min": 0.6496382355690002,
"max": 600.4512939453125,
"count": 666
},
"WalkerStatic.Losses.PolicyLoss.mean": {
"value": 0.01793554611504078,
"min": 0.011666166596114635,
"max": 0.045658644288778305,
"count": 666
},
"WalkerStatic.Policy.LearningRate.mean": {
"value": 6.031856969457294e-07,
"min": 6.031856969457294e-07,
"max": 0.00029969253228046,
"count": 666
},
"WalkerStatic.Policy.Epsilon.mean": {
"value": 0.10020104795694351,
"min": 0.10020104795694351,
"max": 0.19989757239818573,
"count": 666
},
"WalkerStatic.Policy.Beta.mean": {
"value": 2.0031329768244177e-05,
"min": 2.0031329768244177e-05,
"max": 0.0049948859959840775,
"count": 666
},
"WalkerStatic.IsTraining.mean": {
"value": 1.0,
"min": 1.0,
"max": 1.0,
"count": 666
}
},
"metadata": {
"timer_format_version": "0.1.0",
"start_time_seconds": "1592941558",
"python_version": "3.6.9 (default, Apr 18 2020, 01:56:04) \n[GCC 8.4.0]",
"command_line_arguments": "/usr/local/bin/mlagents-learn /unity-volume/config/ppo/WalkerStatic.yaml --no-graphics --env=/unity-volume/executable --run-id=wst-ppo --train --num-envs=8",
"mlagents_version": "0.18.0.dev0",
"mlagents_envs_version": "0.18.0.dev0",
"communication_protocol_version": "1.0.0",
"tensorflow_version": "2.2.0",
"end_time_seconds": "1592960254"
},
"total": 18696.346647177008,
"count": 1,
"self": 3.3935399079928175,
"children": {
"run_training.setup": {
"total": 0.04523291898658499,
"count": 1,
"self": 0.04523291898658499
},
"TrainerController.start_learning": {
"total": 18692.90787435003,
"count": 1,
"self": 7724.8386514367885,
"children": {
"TrainerController._reset_env": {
"total": 5.98037285002647,
"count": 1,
"self": 5.98037285002647
},
"TrainerController.advance": {
"total": 10961.930494949222,
"count": 641043,
"self": 16.358328600239474,
"children": {
"env_step": {
"total": 10945.572166348982,
"count": 641043,
"self": 5405.399530877301,
"children": {
"SubprocessEnvManager._take_step": {
"total": 5513.383768304659,
"count": 2067785,
"self": 123.3252387821558,
"children": {
"NNPolicy.evaluate": {
"total": 5390.058529522503,
"count": 2004163,
"self": 5390.058529522503
}
}
},
"workers": {
"total": 26.788867167022545,
"count": 641043,
"self": 0.0,
"children": {
"worker_root": {
"total": 149372.61837113614,
"count": 2067780,
"is_parallel": true,
"self": 71730.36220776226,
"children": {
"run_training.setup": {
"total": 0.0,
"count": 0,
"is_parallel": true,
"self": 0.0,
"children": {
"steps_from_proto": {
"total": 0.009601829922758043,
"count": 8,
"is_parallel": true,
"self": 0.002117355994414538,
"children": {
"_process_vector_observation": {
"total": 0.007484473928343505,
"count": 16,
"is_parallel": true,
"self": 0.007484473928343505
}
}
},
"UnityEnvironment.step": {
"total": 0.4091312560485676,
"count": 8,
"is_parallel": true,
"self": 0.002098764176480472,
"children": {
"UnityEnvironment._generate_step_input": {
"total": 0.0020347919780761003,
"count": 8,
"is_parallel": true,
"self": 0.0020347919780761003
},
"communicator.exchange": {
"total": 0.39732585789170116,
"count": 8,
"is_parallel": true,
"self": 0.39732585789170116
},
"steps_from_proto": {
"total": 0.007671842002309859,
"count": 8,
"is_parallel": true,
"self": 0.001171778014395386,
"children": {
"_process_vector_observation": {
"total": 0.006500063987914473,
"count": 16,
"is_parallel": true,
"self": 0.006500063987914473
}
}
}
}
}
}
},
"UnityEnvironment.step": {
"total": 77642.25616337388,
"count": 2067772,
"is_parallel": true,
"self": 682.1374540438992,
"children": {
"UnityEnvironment._generate_step_input": {
"total": 666.6127658021287,
"count": 2067772,
"is_parallel": true,
"self": 666.6127658021287
},
"communicator.exchange": {
"total": 73484.35254396428,
"count": 2067772,
"is_parallel": true,
"self": 73484.35254396428
},
"steps_from_proto": {
"total": 2809.153399563569,
"count": 2067772,
"is_parallel": true,
"self": 364.1548632780905,
"children": {
"_process_vector_observation": {
"total": 2444.9985362854786,
"count": 4135544,
"is_parallel": true,
"self": 2444.9985362854786
}
}
}
}
}
}
}
}
}
}
}
}
},
"trainer_threads": {
"total": 5.791999865323305e-05,
"count": 1,
"self": 5.791999865323305e-05,
"children": {
"thread_root": {
"total": 0.0,
"count": 0,
"is_parallel": true,
"self": 0.0,
"children": {
"trainer_advance": {
"total": 18341.90605110576,
"count": 31192350,
"is_parallel": true,
"self": 788.756272311788,
"children": {
"process_trajectory": {
"total": 12639.446193398966,
"count": 31192350,
"is_parallel": true,
"self": 12633.074971054913,
"children": {
"Trainer.save_model": {
"total": 6.371222344052512,
"count": 40,
"is_parallel": true,
"self": 6.371222344052512
}
}
},
"_update_policy": {
"total": 4913.703585395007,
"count": 955,
"is_parallel": true,
"self": 2077.309252145933,
"children": {
"PPOOptimizer.update": {
"total": 2836.3943332490744,
"count": 28650,
"is_parallel": true,
"self": 2836.3943332490744
}
}
}
}
}
}
}
}
},
"TrainerController._save_model": {
"total": 0.15829719399334863,
"count": 1,
"self": 0.0023166430182754993,
"children": {
"Trainer.save_model": {
"total": 0.15598055097507313,
"count": 1,
"self": 0.15598055097507313
}
}
}
}
}
}
}