{
  "created_at": "2026-07-12 16:41:43",
  "python": "3.13.5 | packaged by Anaconda, Inc. | (main, Jun 12 2025, 11:23:37) [Clang 14.0.6 ]",
  "torch": "2.9.0",
  "device": "CPU",
  "seeds": [
    0,
    1,
    2
  ],
  "steps": {
    "ppo": 12000,
    "sac": 12000
  },
  "runs": [
    {
      "seed": 0,
      "clip_eps": 0.05,
      "gae_lambda": 0.95,
      "final_return": 37.05,
      "best_return": 46.85,
      "seconds": 1.7116458415985107,
      "algo": "ppo",
      "experiment": "clip",
      "final_kl": 0.0003583758881256366,
      "final_clip_fraction": 0.011548912851139903
    },
    {
      "seed": 1,
      "clip_eps": 0.05,
      "gae_lambda": 0.95,
      "final_return": 53.1,
      "best_return": 53.1,
      "seconds": 1.811697006225586,
      "algo": "ppo",
      "experiment": "clip",
      "final_kl": 0.000681581175345143,
      "final_clip_fraction": 0.020210597780533135
    },
    {
      "seed": 2,
      "clip_eps": 0.05,
      "gae_lambda": 0.95,
      "final_return": 49.1,
      "best_return": 49.1,
      "seconds": 1.6668939590454102,
      "algo": "ppo",
      "experiment": "clip",
      "final_kl": -0.00016262007849708604,
      "final_clip_fraction": 0.00305706518702209
    },
    {
      "seed": 0,
      "clip_eps": 0.2,
      "gae_lambda": 0.95,
      "final_return": 50.6,
      "best_return": 58.0,
      "seconds": 1.6670432090759277,
      "algo": "ppo",
      "experiment": "clip",
      "final_kl": 0.000821556990598292,
      "final_clip_fraction": 0.0
    },
    {
      "seed": 1,
      "clip_eps": 0.2,
      "gae_lambda": 0.95,
      "final_return": 66.25,
      "best_return": 73.35,
      "seconds": 1.8028490543365479,
      "algo": "ppo",
      "experiment": "clip",
      "final_kl": 0.002914823137594502,
      "final_clip_fraction": 0.0006793478387407959
    },
    {
      "seed": 2,
      "clip_eps": 0.2,
      "gae_lambda": 0.95,
      "final_return": 87.7,
      "best_return": 87.7,
      "seconds": 1.6420800685882568,
      "algo": "ppo",
      "experiment": "clip",
      "final_kl": 0.00034384168544993043,
      "final_clip_fraction": 0.0
    },
    {
      "seed": 0,
      "clip_eps": 0.4,
      "gae_lambda": 0.95,
      "final_return": 64.4,
      "best_return": 70.4,
      "seconds": 1.6631369590759277,
      "algo": "ppo",
      "experiment": "clip",
      "final_kl": 0.0003426978460479291,
      "final_clip_fraction": 0.0
    },
    {
      "seed": 1,
      "clip_eps": 0.4,
      "gae_lambda": 0.95,
      "final_return": 70.9,
      "best_return": 70.9,
      "seconds": 1.7238690853118896,
      "algo": "ppo",
      "experiment": "clip",
      "final_kl": 0.0014689532239097608,
      "final_clip_fraction": 0.0
    },
    {
      "seed": 2,
      "clip_eps": 0.4,
      "gae_lambda": 0.95,
      "final_return": 71.4,
      "best_return": 72.3,
      "seconds": 1.6732900142669678,
      "algo": "ppo",
      "experiment": "clip",
      "final_kl": -0.0002984379137199976,
      "final_clip_fraction": 0.0
    },
    {
      "seed": 0,
      "clip_eps": 0.2,
      "gae_lambda": 0.8,
      "final_return": 70.25,
      "best_return": 73.1,
      "seconds": 1.679368019104004,
      "algo": "ppo",
      "experiment": "gae",
      "final_kl": 0.0003794960575958031,
      "final_clip_fraction": 0.00016983695968519896
    },
    {
      "seed": 1,
      "clip_eps": 0.2,
      "gae_lambda": 0.8,
      "final_return": 81.0,
      "best_return": 81.0,
      "seconds": 1.6972320079803467,
      "algo": "ppo",
      "experiment": "gae",
      "final_kl": 0.0019144980306800424,
      "final_clip_fraction": 0.0028872282127849758
    },
    {
      "seed": 2,
      "clip_eps": 0.2,
      "gae_lambda": 0.8,
      "final_return": 71.0,
      "best_return": 71.0,
      "seconds": 1.6474850177764893,
      "algo": "ppo",
      "experiment": "gae",
      "final_kl": 8.42583991784579e-05,
      "final_clip_fraction": 0.0
    },
    {
      "seed": 0,
      "clip_eps": 0.2,
      "gae_lambda": 0.95,
      "final_return": 50.6,
      "best_return": 58.0,
      "seconds": 1.6178691387176514,
      "algo": "ppo",
      "experiment": "gae",
      "final_kl": 0.000821556990598292,
      "final_clip_fraction": 0.0
    },
    {
      "seed": 1,
      "clip_eps": 0.2,
      "gae_lambda": 0.95,
      "final_return": 66.25,
      "best_return": 73.35,
      "seconds": 1.9415416717529297,
      "algo": "ppo",
      "experiment": "gae",
      "final_kl": 0.002914823137594502,
      "final_clip_fraction": 0.0006793478387407959
    },
    {
      "seed": 2,
      "clip_eps": 0.2,
      "gae_lambda": 0.95,
      "final_return": 87.7,
      "best_return": 87.7,
      "seconds": 2.0326571464538574,
      "algo": "ppo",
      "experiment": "gae",
      "final_kl": 0.00034384168544993043,
      "final_clip_fraction": 0.0
    },
    {
      "seed": 0,
      "clip_eps": 0.2,
      "gae_lambda": 1.0,
      "final_return": 47.35,
      "best_return": 51.45,
      "seconds": 1.8349149227142334,
      "algo": "ppo",
      "experiment": "gae",
      "final_kl": 0.00021542065476987116,
      "final_clip_fraction": 0.0
    },
    {
      "seed": 1,
      "clip_eps": 0.2,
      "gae_lambda": 1.0,
      "final_return": 61.65,
      "best_return": 67.5,
      "seconds": 1.6156480312347412,
      "algo": "ppo",
      "experiment": "gae",
      "final_kl": -5.9970092103830974e-05,
      "final_clip_fraction": 0.0
    },
    {
      "seed": 2,
      "clip_eps": 0.2,
      "gae_lambda": 1.0,
      "final_return": 62.5,
      "best_return": 65.6,
      "seconds": 1.6175601482391357,
      "algo": "ppo",
      "experiment": "gae",
      "final_kl": -6.710091081360559e-05,
      "final_clip_fraction": 0.0
    },
    {
      "seed": 0,
      "auto_alpha": false,
      "final_return": 161.72848333838198,
      "best_return": 161.72848333838198,
      "final_alpha": 0.19999998807907104,
      "seconds": 22.50181484222412,
      "algo": "sac",
      "experiment": "temperature"
    },
    {
      "seed": 1,
      "auto_alpha": false,
      "final_return": 172.86769079742794,
      "best_return": 174.77523361552954,
      "final_alpha": 0.19999998807907104,
      "seconds": 22.412888765335083,
      "algo": "sac",
      "experiment": "temperature"
    },
    {
      "seed": 2,
      "auto_alpha": false,
      "final_return": 165.97768580339797,
      "best_return": 170.45933848663262,
      "final_alpha": 0.19999998807907104,
      "seconds": 22.412549018859863,
      "algo": "sac",
      "experiment": "temperature"
    },
    {
      "seed": 0,
      "auto_alpha": true,
      "final_return": 133.78836109292962,
      "best_return": 133.78836109292962,
      "final_alpha": 0.08820080757141113,
      "seconds": 21.707631826400757,
      "algo": "sac",
      "experiment": "temperature"
    },
    {
      "seed": 1,
      "auto_alpha": true,
      "final_return": 175.84463585599707,
      "best_return": 175.87733497496404,
      "final_alpha": 0.13355103135108948,
      "seconds": 20.852226734161377,
      "algo": "sac",
      "experiment": "temperature"
    },
    {
      "seed": 2,
      "auto_alpha": true,
      "final_return": 152.3544233529322,
      "best_return": 152.3544233529322,
      "final_alpha": 0.1196541115641594,
      "seconds": 20.838349103927612,
      "algo": "sac",
      "experiment": "temperature"
    }
  ],
  "aggregate": {
    "ppo_clip_005": {
      "mean": 46.416666666666664,
      "sd": 8.354689302022749,
      "ci_low": 37.05,
      "ci_high": 53.1
    },
    "ppo_clip_02": {
      "mean": 68.18333333333334,
      "sd": 18.62540827293011,
      "ci_low": 50.6,
      "ci_high": 87.7
    },
    "ppo_clip_04": {
      "mean": 68.9,
      "sd": 3.905124837953327,
      "ci_low": 64.4,
      "ci_high": 71.4
    },
    "sac_fixed": {
      "mean": 166.8579533130693,
      "sd": 5.62153349810828,
      "ci_low": 161.72848333838198,
      "ci_high": 172.86769079742794
    },
    "sac_auto": {
      "mean": 153.9958067672863,
      "sd": 21.07612787898535,
      "ci_low": 133.78836109292962,
      "ci_high": 175.8446358559971
    }
  }
}