{
  "best_metric": 0.13906994462013245,
  "best_model_checkpoint": "limbxy_pose/checkpoint-1780",
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 1780,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2808988764044944,
      "grad_norm": 8.567283167615347e-13,
      "learning_rate": 5e-06,
      "loss": 0.3282,
      "step": 25
    },
    {
      "epoch": 0.5617977528089888,
      "grad_norm": 110866.890625,
      "learning_rate": 1e-05,
      "loss": 0.2853,
      "step": 50
    },
    {
      "epoch": 0.8426966292134831,
      "grad_norm": 1257591.75,
      "learning_rate": 1.5e-05,
      "loss": 0.214,
      "step": 75
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.19327276945114136,
      "eval_rmse": 0.4396279752254486,
      "eval_runtime": 9.4477,
      "eval_samples_per_second": 105.845,
      "eval_steps_per_second": 1.694,
      "step": 89
    },
    {
      "epoch": 1.1235955056179776,
      "grad_norm": 1151241.875,
      "learning_rate": 2e-05,
      "loss": 0.1918,
      "step": 100
    },
    {
      "epoch": 1.404494382022472,
      "grad_norm": 2122279.75,
      "learning_rate": 2.5e-05,
      "loss": 0.1868,
      "step": 125
    },
    {
      "epoch": 1.6853932584269664,
      "grad_norm": 2307076.25,
      "learning_rate": 3e-05,
      "loss": 0.1788,
      "step": 150
    },
    {
      "epoch": 1.9662921348314608,
      "grad_norm": 1163768.75,
      "learning_rate": 3.5e-05,
      "loss": 0.161,
      "step": 175
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.14737975597381592,
      "eval_rmse": 0.3839007019996643,
      "eval_runtime": 9.2409,
      "eval_samples_per_second": 108.215,
      "eval_steps_per_second": 1.731,
      "step": 178
    },
    {
      "epoch": 2.247191011235955,
      "grad_norm": 3384581.5,
      "learning_rate": 4e-05,
      "loss": 0.1542,
      "step": 200
    },
    {
      "epoch": 2.5280898876404496,
      "grad_norm": 5387857.0,
      "learning_rate": 4.5e-05,
      "loss": 0.1643,
      "step": 225
    },
    {
      "epoch": 2.808988764044944,
      "grad_norm": 2503687.0,
      "learning_rate": 5e-05,
      "loss": 0.1634,
      "step": 250
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.15408433973789215,
      "eval_rmse": 0.3925357758998871,
      "eval_runtime": 9.3482,
      "eval_samples_per_second": 106.973,
      "eval_steps_per_second": 1.712,
      "step": 267
    },
    {
      "epoch": 3.0898876404494384,
      "grad_norm": 1989428.125,
      "learning_rate": 4.918300653594771e-05,
      "loss": 0.1626,
      "step": 275
    },
    {
      "epoch": 3.370786516853933,
      "grad_norm": 2154801.5,
      "learning_rate": 4.8366013071895424e-05,
      "loss": 0.1588,
      "step": 300
    },
    {
      "epoch": 3.6516853932584272,
      "grad_norm": 1134807.0,
      "learning_rate": 4.7549019607843135e-05,
      "loss": 0.1532,
      "step": 325
    },
    {
      "epoch": 3.932584269662921,
      "grad_norm": 2527826.5,
      "learning_rate": 4.673202614379085e-05,
      "loss": 0.1547,
      "step": 350
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.1495393067598343,
      "eval_rmse": 0.38670313358306885,
      "eval_runtime": 9.6568,
      "eval_samples_per_second": 103.554,
      "eval_steps_per_second": 1.657,
      "step": 356
    },
    {
      "epoch": 4.213483146067416,
      "grad_norm": 830469.375,
      "learning_rate": 4.5915032679738564e-05,
      "loss": 0.1496,
      "step": 375
    },
    {
      "epoch": 4.49438202247191,
      "grad_norm": 823327.75,
      "learning_rate": 4.5098039215686275e-05,
      "loss": 0.1554,
      "step": 400
    },
    {
      "epoch": 4.775280898876405,
      "grad_norm": 1184849.0,
      "learning_rate": 4.4281045751633986e-05,
      "loss": 0.1674,
      "step": 425
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.1679653525352478,
      "eval_rmse": 0.4098357558250427,
      "eval_runtime": 9.548,
      "eval_samples_per_second": 104.734,
      "eval_steps_per_second": 1.676,
      "step": 445
    },
    {
      "epoch": 5.056179775280899,
      "grad_norm": 169527.875,
      "learning_rate": 4.3464052287581704e-05,
      "loss": 0.1676,
      "step": 450
    },
    {
      "epoch": 5.337078651685394,
      "grad_norm": 1563091.5,
      "learning_rate": 4.2647058823529415e-05,
      "loss": 0.1821,
      "step": 475
    },
    {
      "epoch": 5.617977528089888,
      "grad_norm": 1015313.25,
      "learning_rate": 4.1830065359477126e-05,
      "loss": 0.176,
      "step": 500
    },
    {
      "epoch": 5.898876404494382,
      "grad_norm": 2744912.25,
      "learning_rate": 4.101307189542484e-05,
      "loss": 0.1644,
      "step": 525
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.14729027450084686,
      "eval_rmse": 0.3837841749191284,
      "eval_runtime": 9.7012,
      "eval_samples_per_second": 103.08,
      "eval_steps_per_second": 1.649,
      "step": 534
    },
    {
      "epoch": 6.179775280898877,
      "grad_norm": 116897.2265625,
      "learning_rate": 4.0196078431372555e-05,
      "loss": 0.1588,
      "step": 550
    },
    {
      "epoch": 6.460674157303371,
      "grad_norm": 1905296.75,
      "learning_rate": 3.9379084967320266e-05,
      "loss": 0.1517,
      "step": 575
    },
    {
      "epoch": 6.741573033707866,
      "grad_norm": 2599395.75,
      "learning_rate": 3.8562091503267977e-05,
      "loss": 0.1459,
      "step": 600
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.1404775083065033,
      "eval_rmse": 0.3748033046722412,
      "eval_runtime": 9.6204,
      "eval_samples_per_second": 103.945,
      "eval_steps_per_second": 1.663,
      "step": 623
    },
    {
      "epoch": 7.022471910112359,
      "grad_norm": 2782315.75,
      "learning_rate": 3.774509803921569e-05,
      "loss": 0.1479,
      "step": 625
    },
    {
      "epoch": 7.303370786516854,
      "grad_norm": 1166099.875,
      "learning_rate": 3.6928104575163405e-05,
      "loss": 0.1553,
      "step": 650
    },
    {
      "epoch": 7.584269662921348,
      "grad_norm": 2184319.0,
      "learning_rate": 3.611111111111111e-05,
      "loss": 0.1522,
      "step": 675
    },
    {
      "epoch": 7.865168539325842,
      "grad_norm": 1279161.625,
      "learning_rate": 3.529411764705883e-05,
      "loss": 0.1544,
      "step": 700
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.15137900412082672,
      "eval_rmse": 0.38907453417778015,
      "eval_runtime": 9.7104,
      "eval_samples_per_second": 102.982,
      "eval_steps_per_second": 1.648,
      "step": 712
    },
    {
      "epoch": 8.146067415730338,
      "grad_norm": 2559078.25,
      "learning_rate": 3.447712418300654e-05,
      "loss": 0.1521,
      "step": 725
    },
    {
      "epoch": 8.426966292134832,
      "grad_norm": 257444.171875,
      "learning_rate": 3.366013071895425e-05,
      "loss": 0.143,
      "step": 750
    },
    {
      "epoch": 8.707865168539326,
      "grad_norm": 927010.125,
      "learning_rate": 3.284313725490196e-05,
      "loss": 0.1477,
      "step": 775
    },
    {
      "epoch": 8.98876404494382,
      "grad_norm": 159431.078125,
      "learning_rate": 3.202614379084967e-05,
      "loss": 0.1531,
      "step": 800
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.14193063974380493,
      "eval_rmse": 0.3767368197441101,
      "eval_runtime": 9.5944,
      "eval_samples_per_second": 104.227,
      "eval_steps_per_second": 1.668,
      "step": 801
    },
    {
      "epoch": 9.269662921348315,
      "grad_norm": 1847863.875,
      "learning_rate": 3.120915032679739e-05,
      "loss": 0.1485,
      "step": 825
    },
    {
      "epoch": 9.55056179775281,
      "grad_norm": 952230.1875,
      "learning_rate": 3.0392156862745097e-05,
      "loss": 0.154,
      "step": 850
    },
    {
      "epoch": 9.831460674157304,
      "grad_norm": 88050.4765625,
      "learning_rate": 2.957516339869281e-05,
      "loss": 0.1468,
      "step": 875
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.14216481149196625,
      "eval_rmse": 0.37704750895500183,
      "eval_runtime": 9.601,
      "eval_samples_per_second": 104.156,
      "eval_steps_per_second": 1.666,
      "step": 890
    },
    {
      "epoch": 10.112359550561798,
      "grad_norm": 837362.1875,
      "learning_rate": 2.8758169934640522e-05,
      "loss": 0.1504,
      "step": 900
    },
    {
      "epoch": 10.393258426966293,
      "grad_norm": 1949830.125,
      "learning_rate": 2.7941176470588236e-05,
      "loss": 0.1456,
      "step": 925
    },
    {
      "epoch": 10.674157303370787,
      "grad_norm": 3423746.0,
      "learning_rate": 2.7124183006535947e-05,
      "loss": 0.1445,
      "step": 950
    },
    {
      "epoch": 10.955056179775282,
      "grad_norm": 267509.53125,
      "learning_rate": 2.630718954248366e-05,
      "loss": 0.143,
      "step": 975
    },
    {
      "epoch": 11.0,
      "eval_loss": 0.14469632506370544,
      "eval_rmse": 0.3803896903991699,
      "eval_runtime": 9.5391,
      "eval_samples_per_second": 104.832,
      "eval_steps_per_second": 1.677,
      "step": 979
    },
    {
      "epoch": 11.235955056179776,
      "grad_norm": 1893457.375,
      "learning_rate": 2.5490196078431373e-05,
      "loss": 0.14,
      "step": 1000
    },
    {
      "epoch": 11.51685393258427,
      "grad_norm": 1374771.875,
      "learning_rate": 2.4673202614379087e-05,
      "loss": 0.1458,
      "step": 1025
    },
    {
      "epoch": 11.797752808988765,
      "grad_norm": 2660207.0,
      "learning_rate": 2.38562091503268e-05,
      "loss": 0.1467,
      "step": 1050
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.1596774458885193,
      "eval_rmse": 0.39959660172462463,
      "eval_runtime": 9.752,
      "eval_samples_per_second": 102.543,
      "eval_steps_per_second": 1.641,
      "step": 1068
    },
    {
      "epoch": 12.07865168539326,
      "grad_norm": 4057213.75,
      "learning_rate": 2.303921568627451e-05,
      "loss": 0.1556,
      "step": 1075
    },
    {
      "epoch": 12.359550561797754,
      "grad_norm": 551497.625,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.1514,
      "step": 1100
    },
    {
      "epoch": 12.640449438202246,
      "grad_norm": 61427.015625,
      "learning_rate": 2.1405228758169934e-05,
      "loss": 0.1446,
      "step": 1125
    },
    {
      "epoch": 12.921348314606742,
      "grad_norm": 789958.0,
      "learning_rate": 2.058823529411765e-05,
      "loss": 0.149,
      "step": 1150
    },
    {
      "epoch": 13.0,
      "eval_loss": 0.1393509805202484,
      "eval_rmse": 0.37329748272895813,
      "eval_runtime": 9.3006,
      "eval_samples_per_second": 107.52,
      "eval_steps_per_second": 1.72,
      "step": 1157
    },
    {
      "epoch": 13.202247191011235,
      "grad_norm": 1956022.25,
      "learning_rate": 1.977124183006536e-05,
      "loss": 0.1439,
      "step": 1175
    },
    {
      "epoch": 13.48314606741573,
      "grad_norm": 1201509.0,
      "learning_rate": 1.895424836601307e-05,
      "loss": 0.1423,
      "step": 1200
    },
    {
      "epoch": 13.764044943820224,
      "grad_norm": 1603021.5,
      "learning_rate": 1.8137254901960785e-05,
      "loss": 0.139,
      "step": 1225
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.14120306074619293,
      "eval_rmse": 0.3757699429988861,
      "eval_runtime": 9.6552,
      "eval_samples_per_second": 103.571,
      "eval_steps_per_second": 1.657,
      "step": 1246
    },
    {
      "epoch": 14.044943820224718,
      "grad_norm": 750973.0625,
      "learning_rate": 1.7320261437908496e-05,
      "loss": 0.1442,
      "step": 1250
    },
    {
      "epoch": 14.325842696629213,
      "grad_norm": 380421.21875,
      "learning_rate": 1.650326797385621e-05,
      "loss": 0.1413,
      "step": 1275
    },
    {
      "epoch": 14.606741573033707,
      "grad_norm": 451599.53125,
      "learning_rate": 1.568627450980392e-05,
      "loss": 0.141,
      "step": 1300
    },
    {
      "epoch": 14.887640449438202,
      "grad_norm": 1830802.625,
      "learning_rate": 1.4869281045751634e-05,
      "loss": 0.1433,
      "step": 1325
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.13972964882850647,
      "eval_rmse": 0.3738043010234833,
      "eval_runtime": 9.3491,
      "eval_samples_per_second": 106.962,
      "eval_steps_per_second": 1.711,
      "step": 1335
    },
    {
      "epoch": 15.168539325842696,
      "grad_norm": 241513.265625,
      "learning_rate": 1.4052287581699347e-05,
      "loss": 0.142,
      "step": 1350
    },
    {
      "epoch": 15.44943820224719,
      "grad_norm": 1214894.5,
      "learning_rate": 1.323529411764706e-05,
      "loss": 0.145,
      "step": 1375
    },
    {
      "epoch": 15.730337078651685,
      "grad_norm": 692089.5625,
      "learning_rate": 1.2418300653594772e-05,
      "loss": 0.1426,
      "step": 1400
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.13960076868534088,
      "eval_rmse": 0.37363186478614807,
      "eval_runtime": 9.7085,
      "eval_samples_per_second": 103.003,
      "eval_steps_per_second": 1.648,
      "step": 1424
    },
    {
      "epoch": 16.01123595505618,
      "grad_norm": 21341.8984375,
      "learning_rate": 1.1601307189542485e-05,
      "loss": 0.1419,
      "step": 1425
    },
    {
      "epoch": 16.292134831460675,
      "grad_norm": 361124.03125,
      "learning_rate": 1.0784313725490197e-05,
      "loss": 0.1409,
      "step": 1450
    },
    {
      "epoch": 16.573033707865168,
      "grad_norm": 563769.3125,
      "learning_rate": 9.96732026143791e-06,
      "loss": 0.1427,
      "step": 1475
    },
    {
      "epoch": 16.853932584269664,
      "grad_norm": 67665.75,
      "learning_rate": 9.150326797385621e-06,
      "loss": 0.1403,
      "step": 1500
    },
    {
      "epoch": 17.0,
      "eval_loss": 0.13963405787944794,
      "eval_rmse": 0.3736764192581177,
      "eval_runtime": 9.5629,
      "eval_samples_per_second": 104.571,
      "eval_steps_per_second": 1.673,
      "step": 1513
    },
    {
      "epoch": 17.134831460674157,
      "grad_norm": 661210.5625,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.1424,
      "step": 1525
    },
    {
      "epoch": 17.415730337078653,
      "grad_norm": 708130.875,
      "learning_rate": 7.5163398692810456e-06,
      "loss": 0.1424,
      "step": 1550
    },
    {
      "epoch": 17.696629213483146,
      "grad_norm": 93833.1875,
      "learning_rate": 6.699346405228758e-06,
      "loss": 0.139,
      "step": 1575
    },
    {
      "epoch": 17.97752808988764,
      "grad_norm": 154237.1875,
      "learning_rate": 5.882352941176471e-06,
      "loss": 0.1393,
      "step": 1600
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.139328271150589,
      "eval_rmse": 0.3732670545578003,
      "eval_runtime": 9.4037,
      "eval_samples_per_second": 106.341,
      "eval_steps_per_second": 1.701,
      "step": 1602
    },
    {
      "epoch": 18.258426966292134,
      "grad_norm": 451971.6875,
      "learning_rate": 5.065359477124184e-06,
      "loss": 0.1388,
      "step": 1625
    },
    {
      "epoch": 18.53932584269663,
      "grad_norm": 516973.15625,
      "learning_rate": 4.2483660130718954e-06,
      "loss": 0.1383,
      "step": 1650
    },
    {
      "epoch": 18.820224719101123,
      "grad_norm": 111973.78125,
      "learning_rate": 3.431372549019608e-06,
      "loss": 0.1426,
      "step": 1675
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.13913311064243317,
      "eval_rmse": 0.3730055093765259,
      "eval_runtime": 9.78,
      "eval_samples_per_second": 102.25,
      "eval_steps_per_second": 1.636,
      "step": 1691
    },
    {
      "epoch": 19.10112359550562,
      "grad_norm": 390392.875,
      "learning_rate": 2.6143790849673204e-06,
      "loss": 0.1404,
      "step": 1700
    },
    {
      "epoch": 19.382022471910112,
      "grad_norm": 167872.421875,
      "learning_rate": 1.7973856209150326e-06,
      "loss": 0.1402,
      "step": 1725
    },
    {
      "epoch": 19.662921348314608,
      "grad_norm": 446064.03125,
      "learning_rate": 9.80392156862745e-07,
      "loss": 0.14,
      "step": 1750
    },
    {
      "epoch": 19.9438202247191,
      "grad_norm": 133050.875,
      "learning_rate": 1.6339869281045752e-07,
      "loss": 0.1387,
      "step": 1775
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.13906994462013245,
      "eval_rmse": 0.37292084097862244,
      "eval_runtime": 9.6305,
      "eval_samples_per_second": 103.837,
      "eval_steps_per_second": 1.661,
      "step": 1780
    },
    {
      "epoch": 20.0,
      "step": 1780,
      "total_flos": 0.0,
      "train_loss": 0.15599652267573924,
      "train_runtime": 3327.495,
      "train_samples_per_second": 34.032,
      "train_steps_per_second": 0.535
    }
  ],
  "logging_steps": 25,
  "max_steps": 1780,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}