{
  "best_metric": 0.8263623714447021,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.4434589800443459,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0022172949002217295,
      "grad_norm": 43.49946975708008,
      "learning_rate": 1.004e-05,
      "loss": 2.8648,
      "step": 1
    },
    {
      "epoch": 0.0022172949002217295,
      "eval_loss": 2.7855024337768555,
      "eval_runtime": 20.2353,
      "eval_samples_per_second": 9.39,
      "eval_steps_per_second": 2.372,
      "step": 1
    },
    {
      "epoch": 0.004434589800443459,
      "grad_norm": 49.54320526123047,
      "learning_rate": 2.008e-05,
      "loss": 3.479,
      "step": 2
    },
    {
      "epoch": 0.0066518847006651885,
      "grad_norm": 43.55087661743164,
      "learning_rate": 3.012e-05,
      "loss": 3.174,
      "step": 3
    },
    {
      "epoch": 0.008869179600886918,
      "grad_norm": 34.47139358520508,
      "learning_rate": 4.016e-05,
      "loss": 2.9292,
      "step": 4
    },
    {
      "epoch": 0.011086474501108648,
      "grad_norm": 33.695804595947266,
      "learning_rate": 5.02e-05,
      "loss": 2.9573,
      "step": 5
    },
    {
      "epoch": 0.013303769401330377,
      "grad_norm": 35.9599609375,
      "learning_rate": 6.024e-05,
      "loss": 3.0289,
      "step": 6
    },
    {
      "epoch": 0.015521064301552107,
      "grad_norm": 30.871747970581055,
      "learning_rate": 7.028e-05,
      "loss": 2.7663,
      "step": 7
    },
    {
      "epoch": 0.017738359201773836,
      "grad_norm": 21.825246810913086,
      "learning_rate": 8.032e-05,
      "loss": 2.2641,
      "step": 8
    },
    {
      "epoch": 0.019955654101995565,
      "grad_norm": 26.116636276245117,
      "learning_rate": 9.036000000000001e-05,
      "loss": 2.573,
      "step": 9
    },
    {
      "epoch": 0.022172949002217297,
      "grad_norm": 22.22761344909668,
      "learning_rate": 0.0001004,
      "loss": 2.4794,
      "step": 10
    },
    {
      "epoch": 0.024390243902439025,
      "grad_norm": 26.01641082763672,
      "learning_rate": 9.987157894736842e-05,
      "loss": 2.447,
      "step": 11
    },
    {
      "epoch": 0.026607538802660754,
      "grad_norm": 21.385372161865234,
      "learning_rate": 9.934315789473684e-05,
      "loss": 2.0634,
      "step": 12
    },
    {
      "epoch": 0.028824833702882482,
      "grad_norm": 34.55962371826172,
      "learning_rate": 9.881473684210525e-05,
      "loss": 2.431,
      "step": 13
    },
    {
      "epoch": 0.031042128603104215,
      "grad_norm": 29.898178100585938,
      "learning_rate": 9.828631578947369e-05,
      "loss": 2.6574,
      "step": 14
    },
    {
      "epoch": 0.03325942350332594,
      "grad_norm": 27.43971824645996,
      "learning_rate": 9.77578947368421e-05,
      "loss": 1.9155,
      "step": 15
    },
    {
      "epoch": 0.03547671840354767,
      "grad_norm": 27.087106704711914,
      "learning_rate": 9.722947368421052e-05,
      "loss": 2.1793,
      "step": 16
    },
    {
      "epoch": 0.037694013303769404,
      "grad_norm": 30.93043327331543,
      "learning_rate": 9.670105263157895e-05,
      "loss": 2.4952,
      "step": 17
    },
    {
      "epoch": 0.03991130820399113,
      "grad_norm": 27.66971206665039,
      "learning_rate": 9.617263157894737e-05,
      "loss": 2.2107,
      "step": 18
    },
    {
      "epoch": 0.04212860310421286,
      "grad_norm": 37.52210235595703,
      "learning_rate": 9.564421052631579e-05,
      "loss": 2.235,
      "step": 19
    },
    {
      "epoch": 0.04434589800443459,
      "grad_norm": 23.63651466369629,
      "learning_rate": 9.511578947368421e-05,
      "loss": 2.1324,
      "step": 20
    },
    {
      "epoch": 0.04656319290465632,
      "grad_norm": 24.720478057861328,
      "learning_rate": 9.458736842105264e-05,
      "loss": 2.5987,
      "step": 21
    },
    {
      "epoch": 0.04878048780487805,
      "grad_norm": 23.012195587158203,
      "learning_rate": 9.405894736842106e-05,
      "loss": 2.3515,
      "step": 22
    },
    {
      "epoch": 0.050997782705099776,
      "grad_norm": 24.576961517333984,
      "learning_rate": 9.353052631578947e-05,
      "loss": 2.2289,
      "step": 23
    },
    {
      "epoch": 0.05321507760532151,
      "grad_norm": 24.782127380371094,
      "learning_rate": 9.300210526315789e-05,
      "loss": 2.2446,
      "step": 24
    },
    {
      "epoch": 0.05543237250554324,
      "grad_norm": 29.085662841796875,
      "learning_rate": 9.247368421052631e-05,
      "loss": 2.4364,
      "step": 25
    },
    {
      "epoch": 0.057649667405764965,
      "grad_norm": 26.606853485107422,
      "learning_rate": 9.194526315789473e-05,
      "loss": 2.2234,
      "step": 26
    },
    {
      "epoch": 0.0598669623059867,
      "grad_norm": 30.282766342163086,
      "learning_rate": 9.141684210526316e-05,
      "loss": 2.5798,
      "step": 27
    },
    {
      "epoch": 0.06208425720620843,
      "grad_norm": 29.041223526000977,
      "learning_rate": 9.088842105263158e-05,
      "loss": 2.4436,
      "step": 28
    },
    {
      "epoch": 0.06430155210643015,
      "grad_norm": 32.80604553222656,
      "learning_rate": 9.036000000000001e-05,
      "loss": 2.8015,
      "step": 29
    },
    {
      "epoch": 0.06651884700665188,
      "grad_norm": 28.190860748291016,
      "learning_rate": 8.983157894736843e-05,
      "loss": 2.359,
      "step": 30
    },
    {
      "epoch": 0.06873614190687362,
      "grad_norm": 35.894405364990234,
      "learning_rate": 8.930315789473684e-05,
      "loss": 1.9933,
      "step": 31
    },
    {
      "epoch": 0.07095343680709534,
      "grad_norm": 27.286930084228516,
      "learning_rate": 8.877473684210526e-05,
      "loss": 2.4784,
      "step": 32
    },
    {
      "epoch": 0.07317073170731707,
      "grad_norm": 31.852373123168945,
      "learning_rate": 8.824631578947368e-05,
      "loss": 2.2698,
      "step": 33
    },
    {
      "epoch": 0.07538802660753881,
      "grad_norm": 30.357845306396484,
      "learning_rate": 8.771789473684211e-05,
      "loss": 2.2334,
      "step": 34
    },
    {
      "epoch": 0.07760532150776053,
      "grad_norm": 34.58502197265625,
      "learning_rate": 8.718947368421053e-05,
      "loss": 2.6314,
      "step": 35
    },
    {
      "epoch": 0.07982261640798226,
      "grad_norm": 29.154176712036133,
      "learning_rate": 8.666105263157895e-05,
      "loss": 2.1524,
      "step": 36
    },
    {
      "epoch": 0.082039911308204,
      "grad_norm": 28.38983917236328,
      "learning_rate": 8.613263157894737e-05,
      "loss": 1.7586,
      "step": 37
    },
    {
      "epoch": 0.08425720620842572,
      "grad_norm": 22.648962020874023,
      "learning_rate": 8.560421052631578e-05,
      "loss": 1.4079,
      "step": 38
    },
    {
      "epoch": 0.08647450110864745,
      "grad_norm": 43.15308380126953,
      "learning_rate": 8.50757894736842e-05,
      "loss": 2.1336,
      "step": 39
    },
    {
      "epoch": 0.08869179600886919,
      "grad_norm": 41.409969329833984,
      "learning_rate": 8.454736842105263e-05,
      "loss": 2.5629,
      "step": 40
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 32.42283248901367,
      "learning_rate": 8.401894736842106e-05,
      "loss": 2.0355,
      "step": 41
    },
    {
      "epoch": 0.09312638580931264,
      "grad_norm": 64.17134094238281,
      "learning_rate": 8.349052631578948e-05,
      "loss": 2.8994,
      "step": 42
    },
    {
      "epoch": 0.09534368070953436,
      "grad_norm": 51.92913055419922,
      "learning_rate": 8.29621052631579e-05,
      "loss": 1.9425,
      "step": 43
    },
    {
      "epoch": 0.0975609756097561,
      "grad_norm": 60.37343215942383,
      "learning_rate": 8.243368421052632e-05,
      "loss": 2.5374,
      "step": 44
    },
    {
      "epoch": 0.09977827050997783,
      "grad_norm": 56.61979675292969,
      "learning_rate": 8.190526315789474e-05,
      "loss": 2.7696,
      "step": 45
    },
    {
      "epoch": 0.10199556541019955,
      "grad_norm": 45.54286575317383,
      "learning_rate": 8.137684210526315e-05,
      "loss": 2.2621,
      "step": 46
    },
    {
      "epoch": 0.10421286031042129,
      "grad_norm": 48.23731231689453,
      "learning_rate": 8.084842105263157e-05,
      "loss": 3.2684,
      "step": 47
    },
    {
      "epoch": 0.10643015521064302,
      "grad_norm": 52.1467399597168,
      "learning_rate": 8.032e-05,
      "loss": 3.0954,
      "step": 48
    },
    {
      "epoch": 0.10864745011086474,
      "grad_norm": 57.875450134277344,
      "learning_rate": 7.979157894736842e-05,
      "loss": 2.946,
      "step": 49
    },
    {
      "epoch": 0.11086474501108648,
      "grad_norm": 88.02739715576172,
      "learning_rate": 7.926315789473684e-05,
      "loss": 4.3883,
      "step": 50
    },
    {
      "epoch": 0.11086474501108648,
      "eval_loss": 1.270225167274475,
      "eval_runtime": 20.2194,
      "eval_samples_per_second": 9.397,
      "eval_steps_per_second": 2.374,
      "step": 50
    },
    {
      "epoch": 0.1130820399113082,
      "grad_norm": 26.53325653076172,
      "learning_rate": 7.873473684210526e-05,
      "loss": 3.0032,
      "step": 51
    },
    {
      "epoch": 0.11529933481152993,
      "grad_norm": 17.52619743347168,
      "learning_rate": 7.820631578947369e-05,
      "loss": 2.7827,
      "step": 52
    },
    {
      "epoch": 0.11751662971175167,
      "grad_norm": 14.779261589050293,
      "learning_rate": 7.76778947368421e-05,
      "loss": 2.2784,
      "step": 53
    },
    {
      "epoch": 0.1197339246119734,
      "grad_norm": 12.997217178344727,
      "learning_rate": 7.714947368421052e-05,
      "loss": 2.2414,
      "step": 54
    },
    {
      "epoch": 0.12195121951219512,
      "grad_norm": 14.368338584899902,
      "learning_rate": 7.662105263157896e-05,
      "loss": 2.2074,
      "step": 55
    },
    {
      "epoch": 0.12416851441241686,
      "grad_norm": 15.52922534942627,
      "learning_rate": 7.609263157894737e-05,
      "loss": 2.1736,
      "step": 56
    },
    {
      "epoch": 0.12638580931263857,
      "grad_norm": 18.17425537109375,
      "learning_rate": 7.556421052631579e-05,
      "loss": 2.3588,
      "step": 57
    },
    {
      "epoch": 0.1286031042128603,
      "grad_norm": 16.153467178344727,
      "learning_rate": 7.503578947368421e-05,
      "loss": 2.4205,
      "step": 58
    },
    {
      "epoch": 0.13082039911308205,
      "grad_norm": 15.452371597290039,
      "learning_rate": 7.450736842105263e-05,
      "loss": 2.2599,
      "step": 59
    },
    {
      "epoch": 0.13303769401330376,
      "grad_norm": 18.882034301757812,
      "learning_rate": 7.397894736842105e-05,
      "loss": 1.9239,
      "step": 60
    },
    {
      "epoch": 0.1352549889135255,
      "grad_norm": 15.408071517944336,
      "learning_rate": 7.345052631578948e-05,
      "loss": 2.0407,
      "step": 61
    },
    {
      "epoch": 0.13747228381374724,
      "grad_norm": 18.57291603088379,
      "learning_rate": 7.29221052631579e-05,
      "loss": 2.1355,
      "step": 62
    },
    {
      "epoch": 0.13968957871396895,
      "grad_norm": 14.49044418334961,
      "learning_rate": 7.239368421052631e-05,
      "loss": 1.6433,
      "step": 63
    },
    {
      "epoch": 0.1419068736141907,
      "grad_norm": 21.082347869873047,
      "learning_rate": 7.186526315789474e-05,
      "loss": 2.5801,
      "step": 64
    },
    {
      "epoch": 0.14412416851441243,
      "grad_norm": 18.55680274963379,
      "learning_rate": 7.133684210526316e-05,
      "loss": 1.7756,
      "step": 65
    },
    {
      "epoch": 0.14634146341463414,
      "grad_norm": 17.898927688598633,
      "learning_rate": 7.080842105263158e-05,
      "loss": 1.6243,
      "step": 66
    },
    {
      "epoch": 0.14855875831485588,
      "grad_norm": 17.05344581604004,
      "learning_rate": 7.028e-05,
      "loss": 1.7034,
      "step": 67
    },
    {
      "epoch": 0.15077605321507762,
      "grad_norm": 19.20818328857422,
      "learning_rate": 6.975157894736843e-05,
      "loss": 1.8613,
      "step": 68
    },
    {
      "epoch": 0.15299334811529933,
      "grad_norm": 14.604522705078125,
      "learning_rate": 6.922315789473685e-05,
      "loss": 1.3704,
      "step": 69
    },
    {
      "epoch": 0.15521064301552107,
      "grad_norm": 18.43328285217285,
      "learning_rate": 6.869473684210527e-05,
      "loss": 1.9582,
      "step": 70
    },
    {
      "epoch": 0.1574279379157428,
      "grad_norm": 19.990154266357422,
      "learning_rate": 6.816631578947368e-05,
      "loss": 1.7685,
      "step": 71
    },
    {
      "epoch": 0.15964523281596452,
      "grad_norm": 17.83521842956543,
      "learning_rate": 6.76378947368421e-05,
      "loss": 2.1192,
      "step": 72
    },
    {
      "epoch": 0.16186252771618626,
      "grad_norm": 21.896644592285156,
      "learning_rate": 6.710947368421052e-05,
      "loss": 1.6801,
      "step": 73
    },
    {
      "epoch": 0.164079822616408,
      "grad_norm": 17.964820861816406,
      "learning_rate": 6.658105263157894e-05,
      "loss": 1.7591,
      "step": 74
    },
    {
      "epoch": 0.1662971175166297,
      "grad_norm": 22.256559371948242,
      "learning_rate": 6.605263157894737e-05,
      "loss": 2.5407,
      "step": 75
    },
    {
      "epoch": 0.16851441241685144,
      "grad_norm": 19.16011619567871,
      "learning_rate": 6.55242105263158e-05,
      "loss": 1.7553,
      "step": 76
    },
    {
      "epoch": 0.17073170731707318,
      "grad_norm": 24.629257202148438,
      "learning_rate": 6.499578947368422e-05,
      "loss": 2.3935,
      "step": 77
    },
    {
      "epoch": 0.1729490022172949,
      "grad_norm": 19.385765075683594,
      "learning_rate": 6.446736842105264e-05,
      "loss": 1.8258,
      "step": 78
    },
    {
      "epoch": 0.17516629711751663,
      "grad_norm": 20.863401412963867,
      "learning_rate": 6.393894736842105e-05,
      "loss": 1.9962,
      "step": 79
    },
    {
      "epoch": 0.17738359201773837,
      "grad_norm": 24.57994842529297,
      "learning_rate": 6.341052631578947e-05,
      "loss": 2.5787,
      "step": 80
    },
    {
      "epoch": 0.17960088691796008,
      "grad_norm": 23.363492965698242,
      "learning_rate": 6.288210526315789e-05,
      "loss": 1.577,
      "step": 81
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 22.762670516967773,
      "learning_rate": 6.235368421052632e-05,
      "loss": 2.2144,
      "step": 82
    },
    {
      "epoch": 0.18403547671840353,
      "grad_norm": 19.128597259521484,
      "learning_rate": 6.182526315789474e-05,
      "loss": 1.6379,
      "step": 83
    },
    {
      "epoch": 0.18625277161862527,
      "grad_norm": 20.36613655090332,
      "learning_rate": 6.129684210526316e-05,
      "loss": 2.1041,
      "step": 84
    },
    {
      "epoch": 0.188470066518847,
      "grad_norm": 30.103256225585938,
      "learning_rate": 6.076842105263158e-05,
      "loss": 1.8926,
      "step": 85
    },
    {
      "epoch": 0.19068736141906872,
      "grad_norm": 25.986953735351562,
      "learning_rate": 6.024e-05,
      "loss": 1.9419,
      "step": 86
    },
    {
      "epoch": 0.19290465631929046,
      "grad_norm": 29.100027084350586,
      "learning_rate": 5.971157894736842e-05,
      "loss": 2.3349,
      "step": 87
    },
    {
      "epoch": 0.1951219512195122,
      "grad_norm": 27.938316345214844,
      "learning_rate": 5.9183157894736835e-05,
      "loss": 1.8313,
      "step": 88
    },
    {
      "epoch": 0.1973392461197339,
      "grad_norm": 36.821903228759766,
      "learning_rate": 5.8654736842105267e-05,
      "loss": 2.5949,
      "step": 89
    },
    {
      "epoch": 0.19955654101995565,
      "grad_norm": 28.251705169677734,
      "learning_rate": 5.8126315789473684e-05,
      "loss": 2.038,
      "step": 90
    },
    {
      "epoch": 0.2017738359201774,
      "grad_norm": 20.87628936767578,
      "learning_rate": 5.759789473684211e-05,
      "loss": 1.4964,
      "step": 91
    },
    {
      "epoch": 0.2039911308203991,
      "grad_norm": 36.82621383666992,
      "learning_rate": 5.706947368421053e-05,
      "loss": 2.1786,
      "step": 92
    },
    {
      "epoch": 0.20620842572062084,
      "grad_norm": 34.26662826538086,
      "learning_rate": 5.6541052631578945e-05,
      "loss": 2.3302,
      "step": 93
    },
    {
      "epoch": 0.20842572062084258,
      "grad_norm": 29.61810874938965,
      "learning_rate": 5.601263157894736e-05,
      "loss": 2.0587,
      "step": 94
    },
    {
      "epoch": 0.2106430155210643,
      "grad_norm": 31.78813934326172,
      "learning_rate": 5.5484210526315794e-05,
      "loss": 2.5443,
      "step": 95
    },
    {
      "epoch": 0.21286031042128603,
      "grad_norm": 37.09660339355469,
      "learning_rate": 5.495578947368421e-05,
      "loss": 1.9665,
      "step": 96
    },
    {
      "epoch": 0.21507760532150777,
      "grad_norm": 24.321231842041016,
      "learning_rate": 5.442736842105264e-05,
      "loss": 2.1167,
      "step": 97
    },
    {
      "epoch": 0.21729490022172948,
      "grad_norm": 28.0638484954834,
      "learning_rate": 5.3898947368421055e-05,
      "loss": 1.2507,
      "step": 98
    },
    {
      "epoch": 0.21951219512195122,
      "grad_norm": 43.129005432128906,
      "learning_rate": 5.337052631578947e-05,
      "loss": 3.1423,
      "step": 99
    },
    {
      "epoch": 0.22172949002217296,
      "grad_norm": 90.48893737792969,
      "learning_rate": 5.284210526315789e-05,
      "loss": 2.6743,
      "step": 100
    },
    {
      "epoch": 0.22172949002217296,
      "eval_loss": 1.0482261180877686,
      "eval_runtime": 20.243,
      "eval_samples_per_second": 9.386,
      "eval_steps_per_second": 2.371,
      "step": 100
    },
    {
      "epoch": 0.22394678492239467,
      "grad_norm": 17.120187759399414,
      "learning_rate": 5.231368421052631e-05,
      "loss": 2.3941,
      "step": 101
    },
    {
      "epoch": 0.2261640798226164,
      "grad_norm": 16.493423461914062,
      "learning_rate": 5.178526315789474e-05,
      "loss": 2.3262,
      "step": 102
    },
    {
      "epoch": 0.22838137472283815,
      "grad_norm": 12.841573715209961,
      "learning_rate": 5.1256842105263165e-05,
      "loss": 2.0664,
      "step": 103
    },
    {
      "epoch": 0.23059866962305986,
      "grad_norm": 12.463582992553711,
      "learning_rate": 5.072842105263158e-05,
      "loss": 2.0229,
      "step": 104
    },
    {
      "epoch": 0.2328159645232816,
      "grad_norm": 13.76339054107666,
      "learning_rate": 5.02e-05,
      "loss": 2.0176,
      "step": 105
    },
    {
      "epoch": 0.23503325942350334,
      "grad_norm": 12.230353355407715,
      "learning_rate": 4.967157894736842e-05,
      "loss": 1.6626,
      "step": 106
    },
    {
      "epoch": 0.23725055432372505,
      "grad_norm": 14.192876815795898,
      "learning_rate": 4.914315789473684e-05,
      "loss": 2.0265,
      "step": 107
    },
    {
      "epoch": 0.2394678492239468,
      "grad_norm": 13.952037811279297,
      "learning_rate": 4.861473684210526e-05,
      "loss": 1.7241,
      "step": 108
    },
    {
      "epoch": 0.24168514412416853,
      "grad_norm": 14.577288627624512,
      "learning_rate": 4.8086315789473686e-05,
      "loss": 1.7365,
      "step": 109
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 19.91164779663086,
      "learning_rate": 4.7557894736842104e-05,
      "loss": 1.8842,
      "step": 110
    },
    {
      "epoch": 0.24611973392461198,
      "grad_norm": 14.9652681350708,
      "learning_rate": 4.702947368421053e-05,
      "loss": 1.7975,
      "step": 111
    },
    {
      "epoch": 0.24833702882483372,
      "grad_norm": 13.924230575561523,
      "learning_rate": 4.6501052631578946e-05,
      "loss": 1.9894,
      "step": 112
    },
    {
      "epoch": 0.25055432372505543,
      "grad_norm": 13.089156150817871,
      "learning_rate": 4.5972631578947364e-05,
      "loss": 1.758,
      "step": 113
    },
    {
      "epoch": 0.25277161862527714,
      "grad_norm": 12.60175895690918,
      "learning_rate": 4.544421052631579e-05,
      "loss": 1.6666,
      "step": 114
    },
    {
      "epoch": 0.2549889135254989,
      "grad_norm": 17.490257263183594,
      "learning_rate": 4.4915789473684213e-05,
      "loss": 1.9925,
      "step": 115
    },
    {
      "epoch": 0.2572062084257206,
      "grad_norm": 12.952994346618652,
      "learning_rate": 4.438736842105263e-05,
      "loss": 1.6062,
      "step": 116
    },
    {
      "epoch": 0.25942350332594233,
      "grad_norm": 15.763453483581543,
      "learning_rate": 4.3858947368421056e-05,
      "loss": 2.0694,
      "step": 117
    },
    {
      "epoch": 0.2616407982261641,
      "grad_norm": 14.717920303344727,
      "learning_rate": 4.3330526315789474e-05,
      "loss": 1.7913,
      "step": 118
    },
    {
      "epoch": 0.2638580931263858,
      "grad_norm": 16.53207778930664,
      "learning_rate": 4.280210526315789e-05,
      "loss": 1.7959,
      "step": 119
    },
    {
      "epoch": 0.2660753880266075,
      "grad_norm": 13.253047943115234,
      "learning_rate": 4.2273684210526317e-05,
      "loss": 1.6143,
      "step": 120
    },
    {
      "epoch": 0.2682926829268293,
      "grad_norm": 15.786408424377441,
      "learning_rate": 4.174526315789474e-05,
      "loss": 1.6721,
      "step": 121
    },
    {
      "epoch": 0.270509977827051,
      "grad_norm": 12.185956001281738,
      "learning_rate": 4.121684210526316e-05,
      "loss": 1.5245,
      "step": 122
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 21.057693481445312,
      "learning_rate": 4.068842105263158e-05,
      "loss": 2.1359,
      "step": 123
    },
    {
      "epoch": 0.2749445676274945,
      "grad_norm": 13.455188751220703,
      "learning_rate": 4.016e-05,
      "loss": 2.0307,
      "step": 124
    },
    {
      "epoch": 0.2771618625277162,
      "grad_norm": 15.558540344238281,
      "learning_rate": 3.963157894736842e-05,
      "loss": 1.7351,
      "step": 125
    },
    {
      "epoch": 0.2793791574279379,
      "grad_norm": 16.883926391601562,
      "learning_rate": 3.9103157894736844e-05,
      "loss": 2.1596,
      "step": 126
    },
    {
      "epoch": 0.28159645232815966,
      "grad_norm": 20.178613662719727,
      "learning_rate": 3.857473684210526e-05,
      "loss": 1.8029,
      "step": 127
    },
    {
      "epoch": 0.2838137472283814,
      "grad_norm": 18.442106246948242,
      "learning_rate": 3.804631578947369e-05,
      "loss": 2.0837,
      "step": 128
    },
    {
      "epoch": 0.2860310421286031,
      "grad_norm": 19.457265853881836,
      "learning_rate": 3.7517894736842105e-05,
      "loss": 1.6284,
      "step": 129
    },
    {
      "epoch": 0.28824833702882485,
      "grad_norm": 26.571035385131836,
      "learning_rate": 3.698947368421052e-05,
      "loss": 1.7662,
      "step": 130
    },
    {
      "epoch": 0.29046563192904656,
      "grad_norm": 19.368440628051758,
      "learning_rate": 3.646105263157895e-05,
      "loss": 1.8661,
      "step": 131
    },
    {
      "epoch": 0.2926829268292683,
      "grad_norm": 22.966577529907227,
      "learning_rate": 3.593263157894737e-05,
      "loss": 2.619,
      "step": 132
    },
    {
      "epoch": 0.29490022172949004,
      "grad_norm": 24.9805965423584,
      "learning_rate": 3.540421052631579e-05,
      "loss": 1.5081,
      "step": 133
    },
    {
      "epoch": 0.29711751662971175,
      "grad_norm": 19.5284366607666,
      "learning_rate": 3.4875789473684215e-05,
      "loss": 1.7447,
      "step": 134
    },
    {
      "epoch": 0.29933481152993346,
      "grad_norm": 21.54673194885254,
      "learning_rate": 3.434736842105263e-05,
      "loss": 1.6988,
      "step": 135
    },
    {
      "epoch": 0.30155210643015523,
      "grad_norm": 23.316730499267578,
      "learning_rate": 3.381894736842105e-05,
      "loss": 2.2736,
      "step": 136
    },
    {
      "epoch": 0.30376940133037694,
      "grad_norm": 22.004526138305664,
      "learning_rate": 3.329052631578947e-05,
      "loss": 1.9549,
      "step": 137
    },
    {
      "epoch": 0.30598669623059865,
      "grad_norm": 23.052682876586914,
      "learning_rate": 3.27621052631579e-05,
      "loss": 2.0963,
      "step": 138
    },
    {
      "epoch": 0.3082039911308204,
      "grad_norm": 20.599315643310547,
      "learning_rate": 3.223368421052632e-05,
      "loss": 1.2949,
      "step": 139
    },
    {
      "epoch": 0.31042128603104213,
      "grad_norm": 20.93349266052246,
      "learning_rate": 3.1705263157894736e-05,
      "loss": 1.9277,
      "step": 140
    },
    {
      "epoch": 0.31263858093126384,
      "grad_norm": 27.307830810546875,
      "learning_rate": 3.117684210526316e-05,
      "loss": 1.122,
      "step": 141
    },
    {
      "epoch": 0.3148558758314856,
      "grad_norm": 17.325700759887695,
      "learning_rate": 3.064842105263158e-05,
      "loss": 1.2689,
      "step": 142
    },
    {
      "epoch": 0.3170731707317073,
      "grad_norm": 30.8470458984375,
      "learning_rate": 3.012e-05,
      "loss": 2.1764,
      "step": 143
    },
    {
      "epoch": 0.31929046563192903,
      "grad_norm": 19.84284210205078,
      "learning_rate": 2.9591578947368418e-05,
      "loss": 1.059,
      "step": 144
    },
    {
      "epoch": 0.3215077605321508,
      "grad_norm": 23.532106399536133,
      "learning_rate": 2.9063157894736842e-05,
      "loss": 1.9666,
      "step": 145
    },
    {
      "epoch": 0.3237250554323725,
      "grad_norm": 17.46700668334961,
      "learning_rate": 2.8534736842105264e-05,
      "loss": 1.5324,
      "step": 146
    },
    {
      "epoch": 0.3259423503325942,
      "grad_norm": 23.52845001220703,
      "learning_rate": 2.800631578947368e-05,
      "loss": 1.3784,
      "step": 147
    },
    {
      "epoch": 0.328159645232816,
      "grad_norm": 27.91365623474121,
      "learning_rate": 2.7477894736842106e-05,
      "loss": 1.6899,
      "step": 148
    },
    {
      "epoch": 0.3303769401330377,
      "grad_norm": 32.38272476196289,
      "learning_rate": 2.6949473684210527e-05,
      "loss": 1.6625,
      "step": 149
    },
    {
      "epoch": 0.3325942350332594,
      "grad_norm": 51.69731140136719,
      "learning_rate": 2.6421052631578945e-05,
      "loss": 1.9981,
      "step": 150
    },
    {
      "epoch": 0.3325942350332594,
      "eval_loss": 0.8841075897216797,
      "eval_runtime": 20.2428,
      "eval_samples_per_second": 9.386,
      "eval_steps_per_second": 2.371,
      "step": 150
    },
    {
      "epoch": 0.3348115299334812,
      "grad_norm": 9.736676216125488,
      "learning_rate": 2.589263157894737e-05,
      "loss": 1.6967,
      "step": 151
    },
    {
      "epoch": 0.3370288248337029,
      "grad_norm": 10.335311889648438,
      "learning_rate": 2.536421052631579e-05,
      "loss": 1.6221,
      "step": 152
    },
    {
      "epoch": 0.3392461197339246,
      "grad_norm": 10.948195457458496,
      "learning_rate": 2.483578947368421e-05,
      "loss": 1.891,
      "step": 153
    },
    {
      "epoch": 0.34146341463414637,
      "grad_norm": 10.740894317626953,
      "learning_rate": 2.430736842105263e-05,
      "loss": 1.5576,
      "step": 154
    },
    {
      "epoch": 0.3436807095343681,
      "grad_norm": 10.873139381408691,
      "learning_rate": 2.3778947368421052e-05,
      "loss": 1.5878,
      "step": 155
    },
    {
      "epoch": 0.3458980044345898,
      "grad_norm": 14.580072402954102,
      "learning_rate": 2.3250526315789473e-05,
      "loss": 1.9496,
      "step": 156
    },
    {
      "epoch": 0.34811529933481156,
      "grad_norm": 12.087544441223145,
      "learning_rate": 2.2722105263157894e-05,
      "loss": 1.8294,
      "step": 157
    },
    {
      "epoch": 0.35033259423503327,
      "grad_norm": 14.1340913772583,
      "learning_rate": 2.2193684210526316e-05,
      "loss": 1.8763,
      "step": 158
    },
    {
      "epoch": 0.352549889135255,
      "grad_norm": 11.832399368286133,
      "learning_rate": 2.1665263157894737e-05,
      "loss": 1.771,
      "step": 159
    },
    {
      "epoch": 0.35476718403547675,
      "grad_norm": 14.125495910644531,
      "learning_rate": 2.1136842105263158e-05,
      "loss": 2.1433,
      "step": 160
    },
    {
      "epoch": 0.35698447893569846,
      "grad_norm": 16.098318099975586,
      "learning_rate": 2.060842105263158e-05,
      "loss": 1.5098,
      "step": 161
    },
    {
      "epoch": 0.35920177383592017,
      "grad_norm": 11.606100082397461,
      "learning_rate": 2.008e-05,
      "loss": 1.531,
      "step": 162
    },
    {
      "epoch": 0.3614190687361419,
      "grad_norm": 14.126666069030762,
      "learning_rate": 1.9551578947368422e-05,
      "loss": 2.1956,
      "step": 163
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 13.657944679260254,
      "learning_rate": 1.9023157894736843e-05,
      "loss": 2.0216,
      "step": 164
    },
    {
      "epoch": 0.36585365853658536,
      "grad_norm": 16.890853881835938,
      "learning_rate": 1.849473684210526e-05,
      "loss": 2.0006,
      "step": 165
    },
    {
      "epoch": 0.36807095343680707,
      "grad_norm": 12.908790588378906,
      "learning_rate": 1.7966315789473686e-05,
      "loss": 1.6173,
      "step": 166
    },
    {
      "epoch": 0.37028824833702884,
      "grad_norm": 17.241039276123047,
      "learning_rate": 1.7437894736842107e-05,
      "loss": 1.9784,
      "step": 167
    },
    {
      "epoch": 0.37250554323725055,
      "grad_norm": 15.207599639892578,
      "learning_rate": 1.6909473684210525e-05,
      "loss": 1.5988,
      "step": 168
    },
    {
      "epoch": 0.37472283813747226,
      "grad_norm": 15.888737678527832,
      "learning_rate": 1.638105263157895e-05,
      "loss": 1.4648,
      "step": 169
    },
    {
      "epoch": 0.376940133037694,
      "grad_norm": 12.339811325073242,
      "learning_rate": 1.5852631578947368e-05,
      "loss": 1.5447,
      "step": 170
    },
    {
      "epoch": 0.37915742793791574,
      "grad_norm": 12.044404983520508,
      "learning_rate": 1.532421052631579e-05,
      "loss": 1.3339,
      "step": 171
    },
    {
      "epoch": 0.38137472283813745,
      "grad_norm": 11.459868431091309,
      "learning_rate": 1.4795789473684209e-05,
      "loss": 1.4563,
      "step": 172
    },
    {
      "epoch": 0.3835920177383592,
      "grad_norm": 14.248141288757324,
      "learning_rate": 1.4267368421052632e-05,
      "loss": 1.5893,
      "step": 173
    },
    {
      "epoch": 0.3858093126385809,
      "grad_norm": 19.118242263793945,
      "learning_rate": 1.3738947368421053e-05,
      "loss": 1.9648,
      "step": 174
    },
    {
      "epoch": 0.38802660753880264,
      "grad_norm": 15.491469383239746,
      "learning_rate": 1.3210526315789473e-05,
      "loss": 1.5081,
      "step": 175
    },
    {
      "epoch": 0.3902439024390244,
      "grad_norm": 13.472190856933594,
      "learning_rate": 1.2682105263157896e-05,
      "loss": 1.2014,
      "step": 176
    },
    {
      "epoch": 0.3924611973392461,
      "grad_norm": 16.698389053344727,
      "learning_rate": 1.2153684210526315e-05,
      "loss": 1.6185,
      "step": 177
    },
    {
      "epoch": 0.3946784922394678,
      "grad_norm": 14.444684028625488,
      "learning_rate": 1.1625263157894737e-05,
      "loss": 1.663,
      "step": 178
    },
    {
      "epoch": 0.3968957871396896,
      "grad_norm": 12.130651473999023,
      "learning_rate": 1.1096842105263158e-05,
      "loss": 1.1374,
      "step": 179
    },
    {
      "epoch": 0.3991130820399113,
      "grad_norm": 13.900895118713379,
      "learning_rate": 1.0568421052631579e-05,
      "loss": 1.3243,
      "step": 180
    },
    {
      "epoch": 0.401330376940133,
      "grad_norm": 15.653491020202637,
      "learning_rate": 1.004e-05,
      "loss": 1.7645,
      "step": 181
    },
    {
      "epoch": 0.4035476718403548,
      "grad_norm": 15.8237886428833,
      "learning_rate": 9.511578947368422e-06,
      "loss": 1.6201,
      "step": 182
    },
    {
      "epoch": 0.4057649667405765,
      "grad_norm": 32.045345306396484,
      "learning_rate": 8.983157894736843e-06,
      "loss": 1.7167,
      "step": 183
    },
    {
      "epoch": 0.4079822616407982,
      "grad_norm": 13.554454803466797,
      "learning_rate": 8.454736842105263e-06,
      "loss": 1.1857,
      "step": 184
    },
    {
      "epoch": 0.41019955654101997,
      "grad_norm": 13.749832153320312,
      "learning_rate": 7.926315789473684e-06,
      "loss": 1.0891,
      "step": 185
    },
    {
      "epoch": 0.4124168514412417,
      "grad_norm": 16.378047943115234,
      "learning_rate": 7.397894736842104e-06,
      "loss": 1.244,
      "step": 186
    },
    {
      "epoch": 0.4146341463414634,
      "grad_norm": 20.462013244628906,
      "learning_rate": 6.8694736842105265e-06,
      "loss": 1.659,
      "step": 187
    },
    {
      "epoch": 0.41685144124168516,
      "grad_norm": 18.432472229003906,
      "learning_rate": 6.341052631578948e-06,
      "loss": 1.8659,
      "step": 188
    },
    {
      "epoch": 0.4190687361419069,
      "grad_norm": 21.487030029296875,
      "learning_rate": 5.812631578947368e-06,
      "loss": 1.7793,
      "step": 189
    },
    {
      "epoch": 0.4212860310421286,
      "grad_norm": 27.553062438964844,
      "learning_rate": 5.2842105263157896e-06,
      "loss": 2.1433,
      "step": 190
    },
    {
      "epoch": 0.42350332594235035,
      "grad_norm": 20.972902297973633,
      "learning_rate": 4.755789473684211e-06,
      "loss": 1.3501,
      "step": 191
    },
    {
      "epoch": 0.42572062084257206,
      "grad_norm": 21.430740356445312,
      "learning_rate": 4.227368421052631e-06,
      "loss": 1.2867,
      "step": 192
    },
    {
      "epoch": 0.4279379157427938,
      "grad_norm": 21.622007369995117,
      "learning_rate": 3.698947368421052e-06,
      "loss": 1.8264,
      "step": 193
    },
    {
      "epoch": 0.43015521064301554,
      "grad_norm": 18.00780487060547,
      "learning_rate": 3.170526315789474e-06,
      "loss": 1.5396,
      "step": 194
    },
    {
      "epoch": 0.43237250554323725,
      "grad_norm": 21.456939697265625,
      "learning_rate": 2.6421052631578948e-06,
      "loss": 0.8305,
      "step": 195
    },
    {
      "epoch": 0.43458980044345896,
      "grad_norm": 37.9657096862793,
      "learning_rate": 2.1136842105263157e-06,
      "loss": 2.433,
      "step": 196
    },
    {
      "epoch": 0.43680709534368073,
      "grad_norm": 31.422948837280273,
      "learning_rate": 1.585263157894737e-06,
      "loss": 1.5014,
      "step": 197
    },
    {
      "epoch": 0.43902439024390244,
      "grad_norm": 22.746129989624023,
      "learning_rate": 1.0568421052631578e-06,
      "loss": 1.9673,
      "step": 198
    },
    {
      "epoch": 0.44124168514412415,
      "grad_norm": 45.2837028503418,
      "learning_rate": 5.284210526315789e-07,
      "loss": 1.036,
      "step": 199
    },
    {
      "epoch": 0.4434589800443459,
      "grad_norm": 56.693782806396484,
      "learning_rate": 0.0,
      "loss": 2.1921,
      "step": 200
    },
    {
      "epoch": 0.4434589800443459,
      "eval_loss": 0.8263623714447021,
      "eval_runtime": 20.2153,
      "eval_samples_per_second": 9.399,
      "eval_steps_per_second": 2.374,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.15499809800192e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}