{
  "best_metric": 1.5330638885498047,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 1.82648401826484,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0091324200913242,
      "grad_norm": 14.385358810424805,
      "learning_rate": 1e-05,
      "loss": 8.1778,
      "step": 1
    },
    {
      "epoch": 0.0091324200913242,
      "eval_loss": 2.0388219356536865,
      "eval_runtime": 27.1877,
      "eval_samples_per_second": 6.805,
      "eval_steps_per_second": 1.729,
      "step": 1
    },
    {
      "epoch": 0.0182648401826484,
      "grad_norm": 14.151931762695312,
      "learning_rate": 2e-05,
      "loss": 7.9544,
      "step": 2
    },
    {
      "epoch": 0.0273972602739726,
      "grad_norm": 11.721120834350586,
      "learning_rate": 3e-05,
      "loss": 8.0542,
      "step": 3
    },
    {
      "epoch": 0.0365296803652968,
      "grad_norm": 7.312915325164795,
      "learning_rate": 4e-05,
      "loss": 7.4967,
      "step": 4
    },
    {
      "epoch": 0.045662100456621,
      "grad_norm": 7.985286235809326,
      "learning_rate": 5e-05,
      "loss": 7.3073,
      "step": 5
    },
    {
      "epoch": 0.0547945205479452,
      "grad_norm": 7.377535343170166,
      "learning_rate": 6e-05,
      "loss": 7.2824,
      "step": 6
    },
    {
      "epoch": 0.0639269406392694,
      "grad_norm": 6.818817615509033,
      "learning_rate": 7e-05,
      "loss": 7.467,
      "step": 7
    },
    {
      "epoch": 0.0730593607305936,
      "grad_norm": 5.796252250671387,
      "learning_rate": 8e-05,
      "loss": 7.1328,
      "step": 8
    },
    {
      "epoch": 0.0821917808219178,
      "grad_norm": 4.33521032333374,
      "learning_rate": 9e-05,
      "loss": 6.6809,
      "step": 9
    },
    {
      "epoch": 0.091324200913242,
      "grad_norm": 4.325790882110596,
      "learning_rate": 0.0001,
      "loss": 7.0585,
      "step": 10
    },
    {
      "epoch": 0.1004566210045662,
      "grad_norm": 3.940777063369751,
      "learning_rate": 9.999316524962345e-05,
      "loss": 6.9329,
      "step": 11
    },
    {
      "epoch": 0.1095890410958904,
      "grad_norm": 3.775146007537842,
      "learning_rate": 9.997266286704631e-05,
      "loss": 6.7033,
      "step": 12
    },
    {
      "epoch": 0.1187214611872146,
      "grad_norm": 3.593841314315796,
      "learning_rate": 9.993849845741524e-05,
      "loss": 7.0459,
      "step": 13
    },
    {
      "epoch": 0.1278538812785388,
      "grad_norm": 3.878154754638672,
      "learning_rate": 9.989068136093873e-05,
      "loss": 7.1871,
      "step": 14
    },
    {
      "epoch": 0.136986301369863,
      "grad_norm": 3.4116504192352295,
      "learning_rate": 9.98292246503335e-05,
      "loss": 6.8704,
      "step": 15
    },
    {
      "epoch": 0.1461187214611872,
      "grad_norm": 3.4045538902282715,
      "learning_rate": 9.975414512725057e-05,
      "loss": 6.9161,
      "step": 16
    },
    {
      "epoch": 0.1552511415525114,
      "grad_norm": 3.3985300064086914,
      "learning_rate": 9.966546331768191e-05,
      "loss": 7.0666,
      "step": 17
    },
    {
      "epoch": 0.1643835616438356,
      "grad_norm": 3.1416335105895996,
      "learning_rate": 9.956320346634876e-05,
      "loss": 6.6545,
      "step": 18
    },
    {
      "epoch": 0.1735159817351598,
      "grad_norm": 3.2489981651306152,
      "learning_rate": 9.944739353007344e-05,
      "loss": 6.5647,
      "step": 19
    },
    {
      "epoch": 0.182648401826484,
      "grad_norm": 3.2142012119293213,
      "learning_rate": 9.931806517013612e-05,
      "loss": 6.8472,
      "step": 20
    },
    {
      "epoch": 0.1917808219178082,
      "grad_norm": 2.957822561264038,
      "learning_rate": 9.917525374361912e-05,
      "loss": 6.6633,
      "step": 21
    },
    {
      "epoch": 0.2009132420091324,
      "grad_norm": 2.905742645263672,
      "learning_rate": 9.901899829374047e-05,
      "loss": 6.729,
      "step": 22
    },
    {
      "epoch": 0.2100456621004566,
      "grad_norm": 3.0980546474456787,
      "learning_rate": 9.884934153917997e-05,
      "loss": 6.5608,
      "step": 23
    },
    {
      "epoch": 0.2191780821917808,
      "grad_norm": 2.8622286319732666,
      "learning_rate": 9.86663298624003e-05,
      "loss": 6.5351,
      "step": 24
    },
    {
      "epoch": 0.228310502283105,
      "grad_norm": 3.3140783309936523,
      "learning_rate": 9.847001329696653e-05,
      "loss": 7.1661,
      "step": 25
    },
    {
      "epoch": 0.2374429223744292,
      "grad_norm": 3.69408917427063,
      "learning_rate": 9.826044551386744e-05,
      "loss": 7.251,
      "step": 26
    },
    {
      "epoch": 0.2465753424657534,
      "grad_norm": 4.077339172363281,
      "learning_rate": 9.803768380684242e-05,
      "loss": 7.2643,
      "step": 27
    },
    {
      "epoch": 0.2557077625570776,
      "grad_norm": 3.3373944759368896,
      "learning_rate": 9.780178907671789e-05,
      "loss": 6.7573,
      "step": 28
    },
    {
      "epoch": 0.2648401826484018,
      "grad_norm": 2.928842782974243,
      "learning_rate": 9.755282581475769e-05,
      "loss": 6.7302,
      "step": 29
    },
    {
      "epoch": 0.273972602739726,
      "grad_norm": 2.864501714706421,
      "learning_rate": 9.729086208503174e-05,
      "loss": 6.3781,
      "step": 30
    },
    {
      "epoch": 0.2831050228310502,
      "grad_norm": 2.730508327484131,
      "learning_rate": 9.701596950580806e-05,
      "loss": 6.475,
      "step": 31
    },
    {
      "epoch": 0.2922374429223744,
      "grad_norm": 2.8813071250915527,
      "learning_rate": 9.672822322997305e-05,
      "loss": 6.7599,
      "step": 32
    },
    {
      "epoch": 0.3013698630136986,
      "grad_norm": 2.6118905544281006,
      "learning_rate": 9.642770192448536e-05,
      "loss": 6.2425,
      "step": 33
    },
    {
      "epoch": 0.3105022831050228,
      "grad_norm": 2.760718584060669,
      "learning_rate": 9.611448774886924e-05,
      "loss": 6.5296,
      "step": 34
    },
    {
      "epoch": 0.319634703196347,
      "grad_norm": 2.623293876647949,
      "learning_rate": 9.578866633275288e-05,
      "loss": 6.5239,
      "step": 35
    },
    {
      "epoch": 0.3287671232876712,
      "grad_norm": 2.558960199356079,
      "learning_rate": 9.545032675245813e-05,
      "loss": 6.4783,
      "step": 36
    },
    {
      "epoch": 0.3378995433789954,
      "grad_norm": 2.5984208583831787,
      "learning_rate": 9.509956150664796e-05,
      "loss": 6.6265,
      "step": 37
    },
    {
      "epoch": 0.3470319634703196,
      "grad_norm": 2.673401117324829,
      "learning_rate": 9.473646649103818e-05,
      "loss": 6.4988,
      "step": 38
    },
    {
      "epoch": 0.3561643835616438,
      "grad_norm": 2.5916404724121094,
      "learning_rate": 9.43611409721806e-05,
      "loss": 6.5329,
      "step": 39
    },
    {
      "epoch": 0.365296803652968,
      "grad_norm": 2.618675947189331,
      "learning_rate": 9.397368756032445e-05,
      "loss": 6.3273,
      "step": 40
    },
    {
      "epoch": 0.3744292237442922,
      "grad_norm": 2.535189151763916,
      "learning_rate": 9.357421218136386e-05,
      "loss": 6.4054,
      "step": 41
    },
    {
      "epoch": 0.3835616438356164,
      "grad_norm": 2.5565288066864014,
      "learning_rate": 9.316282404787871e-05,
      "loss": 6.6651,
      "step": 42
    },
    {
      "epoch": 0.3926940639269406,
      "grad_norm": 2.619098663330078,
      "learning_rate": 9.273963562927695e-05,
      "loss": 6.6821,
      "step": 43
    },
    {
      "epoch": 0.4018264840182648,
      "grad_norm": 2.448279857635498,
      "learning_rate": 9.230476262104677e-05,
      "loss": 6.2822,
      "step": 44
    },
    {
      "epoch": 0.410958904109589,
      "grad_norm": 2.4789090156555176,
      "learning_rate": 9.185832391312644e-05,
      "loss": 6.4207,
      "step": 45
    },
    {
      "epoch": 0.4200913242009132,
      "grad_norm": 2.5198521614074707,
      "learning_rate": 9.140044155740101e-05,
      "loss": 6.4779,
      "step": 46
    },
    {
      "epoch": 0.4292237442922374,
      "grad_norm": 2.563779830932617,
      "learning_rate": 9.093124073433463e-05,
      "loss": 6.5538,
      "step": 47
    },
    {
      "epoch": 0.4383561643835616,
      "grad_norm": 2.424715280532837,
      "learning_rate": 9.045084971874738e-05,
      "loss": 6.3172,
      "step": 48
    },
    {
      "epoch": 0.4474885844748858,
      "grad_norm": 2.480990409851074,
      "learning_rate": 8.995939984474624e-05,
      "loss": 6.3953,
      "step": 49
    },
    {
      "epoch": 0.45662100456621,
      "grad_norm": 2.5143067836761475,
      "learning_rate": 8.945702546981969e-05,
      "loss": 6.5031,
      "step": 50
    },
    {
      "epoch": 0.45662100456621,
      "eval_loss": 1.6045836210250854,
      "eval_runtime": 27.369,
      "eval_samples_per_second": 6.759,
      "eval_steps_per_second": 1.717,
      "step": 50
    },
    {
      "epoch": 0.4657534246575342,
      "grad_norm": 2.5302507877349854,
      "learning_rate": 8.894386393810563e-05,
      "loss": 6.5664,
      "step": 51
    },
    {
      "epoch": 0.4748858447488584,
      "grad_norm": 2.7475790977478027,
      "learning_rate": 8.842005554284296e-05,
      "loss": 6.9106,
      "step": 52
    },
    {
      "epoch": 0.4840182648401826,
      "grad_norm": 2.9009580612182617,
      "learning_rate": 8.788574348801675e-05,
      "loss": 6.76,
      "step": 53
    },
    {
      "epoch": 0.4931506849315068,
      "grad_norm": 3.160198926925659,
      "learning_rate": 8.73410738492077e-05,
      "loss": 6.7027,
      "step": 54
    },
    {
      "epoch": 0.502283105022831,
      "grad_norm": 2.6028199195861816,
      "learning_rate": 8.678619553365659e-05,
      "loss": 6.3942,
      "step": 55
    },
    {
      "epoch": 0.5114155251141552,
      "grad_norm": 2.5132687091827393,
      "learning_rate": 8.622126023955446e-05,
      "loss": 6.4247,
      "step": 56
    },
    {
      "epoch": 0.5205479452054794,
      "grad_norm": 2.346787929534912,
      "learning_rate": 8.564642241456986e-05,
      "loss": 6.219,
      "step": 57
    },
    {
      "epoch": 0.5296803652968036,
      "grad_norm": 2.4157371520996094,
      "learning_rate": 8.506183921362443e-05,
      "loss": 6.407,
      "step": 58
    },
    {
      "epoch": 0.5388127853881278,
      "grad_norm": 2.3487019538879395,
      "learning_rate": 8.44676704559283e-05,
      "loss": 6.0988,
      "step": 59
    },
    {
      "epoch": 0.547945205479452,
      "grad_norm": 2.297541379928589,
      "learning_rate": 8.386407858128706e-05,
      "loss": 5.9097,
      "step": 60
    },
    {
      "epoch": 0.5570776255707762,
      "grad_norm": 2.353090763092041,
      "learning_rate": 8.32512286056924e-05,
      "loss": 6.2177,
      "step": 61
    },
    {
      "epoch": 0.5662100456621004,
      "grad_norm": 2.456840753555298,
      "learning_rate": 8.262928807620843e-05,
      "loss": 6.2388,
      "step": 62
    },
    {
      "epoch": 0.5753424657534246,
      "grad_norm": 2.3819286823272705,
      "learning_rate": 8.199842702516583e-05,
      "loss": 6.1685,
      "step": 63
    },
    {
      "epoch": 0.5844748858447488,
      "grad_norm": 2.324075937271118,
      "learning_rate": 8.135881792367686e-05,
      "loss": 6.3753,
      "step": 64
    },
    {
      "epoch": 0.593607305936073,
      "grad_norm": 2.3761610984802246,
      "learning_rate": 8.07106356344834e-05,
      "loss": 6.4645,
      "step": 65
    },
    {
      "epoch": 0.6027397260273972,
      "grad_norm": 2.377534866333008,
      "learning_rate": 8.005405736415126e-05,
      "loss": 6.4852,
      "step": 66
    },
    {
      "epoch": 0.6118721461187214,
      "grad_norm": 2.298407793045044,
      "learning_rate": 7.938926261462366e-05,
      "loss": 6.1425,
      "step": 67
    },
    {
      "epoch": 0.6210045662100456,
      "grad_norm": 2.3773036003112793,
      "learning_rate": 7.871643313414718e-05,
      "loss": 6.4493,
      "step": 68
    },
    {
      "epoch": 0.6301369863013698,
      "grad_norm": 2.4844040870666504,
      "learning_rate": 7.803575286758364e-05,
      "loss": 6.3431,
      "step": 69
    },
    {
      "epoch": 0.639269406392694,
      "grad_norm": 2.3018264770507812,
      "learning_rate": 7.734740790612136e-05,
      "loss": 6.4451,
      "step": 70
    },
    {
      "epoch": 0.6484018264840182,
      "grad_norm": 2.306001901626587,
      "learning_rate": 7.66515864363997e-05,
      "loss": 6.4164,
      "step": 71
    },
    {
      "epoch": 0.6575342465753424,
      "grad_norm": 2.3219869136810303,
      "learning_rate": 7.594847868906076e-05,
      "loss": 6.258,
      "step": 72
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 2.3488428592681885,
      "learning_rate": 7.52382768867422e-05,
      "loss": 6.1906,
      "step": 73
    },
    {
      "epoch": 0.6757990867579908,
      "grad_norm": 2.3201406002044678,
      "learning_rate": 7.452117519152542e-05,
      "loss": 6.335,
      "step": 74
    },
    {
      "epoch": 0.684931506849315,
      "grad_norm": 2.4174585342407227,
      "learning_rate": 7.379736965185368e-05,
      "loss": 6.2573,
      "step": 75
    },
    {
      "epoch": 0.6940639269406392,
      "grad_norm": 2.428056240081787,
      "learning_rate": 7.30670581489344e-05,
      "loss": 6.3608,
      "step": 76
    },
    {
      "epoch": 0.7031963470319634,
      "grad_norm": 2.3151209354400635,
      "learning_rate": 7.233044034264034e-05,
      "loss": 6.3383,
      "step": 77
    },
    {
      "epoch": 0.7123287671232876,
      "grad_norm": 2.2870469093322754,
      "learning_rate": 7.158771761692464e-05,
      "loss": 6.2756,
      "step": 78
    },
    {
      "epoch": 0.7214611872146118,
      "grad_norm": 2.6297852993011475,
      "learning_rate": 7.083909302476453e-05,
      "loss": 6.8066,
      "step": 79
    },
    {
      "epoch": 0.730593607305936,
      "grad_norm": 2.835029125213623,
      "learning_rate": 7.008477123264848e-05,
      "loss": 6.7208,
      "step": 80
    },
    {
      "epoch": 0.7397260273972602,
      "grad_norm": 3.1156086921691895,
      "learning_rate": 6.932495846462261e-05,
      "loss": 6.5322,
      "step": 81
    },
    {
      "epoch": 0.7488584474885844,
      "grad_norm": 2.4558022022247314,
      "learning_rate": 6.855986244591104e-05,
      "loss": 6.1883,
      "step": 82
    },
    {
      "epoch": 0.7579908675799086,
      "grad_norm": 2.3941500186920166,
      "learning_rate": 6.778969234612584e-05,
      "loss": 6.3781,
      "step": 83
    },
    {
      "epoch": 0.7671232876712328,
      "grad_norm": 2.3683080673217773,
      "learning_rate": 6.701465872208216e-05,
      "loss": 6.5191,
      "step": 84
    },
    {
      "epoch": 0.776255707762557,
      "grad_norm": 2.3145902156829834,
      "learning_rate": 6.623497346023418e-05,
      "loss": 5.9855,
      "step": 85
    },
    {
      "epoch": 0.7853881278538812,
      "grad_norm": 2.3052635192871094,
      "learning_rate": 6.545084971874738e-05,
      "loss": 6.2429,
      "step": 86
    },
    {
      "epoch": 0.7945205479452054,
      "grad_norm": 2.2787489891052246,
      "learning_rate": 6.466250186922325e-05,
      "loss": 6.1291,
      "step": 87
    },
    {
      "epoch": 0.8036529680365296,
      "grad_norm": 2.3815743923187256,
      "learning_rate": 6.387014543809223e-05,
      "loss": 6.331,
      "step": 88
    },
    {
      "epoch": 0.8127853881278538,
      "grad_norm": 2.3296122550964355,
      "learning_rate": 6.307399704769099e-05,
      "loss": 6.062,
      "step": 89
    },
    {
      "epoch": 0.821917808219178,
      "grad_norm": 2.364459753036499,
      "learning_rate": 6.227427435703997e-05,
      "loss": 6.2991,
      "step": 90
    },
    {
      "epoch": 0.8310502283105022,
      "grad_norm": 2.296963930130005,
      "learning_rate": 6.147119600233758e-05,
      "loss": 6.3378,
      "step": 91
    },
    {
      "epoch": 0.8401826484018264,
      "grad_norm": 2.2709155082702637,
      "learning_rate": 6.066498153718735e-05,
      "loss": 6.0201,
      "step": 92
    },
    {
      "epoch": 0.8493150684931506,
      "grad_norm": 2.2849385738372803,
      "learning_rate": 5.985585137257401e-05,
      "loss": 6.3139,
      "step": 93
    },
    {
      "epoch": 0.8584474885844748,
      "grad_norm": 2.266611099243164,
      "learning_rate": 5.90440267166055e-05,
      "loss": 5.9835,
      "step": 94
    },
    {
      "epoch": 0.867579908675799,
      "grad_norm": 2.24310564994812,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 6.1997,
      "step": 95
    },
    {
      "epoch": 0.8767123287671232,
      "grad_norm": 2.283054828643799,
      "learning_rate": 5.74131823855921e-05,
      "loss": 6.3574,
      "step": 96
    },
    {
      "epoch": 0.8858447488584474,
      "grad_norm": 2.3451895713806152,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 6.5077,
      "step": 97
    },
    {
      "epoch": 0.8949771689497716,
      "grad_norm": 2.262666702270508,
      "learning_rate": 5.577423184847932e-05,
      "loss": 6.339,
      "step": 98
    },
    {
      "epoch": 0.9041095890410958,
      "grad_norm": 2.266411304473877,
      "learning_rate": 5.495227651252315e-05,
      "loss": 6.2461,
      "step": 99
    },
    {
      "epoch": 0.91324200913242,
      "grad_norm": 2.2958450317382812,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 6.3088,
      "step": 100
    },
    {
      "epoch": 0.91324200913242,
      "eval_loss": 1.5493108034133911,
      "eval_runtime": 27.3972,
      "eval_samples_per_second": 6.753,
      "eval_steps_per_second": 1.716,
      "step": 100
    },
    {
      "epoch": 0.9223744292237442,
      "grad_norm": 2.396834373474121,
      "learning_rate": 5.330452921628497e-05,
      "loss": 6.1975,
      "step": 101
    },
    {
      "epoch": 0.9315068493150684,
      "grad_norm": 2.2903473377227783,
      "learning_rate": 5.247918773366112e-05,
      "loss": 6.1992,
      "step": 102
    },
    {
      "epoch": 0.9406392694063926,
      "grad_norm": 2.2541141510009766,
      "learning_rate": 5.165316846586541e-05,
      "loss": 6.0902,
      "step": 103
    },
    {
      "epoch": 0.9497716894977168,
      "grad_norm": 2.27167010307312,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 6.1416,
      "step": 104
    },
    {
      "epoch": 0.958904109589041,
      "grad_norm": 2.301316261291504,
      "learning_rate": 5e-05,
      "loss": 6.4116,
      "step": 105
    },
    {
      "epoch": 0.9680365296803652,
      "grad_norm": 2.592348575592041,
      "learning_rate": 4.917330276168208e-05,
      "loss": 6.7602,
      "step": 106
    },
    {
      "epoch": 0.9771689497716894,
      "grad_norm": 2.704606294631958,
      "learning_rate": 4.834683153413459e-05,
      "loss": 6.5595,
      "step": 107
    },
    {
      "epoch": 0.9863013698630136,
      "grad_norm": 3.0188896656036377,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 6.7546,
      "step": 108
    },
    {
      "epoch": 0.9954337899543378,
      "grad_norm": 2.263741970062256,
      "learning_rate": 4.669547078371504e-05,
      "loss": 6.3075,
      "step": 109
    },
    {
      "epoch": 1.004566210045662,
      "grad_norm": 3.022768974304199,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 5.8854,
      "step": 110
    },
    {
      "epoch": 1.0136986301369864,
      "grad_norm": 2.2758378982543945,
      "learning_rate": 4.504772348747687e-05,
      "loss": 5.5925,
      "step": 111
    },
    {
      "epoch": 1.0228310502283104,
      "grad_norm": 2.3155620098114014,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 5.518,
      "step": 112
    },
    {
      "epoch": 1.0319634703196348,
      "grad_norm": 2.223759174346924,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 5.8703,
      "step": 113
    },
    {
      "epoch": 1.0410958904109588,
      "grad_norm": 2.2196123600006104,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 5.3927,
      "step": 114
    },
    {
      "epoch": 1.0502283105022832,
      "grad_norm": 2.228165626525879,
      "learning_rate": 4.17702704859633e-05,
      "loss": 5.239,
      "step": 115
    },
    {
      "epoch": 1.0593607305936072,
      "grad_norm": 2.4534826278686523,
      "learning_rate": 4.095597328339452e-05,
      "loss": 5.5712,
      "step": 116
    },
    {
      "epoch": 1.0684931506849316,
      "grad_norm": 2.5110085010528564,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 5.5295,
      "step": 117
    },
    {
      "epoch": 1.0776255707762556,
      "grad_norm": 2.49696683883667,
      "learning_rate": 3.933501846281267e-05,
      "loss": 5.5812,
      "step": 118
    },
    {
      "epoch": 1.08675799086758,
      "grad_norm": 2.450960159301758,
      "learning_rate": 3.852880399766243e-05,
      "loss": 5.4638,
      "step": 119
    },
    {
      "epoch": 1.095890410958904,
      "grad_norm": 2.3998756408691406,
      "learning_rate": 3.772572564296005e-05,
      "loss": 5.5282,
      "step": 120
    },
    {
      "epoch": 1.1050228310502284,
      "grad_norm": 2.3926761150360107,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 5.4748,
      "step": 121
    },
    {
      "epoch": 1.1141552511415524,
      "grad_norm": 2.308351516723633,
      "learning_rate": 3.612985456190778e-05,
      "loss": 5.533,
      "step": 122
    },
    {
      "epoch": 1.1232876712328768,
      "grad_norm": 2.2923483848571777,
      "learning_rate": 3.533749813077677e-05,
      "loss": 5.3549,
      "step": 123
    },
    {
      "epoch": 1.1324200913242009,
      "grad_norm": 2.2991902828216553,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 5.6274,
      "step": 124
    },
    {
      "epoch": 1.1415525114155252,
      "grad_norm": 2.311222553253174,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 5.5794,
      "step": 125
    },
    {
      "epoch": 1.1506849315068493,
      "grad_norm": 2.407327890396118,
      "learning_rate": 3.298534127791785e-05,
      "loss": 5.5524,
      "step": 126
    },
    {
      "epoch": 1.1598173515981736,
      "grad_norm": 2.2840681076049805,
      "learning_rate": 3.221030765387417e-05,
      "loss": 5.4964,
      "step": 127
    },
    {
      "epoch": 1.1689497716894977,
      "grad_norm": 2.29766845703125,
      "learning_rate": 3.144013755408895e-05,
      "loss": 5.4344,
      "step": 128
    },
    {
      "epoch": 1.178082191780822,
      "grad_norm": 2.382521390914917,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 5.4754,
      "step": 129
    },
    {
      "epoch": 1.187214611872146,
      "grad_norm": 2.4657342433929443,
      "learning_rate": 2.991522876735154e-05,
      "loss": 5.5409,
      "step": 130
    },
    {
      "epoch": 1.1963470319634704,
      "grad_norm": 2.4315481185913086,
      "learning_rate": 2.916090697523549e-05,
      "loss": 5.5372,
      "step": 131
    },
    {
      "epoch": 1.2054794520547945,
      "grad_norm": 2.3506505489349365,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 5.4611,
      "step": 132
    },
    {
      "epoch": 1.2146118721461188,
      "grad_norm": 2.3935210704803467,
      "learning_rate": 2.766955965735968e-05,
      "loss": 5.5723,
      "step": 133
    },
    {
      "epoch": 1.2237442922374429,
      "grad_norm": 2.5856597423553467,
      "learning_rate": 2.693294185106562e-05,
      "loss": 5.8027,
      "step": 134
    },
    {
      "epoch": 1.2328767123287672,
      "grad_norm": 2.764137029647827,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 5.609,
      "step": 135
    },
    {
      "epoch": 1.2420091324200913,
      "grad_norm": 2.958418846130371,
      "learning_rate": 2.547882480847461e-05,
      "loss": 5.6849,
      "step": 136
    },
    {
      "epoch": 1.2511415525114156,
      "grad_norm": 2.9182913303375244,
      "learning_rate": 2.476172311325783e-05,
      "loss": 5.3936,
      "step": 137
    },
    {
      "epoch": 1.2602739726027397,
      "grad_norm": 2.536137342453003,
      "learning_rate": 2.405152131093926e-05,
      "loss": 5.4941,
      "step": 138
    },
    {
      "epoch": 1.269406392694064,
      "grad_norm": 2.4414384365081787,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 5.395,
      "step": 139
    },
    {
      "epoch": 1.278538812785388,
      "grad_norm": 2.420905113220215,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 5.3234,
      "step": 140
    },
    {
      "epoch": 1.2876712328767124,
      "grad_norm": 2.4749791622161865,
      "learning_rate": 2.196424713241637e-05,
      "loss": 5.3562,
      "step": 141
    },
    {
      "epoch": 1.2968036529680365,
      "grad_norm": 2.415034770965576,
      "learning_rate": 2.128356686585282e-05,
      "loss": 5.3091,
      "step": 142
    },
    {
      "epoch": 1.3059360730593608,
      "grad_norm": 2.386012077331543,
      "learning_rate": 2.061073738537635e-05,
      "loss": 5.2667,
      "step": 143
    },
    {
      "epoch": 1.3150684931506849,
      "grad_norm": 2.4848532676696777,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 5.4513,
      "step": 144
    },
    {
      "epoch": 1.3242009132420092,
      "grad_norm": 2.357064723968506,
      "learning_rate": 1.928936436551661e-05,
      "loss": 5.2696,
      "step": 145
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 2.3330743312835693,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 5.2867,
      "step": 146
    },
    {
      "epoch": 1.3424657534246576,
      "grad_norm": 2.316640853881836,
      "learning_rate": 1.800157297483417e-05,
      "loss": 5.4867,
      "step": 147
    },
    {
      "epoch": 1.3515981735159817,
      "grad_norm": 2.335223436355591,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 5.3442,
      "step": 148
    },
    {
      "epoch": 1.360730593607306,
      "grad_norm": 2.307755470275879,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 5.4288,
      "step": 149
    },
    {
      "epoch": 1.36986301369863,
      "grad_norm": 2.3415632247924805,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 5.4639,
      "step": 150
    },
    {
      "epoch": 1.36986301369863,
      "eval_loss": 1.5374356508255005,
      "eval_runtime": 27.4806,
      "eval_samples_per_second": 6.732,
      "eval_steps_per_second": 1.71,
      "step": 150
    },
    {
      "epoch": 1.3789954337899544,
      "grad_norm": 2.371952772140503,
      "learning_rate": 1.553232954407171e-05,
      "loss": 5.4727,
      "step": 151
    },
    {
      "epoch": 1.3881278538812785,
      "grad_norm": 2.3645217418670654,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 5.4931,
      "step": 152
    },
    {
      "epoch": 1.3972602739726028,
      "grad_norm": 2.39043927192688,
      "learning_rate": 1.435357758543015e-05,
      "loss": 5.5367,
      "step": 153
    },
    {
      "epoch": 1.4063926940639269,
      "grad_norm": 2.3830485343933105,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 5.5211,
      "step": 154
    },
    {
      "epoch": 1.4155251141552512,
      "grad_norm": 2.3963820934295654,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 5.2548,
      "step": 155
    },
    {
      "epoch": 1.4246575342465753,
      "grad_norm": 2.45495867729187,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 5.5063,
      "step": 156
    },
    {
      "epoch": 1.4337899543378996,
      "grad_norm": 2.411886692047119,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 5.3343,
      "step": 157
    },
    {
      "epoch": 1.4429223744292237,
      "grad_norm": 2.4337449073791504,
      "learning_rate": 1.157994445715706e-05,
      "loss": 5.5986,
      "step": 158
    },
    {
      "epoch": 1.452054794520548,
      "grad_norm": 2.3690693378448486,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 5.3817,
      "step": 159
    },
    {
      "epoch": 1.461187214611872,
      "grad_norm": 2.3939321041107178,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 5.5521,
      "step": 160
    },
    {
      "epoch": 1.4703196347031964,
      "grad_norm": 2.4792160987854004,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 5.7174,
      "step": 161
    },
    {
      "epoch": 1.4794520547945205,
      "grad_norm": 2.723461627960205,
      "learning_rate": 9.549150281252633e-06,
      "loss": 5.8516,
      "step": 162
    },
    {
      "epoch": 1.4885844748858448,
      "grad_norm": 2.8717315196990967,
      "learning_rate": 9.068759265665384e-06,
      "loss": 5.9104,
      "step": 163
    },
    {
      "epoch": 1.4977168949771689,
      "grad_norm": 2.8555715084075928,
      "learning_rate": 8.599558442598998e-06,
      "loss": 5.4618,
      "step": 164
    },
    {
      "epoch": 1.5068493150684932,
      "grad_norm": 2.5716333389282227,
      "learning_rate": 8.141676086873572e-06,
      "loss": 5.5172,
      "step": 165
    },
    {
      "epoch": 1.5159817351598175,
      "grad_norm": 2.5310869216918945,
      "learning_rate": 7.695237378953223e-06,
      "loss": 5.5131,
      "step": 166
    },
    {
      "epoch": 1.5251141552511416,
      "grad_norm": 2.482553720474243,
      "learning_rate": 7.260364370723044e-06,
      "loss": 5.6114,
      "step": 167
    },
    {
      "epoch": 1.5342465753424657,
      "grad_norm": 2.4653866291046143,
      "learning_rate": 6.837175952121306e-06,
      "loss": 5.4987,
      "step": 168
    },
    {
      "epoch": 1.54337899543379,
      "grad_norm": 2.417545795440674,
      "learning_rate": 6.425787818636131e-06,
      "loss": 5.0537,
      "step": 169
    },
    {
      "epoch": 1.5525114155251143,
      "grad_norm": 2.4840376377105713,
      "learning_rate": 6.026312439675552e-06,
      "loss": 5.487,
      "step": 170
    },
    {
      "epoch": 1.5616438356164384,
      "grad_norm": 2.5079848766326904,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 5.4829,
      "step": 171
    },
    {
      "epoch": 1.5707762557077625,
      "grad_norm": 2.4372551441192627,
      "learning_rate": 5.263533508961827e-06,
      "loss": 5.2636,
      "step": 172
    },
    {
      "epoch": 1.5799086757990868,
      "grad_norm": 2.441577196121216,
      "learning_rate": 4.900438493352055e-06,
      "loss": 5.5759,
      "step": 173
    },
    {
      "epoch": 1.589041095890411,
      "grad_norm": 2.4013524055480957,
      "learning_rate": 4.549673247541875e-06,
      "loss": 5.3549,
      "step": 174
    },
    {
      "epoch": 1.5981735159817352,
      "grad_norm": 2.3573992252349854,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 5.4232,
      "step": 175
    },
    {
      "epoch": 1.6073059360730593,
      "grad_norm": 2.3793928623199463,
      "learning_rate": 3.885512251130763e-06,
      "loss": 5.3732,
      "step": 176
    },
    {
      "epoch": 1.6164383561643836,
      "grad_norm": 2.334749698638916,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 5.2292,
      "step": 177
    },
    {
      "epoch": 1.625570776255708,
      "grad_norm": 2.452176094055176,
      "learning_rate": 3.271776770026963e-06,
      "loss": 5.5245,
      "step": 178
    },
    {
      "epoch": 1.634703196347032,
      "grad_norm": 2.385550022125244,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 5.5387,
      "step": 179
    },
    {
      "epoch": 1.643835616438356,
      "grad_norm": 2.3744089603424072,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 5.4045,
      "step": 180
    },
    {
      "epoch": 1.6529680365296804,
      "grad_norm": 2.295444965362549,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 5.2876,
      "step": 181
    },
    {
      "epoch": 1.6621004566210047,
      "grad_norm": 2.4016926288604736,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 5.4212,
      "step": 182
    },
    {
      "epoch": 1.6712328767123288,
      "grad_norm": 2.3749890327453613,
      "learning_rate": 1.962316193157593e-06,
      "loss": 5.4061,
      "step": 183
    },
    {
      "epoch": 1.6803652968036529,
      "grad_norm": 2.3701069355010986,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 5.2654,
      "step": 184
    },
    {
      "epoch": 1.6894977168949772,
      "grad_norm": 2.3576807975769043,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 5.2827,
      "step": 185
    },
    {
      "epoch": 1.6986301369863015,
      "grad_norm": 2.338988780975342,
      "learning_rate": 1.333670137599713e-06,
      "loss": 5.202,
      "step": 186
    },
    {
      "epoch": 1.7077625570776256,
      "grad_norm": 2.453352212905884,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 5.6261,
      "step": 187
    },
    {
      "epoch": 1.7168949771689497,
      "grad_norm": 2.4335148334503174,
      "learning_rate": 9.810017062595322e-07,
      "loss": 5.5933,
      "step": 188
    },
    {
      "epoch": 1.726027397260274,
      "grad_norm": 2.6970746517181396,
      "learning_rate": 8.247462563808817e-07,
      "loss": 5.6317,
      "step": 189
    },
    {
      "epoch": 1.7351598173515983,
      "grad_norm": 2.8925790786743164,
      "learning_rate": 6.819348298638839e-07,
      "loss": 5.6674,
      "step": 190
    },
    {
      "epoch": 1.7442922374429224,
      "grad_norm": 2.8324902057647705,
      "learning_rate": 5.526064699265753e-07,
      "loss": 5.219,
      "step": 191
    },
    {
      "epoch": 1.7534246575342465,
      "grad_norm": 2.6506247520446777,
      "learning_rate": 4.367965336512403e-07,
      "loss": 5.7944,
      "step": 192
    },
    {
      "epoch": 1.7625570776255708,
      "grad_norm": 2.4575681686401367,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 5.1708,
      "step": 193
    },
    {
      "epoch": 1.771689497716895,
      "grad_norm": 2.475024700164795,
      "learning_rate": 2.458548727494292e-07,
      "loss": 5.5471,
      "step": 194
    },
    {
      "epoch": 1.7808219178082192,
      "grad_norm": 2.4819118976593018,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 5.4551,
      "step": 195
    },
    {
      "epoch": 1.7899543378995433,
      "grad_norm": 2.43180775642395,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 5.2573,
      "step": 196
    },
    {
      "epoch": 1.7990867579908676,
      "grad_norm": 2.386868715286255,
      "learning_rate": 6.150154258476315e-08,
      "loss": 5.3956,
      "step": 197
    },
    {
      "epoch": 1.808219178082192,
      "grad_norm": 2.47499680519104,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 5.6229,
      "step": 198
    },
    {
      "epoch": 1.817351598173516,
      "grad_norm": 2.3925249576568604,
      "learning_rate": 6.834750376549792e-09,
      "loss": 5.3068,
      "step": 199
    },
    {
      "epoch": 1.82648401826484,
      "grad_norm": 2.396662473678589,
      "learning_rate": 0.0,
      "loss": 5.4254,
      "step": 200
    },
    {
      "epoch": 1.82648401826484,
      "eval_loss": 1.5330638885498047,
      "eval_runtime": 27.3807,
      "eval_samples_per_second": 6.757,
      "eval_steps_per_second": 1.717,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 5.567930082388869e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}