{
  "best_metric": 0.9513868689537048,
  "best_model_checkpoint": "miner_id_24/checkpoint-150",
  "epoch": 1.1816838995568686,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005908419497784343,
      "grad_norm": 51.34055709838867,
      "learning_rate": 1e-05,
      "loss": 6.3437,
      "step": 1
    },
    {
      "epoch": 0.005908419497784343,
      "eval_loss": 1.8814340829849243,
      "eval_runtime": 20.607,
      "eval_samples_per_second": 13.879,
      "eval_steps_per_second": 3.494,
      "step": 1
    },
    {
      "epoch": 0.011816838995568686,
      "grad_norm": 51.71321105957031,
      "learning_rate": 2e-05,
      "loss": 6.4062,
      "step": 2
    },
    {
      "epoch": 0.01772525849335303,
      "grad_norm": 47.652870178222656,
      "learning_rate": 3e-05,
      "loss": 6.0732,
      "step": 3
    },
    {
      "epoch": 0.023633677991137372,
      "grad_norm": 34.534549713134766,
      "learning_rate": 4e-05,
      "loss": 5.0456,
      "step": 4
    },
    {
      "epoch": 0.029542097488921712,
      "grad_norm": 50.97783279418945,
      "learning_rate": 5e-05,
      "loss": 4.1734,
      "step": 5
    },
    {
      "epoch": 0.03545051698670606,
      "grad_norm": 25.17313003540039,
      "learning_rate": 6e-05,
      "loss": 4.1785,
      "step": 6
    },
    {
      "epoch": 0.0413589364844904,
      "grad_norm": 34.59897232055664,
      "learning_rate": 7e-05,
      "loss": 4.1085,
      "step": 7
    },
    {
      "epoch": 0.047267355982274745,
      "grad_norm": 18.07676124572754,
      "learning_rate": 8e-05,
      "loss": 3.6911,
      "step": 8
    },
    {
      "epoch": 0.053175775480059084,
      "grad_norm": 15.92068862915039,
      "learning_rate": 9e-05,
      "loss": 3.6311,
      "step": 9
    },
    {
      "epoch": 0.059084194977843424,
      "grad_norm": 12.019160270690918,
      "learning_rate": 0.0001,
      "loss": 3.1151,
      "step": 10
    },
    {
      "epoch": 0.06499261447562776,
      "grad_norm": 13.678496360778809,
      "learning_rate": 9.999316524962345e-05,
      "loss": 3.5686,
      "step": 11
    },
    {
      "epoch": 0.07090103397341212,
      "grad_norm": 12.054726600646973,
      "learning_rate": 9.997266286704631e-05,
      "loss": 3.6136,
      "step": 12
    },
    {
      "epoch": 0.07680945347119646,
      "grad_norm": 13.952569007873535,
      "learning_rate": 9.993849845741524e-05,
      "loss": 3.4011,
      "step": 13
    },
    {
      "epoch": 0.0827178729689808,
      "grad_norm": 12.28878402709961,
      "learning_rate": 9.989068136093873e-05,
      "loss": 3.6625,
      "step": 14
    },
    {
      "epoch": 0.08862629246676514,
      "grad_norm": 11.535065650939941,
      "learning_rate": 9.98292246503335e-05,
      "loss": 3.5663,
      "step": 15
    },
    {
      "epoch": 0.09453471196454949,
      "grad_norm": 12.758620262145996,
      "learning_rate": 9.975414512725057e-05,
      "loss": 3.5768,
      "step": 16
    },
    {
      "epoch": 0.10044313146233383,
      "grad_norm": 12.950556755065918,
      "learning_rate": 9.966546331768191e-05,
      "loss": 3.3219,
      "step": 17
    },
    {
      "epoch": 0.10635155096011817,
      "grad_norm": 11.887937545776367,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.7698,
      "step": 18
    },
    {
      "epoch": 0.11225997045790251,
      "grad_norm": 11.189031600952148,
      "learning_rate": 9.944739353007344e-05,
      "loss": 3.5234,
      "step": 19
    },
    {
      "epoch": 0.11816838995568685,
      "grad_norm": 11.704141616821289,
      "learning_rate": 9.931806517013612e-05,
      "loss": 3.1704,
      "step": 20
    },
    {
      "epoch": 0.1240768094534712,
      "grad_norm": 15.247882843017578,
      "learning_rate": 9.917525374361912e-05,
      "loss": 3.8576,
      "step": 21
    },
    {
      "epoch": 0.12998522895125553,
      "grad_norm": 14.080589294433594,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.871,
      "step": 22
    },
    {
      "epoch": 0.1358936484490399,
      "grad_norm": 11.416792869567871,
      "learning_rate": 9.884934153917997e-05,
      "loss": 3.5375,
      "step": 23
    },
    {
      "epoch": 0.14180206794682423,
      "grad_norm": 11.188799858093262,
      "learning_rate": 9.86663298624003e-05,
      "loss": 3.6145,
      "step": 24
    },
    {
      "epoch": 0.14771048744460857,
      "grad_norm": 15.087061882019043,
      "learning_rate": 9.847001329696653e-05,
      "loss": 4.0115,
      "step": 25
    },
    {
      "epoch": 0.1536189069423929,
      "grad_norm": 11.513084411621094,
      "learning_rate": 9.826044551386744e-05,
      "loss": 3.9049,
      "step": 26
    },
    {
      "epoch": 0.15952732644017725,
      "grad_norm": 12.167125701904297,
      "learning_rate": 9.803768380684242e-05,
      "loss": 4.083,
      "step": 27
    },
    {
      "epoch": 0.1654357459379616,
      "grad_norm": 12.131103515625,
      "learning_rate": 9.780178907671789e-05,
      "loss": 3.6772,
      "step": 28
    },
    {
      "epoch": 0.17134416543574593,
      "grad_norm": 13.083057403564453,
      "learning_rate": 9.755282581475769e-05,
      "loss": 3.8718,
      "step": 29
    },
    {
      "epoch": 0.17725258493353027,
      "grad_norm": 12.414674758911133,
      "learning_rate": 9.729086208503174e-05,
      "loss": 4.1398,
      "step": 30
    },
    {
      "epoch": 0.1831610044313146,
      "grad_norm": 11.669783592224121,
      "learning_rate": 9.701596950580806e-05,
      "loss": 4.2851,
      "step": 31
    },
    {
      "epoch": 0.18906942392909898,
      "grad_norm": 12.696599006652832,
      "learning_rate": 9.672822322997305e-05,
      "loss": 3.9792,
      "step": 32
    },
    {
      "epoch": 0.19497784342688332,
      "grad_norm": 14.1757173538208,
      "learning_rate": 9.642770192448536e-05,
      "loss": 4.8195,
      "step": 33
    },
    {
      "epoch": 0.20088626292466766,
      "grad_norm": 11.965018272399902,
      "learning_rate": 9.611448774886924e-05,
      "loss": 4.372,
      "step": 34
    },
    {
      "epoch": 0.206794682422452,
      "grad_norm": 11.343331336975098,
      "learning_rate": 9.578866633275288e-05,
      "loss": 4.5424,
      "step": 35
    },
    {
      "epoch": 0.21270310192023634,
      "grad_norm": 11.488658905029297,
      "learning_rate": 9.545032675245813e-05,
      "loss": 4.5233,
      "step": 36
    },
    {
      "epoch": 0.21861152141802068,
      "grad_norm": 13.247932434082031,
      "learning_rate": 9.509956150664796e-05,
      "loss": 5.7089,
      "step": 37
    },
    {
      "epoch": 0.22451994091580502,
      "grad_norm": 13.024702072143555,
      "learning_rate": 9.473646649103818e-05,
      "loss": 5.0666,
      "step": 38
    },
    {
      "epoch": 0.23042836041358936,
      "grad_norm": 15.370957374572754,
      "learning_rate": 9.43611409721806e-05,
      "loss": 5.4644,
      "step": 39
    },
    {
      "epoch": 0.2363367799113737,
      "grad_norm": 12.9291353225708,
      "learning_rate": 9.397368756032445e-05,
      "loss": 5.3628,
      "step": 40
    },
    {
      "epoch": 0.24224519940915806,
      "grad_norm": 14.708918571472168,
      "learning_rate": 9.357421218136386e-05,
      "loss": 5.6903,
      "step": 41
    },
    {
      "epoch": 0.2481536189069424,
      "grad_norm": 14.399056434631348,
      "learning_rate": 9.316282404787871e-05,
      "loss": 6.7158,
      "step": 42
    },
    {
      "epoch": 0.25406203840472674,
      "grad_norm": 23.516857147216797,
      "learning_rate": 9.273963562927695e-05,
      "loss": 5.2023,
      "step": 43
    },
    {
      "epoch": 0.25997045790251105,
      "grad_norm": 11.844979286193848,
      "learning_rate": 9.230476262104677e-05,
      "loss": 4.1234,
      "step": 44
    },
    {
      "epoch": 0.2658788774002954,
      "grad_norm": 8.362876892089844,
      "learning_rate": 9.185832391312644e-05,
      "loss": 3.7978,
      "step": 45
    },
    {
      "epoch": 0.2717872968980798,
      "grad_norm": 8.686746597290039,
      "learning_rate": 9.140044155740101e-05,
      "loss": 4.0384,
      "step": 46
    },
    {
      "epoch": 0.2776957163958641,
      "grad_norm": 8.486406326293945,
      "learning_rate": 9.093124073433463e-05,
      "loss": 3.8218,
      "step": 47
    },
    {
      "epoch": 0.28360413589364847,
      "grad_norm": 11.87731647491455,
      "learning_rate": 9.045084971874738e-05,
      "loss": 3.2001,
      "step": 48
    },
    {
      "epoch": 0.2895125553914328,
      "grad_norm": 7.831921100616455,
      "learning_rate": 8.995939984474624e-05,
      "loss": 3.533,
      "step": 49
    },
    {
      "epoch": 0.29542097488921715,
      "grad_norm": 12.459837913513184,
      "learning_rate": 8.945702546981969e-05,
      "loss": 3.2268,
      "step": 50
    },
    {
      "epoch": 0.29542097488921715,
      "eval_loss": 1.0215978622436523,
      "eval_runtime": 21.0397,
      "eval_samples_per_second": 13.593,
      "eval_steps_per_second": 3.422,
      "step": 50
    },
    {
      "epoch": 0.30132939438700146,
      "grad_norm": 7.991581439971924,
      "learning_rate": 8.894386393810563e-05,
      "loss": 3.5145,
      "step": 51
    },
    {
      "epoch": 0.3072378138847858,
      "grad_norm": 8.76706314086914,
      "learning_rate": 8.842005554284296e-05,
      "loss": 3.1821,
      "step": 52
    },
    {
      "epoch": 0.31314623338257014,
      "grad_norm": 8.565008163452148,
      "learning_rate": 8.788574348801675e-05,
      "loss": 3.3311,
      "step": 53
    },
    {
      "epoch": 0.3190546528803545,
      "grad_norm": 8.128891944885254,
      "learning_rate": 8.73410738492077e-05,
      "loss": 3.1362,
      "step": 54
    },
    {
      "epoch": 0.3249630723781389,
      "grad_norm": 11.025174140930176,
      "learning_rate": 8.678619553365659e-05,
      "loss": 4.0493,
      "step": 55
    },
    {
      "epoch": 0.3308714918759232,
      "grad_norm": 9.589738845825195,
      "learning_rate": 8.622126023955446e-05,
      "loss": 3.247,
      "step": 56
    },
    {
      "epoch": 0.33677991137370755,
      "grad_norm": 8.836793899536133,
      "learning_rate": 8.564642241456986e-05,
      "loss": 3.1632,
      "step": 57
    },
    {
      "epoch": 0.34268833087149186,
      "grad_norm": 8.42935848236084,
      "learning_rate": 8.506183921362443e-05,
      "loss": 3.1582,
      "step": 58
    },
    {
      "epoch": 0.34859675036927623,
      "grad_norm": 10.000310897827148,
      "learning_rate": 8.44676704559283e-05,
      "loss": 3.5462,
      "step": 59
    },
    {
      "epoch": 0.35450516986706054,
      "grad_norm": 10.514399528503418,
      "learning_rate": 8.386407858128706e-05,
      "loss": 3.1922,
      "step": 60
    },
    {
      "epoch": 0.3604135893648449,
      "grad_norm": 10.593103408813477,
      "learning_rate": 8.32512286056924e-05,
      "loss": 3.5886,
      "step": 61
    },
    {
      "epoch": 0.3663220088626292,
      "grad_norm": 9.908804893493652,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.6878,
      "step": 62
    },
    {
      "epoch": 0.3722304283604136,
      "grad_norm": 10.04334545135498,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.6867,
      "step": 63
    },
    {
      "epoch": 0.37813884785819796,
      "grad_norm": 9.433609962463379,
      "learning_rate": 8.135881792367686e-05,
      "loss": 3.4514,
      "step": 64
    },
    {
      "epoch": 0.38404726735598227,
      "grad_norm": 9.633214950561523,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.3021,
      "step": 65
    },
    {
      "epoch": 0.38995568685376664,
      "grad_norm": 8.5519380569458,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.9637,
      "step": 66
    },
    {
      "epoch": 0.39586410635155095,
      "grad_norm": 9.883139610290527,
      "learning_rate": 7.938926261462366e-05,
      "loss": 3.2762,
      "step": 67
    },
    {
      "epoch": 0.4017725258493353,
      "grad_norm": 10.432541847229004,
      "learning_rate": 7.871643313414718e-05,
      "loss": 3.594,
      "step": 68
    },
    {
      "epoch": 0.4076809453471196,
      "grad_norm": 14.026384353637695,
      "learning_rate": 7.803575286758364e-05,
      "loss": 3.6544,
      "step": 69
    },
    {
      "epoch": 0.413589364844904,
      "grad_norm": 9.499879837036133,
      "learning_rate": 7.734740790612136e-05,
      "loss": 3.3933,
      "step": 70
    },
    {
      "epoch": 0.4194977843426883,
      "grad_norm": 20.312835693359375,
      "learning_rate": 7.66515864363997e-05,
      "loss": 4.048,
      "step": 71
    },
    {
      "epoch": 0.4254062038404727,
      "grad_norm": 10.350151062011719,
      "learning_rate": 7.594847868906076e-05,
      "loss": 4.5951,
      "step": 72
    },
    {
      "epoch": 0.43131462333825704,
      "grad_norm": 12.654243469238281,
      "learning_rate": 7.52382768867422e-05,
      "loss": 4.1947,
      "step": 73
    },
    {
      "epoch": 0.43722304283604135,
      "grad_norm": 9.018845558166504,
      "learning_rate": 7.452117519152542e-05,
      "loss": 4.2871,
      "step": 74
    },
    {
      "epoch": 0.4431314623338257,
      "grad_norm": 10.344325065612793,
      "learning_rate": 7.379736965185368e-05,
      "loss": 4.3722,
      "step": 75
    },
    {
      "epoch": 0.44903988183161003,
      "grad_norm": 8.817124366760254,
      "learning_rate": 7.30670581489344e-05,
      "loss": 4.0078,
      "step": 76
    },
    {
      "epoch": 0.4549483013293944,
      "grad_norm": 12.50080680847168,
      "learning_rate": 7.233044034264034e-05,
      "loss": 5.2382,
      "step": 77
    },
    {
      "epoch": 0.4608567208271787,
      "grad_norm": 11.088142395019531,
      "learning_rate": 7.158771761692464e-05,
      "loss": 4.6666,
      "step": 78
    },
    {
      "epoch": 0.4667651403249631,
      "grad_norm": 13.323476791381836,
      "learning_rate": 7.083909302476453e-05,
      "loss": 4.9292,
      "step": 79
    },
    {
      "epoch": 0.4726735598227474,
      "grad_norm": 15.717926979064941,
      "learning_rate": 7.008477123264848e-05,
      "loss": 5.3609,
      "step": 80
    },
    {
      "epoch": 0.47858197932053176,
      "grad_norm": 12.110669136047363,
      "learning_rate": 6.932495846462261e-05,
      "loss": 5.6962,
      "step": 81
    },
    {
      "epoch": 0.4844903988183161,
      "grad_norm": 12.197193145751953,
      "learning_rate": 6.855986244591104e-05,
      "loss": 5.6069,
      "step": 82
    },
    {
      "epoch": 0.49039881831610044,
      "grad_norm": 12.179551124572754,
      "learning_rate": 6.778969234612584e-05,
      "loss": 5.885,
      "step": 83
    },
    {
      "epoch": 0.4963072378138848,
      "grad_norm": 16.919248580932617,
      "learning_rate": 6.701465872208216e-05,
      "loss": 7.1328,
      "step": 84
    },
    {
      "epoch": 0.5022156573116692,
      "grad_norm": 8.936378479003906,
      "learning_rate": 6.623497346023418e-05,
      "loss": 3.6559,
      "step": 85
    },
    {
      "epoch": 0.5081240768094535,
      "grad_norm": 9.675037384033203,
      "learning_rate": 6.545084971874738e-05,
      "loss": 4.3146,
      "step": 86
    },
    {
      "epoch": 0.5140324963072378,
      "grad_norm": 9.35242748260498,
      "learning_rate": 6.466250186922325e-05,
      "loss": 3.8763,
      "step": 87
    },
    {
      "epoch": 0.5199409158050221,
      "grad_norm": 9.486969947814941,
      "learning_rate": 6.387014543809223e-05,
      "loss": 4.0821,
      "step": 88
    },
    {
      "epoch": 0.5258493353028065,
      "grad_norm": 7.812458038330078,
      "learning_rate": 6.307399704769099e-05,
      "loss": 3.2371,
      "step": 89
    },
    {
      "epoch": 0.5317577548005908,
      "grad_norm": 8.720259666442871,
      "learning_rate": 6.227427435703997e-05,
      "loss": 3.2058,
      "step": 90
    },
    {
      "epoch": 0.5376661742983752,
      "grad_norm": 9.179726600646973,
      "learning_rate": 6.147119600233758e-05,
      "loss": 3.5405,
      "step": 91
    },
    {
      "epoch": 0.5435745937961596,
      "grad_norm": 7.906421184539795,
      "learning_rate": 6.066498153718735e-05,
      "loss": 3.1573,
      "step": 92
    },
    {
      "epoch": 0.5494830132939439,
      "grad_norm": 10.167905807495117,
      "learning_rate": 5.985585137257401e-05,
      "loss": 3.4864,
      "step": 93
    },
    {
      "epoch": 0.5553914327917282,
      "grad_norm": 9.9895601272583,
      "learning_rate": 5.90440267166055e-05,
      "loss": 3.3057,
      "step": 94
    },
    {
      "epoch": 0.5612998522895125,
      "grad_norm": 9.537314414978027,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 3.3805,
      "step": 95
    },
    {
      "epoch": 0.5672082717872969,
      "grad_norm": 8.18869686126709,
      "learning_rate": 5.74131823855921e-05,
      "loss": 3.0489,
      "step": 96
    },
    {
      "epoch": 0.5731166912850812,
      "grad_norm": 7.7087273597717285,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 2.9601,
      "step": 97
    },
    {
      "epoch": 0.5790251107828656,
      "grad_norm": 8.3488187789917,
      "learning_rate": 5.577423184847932e-05,
      "loss": 3.3352,
      "step": 98
    },
    {
      "epoch": 0.5849335302806499,
      "grad_norm": 8.766341209411621,
      "learning_rate": 5.495227651252315e-05,
      "loss": 3.5692,
      "step": 99
    },
    {
      "epoch": 0.5908419497784343,
      "grad_norm": 8.597002029418945,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 3.1287,
      "step": 100
    },
    {
      "epoch": 0.5908419497784343,
      "eval_loss": 0.9850096702575684,
      "eval_runtime": 21.0719,
      "eval_samples_per_second": 13.573,
      "eval_steps_per_second": 3.417,
      "step": 100
    },
    {
      "epoch": 0.5967503692762186,
      "grad_norm": 7.717814922332764,
      "learning_rate": 5.330452921628497e-05,
      "loss": 3.2411,
      "step": 101
    },
    {
      "epoch": 0.6026587887740029,
      "grad_norm": 9.996946334838867,
      "learning_rate": 5.247918773366112e-05,
      "loss": 3.7527,
      "step": 102
    },
    {
      "epoch": 0.6085672082717873,
      "grad_norm": 8.351297378540039,
      "learning_rate": 5.165316846586541e-05,
      "loss": 3.4355,
      "step": 103
    },
    {
      "epoch": 0.6144756277695717,
      "grad_norm": 7.991336822509766,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 3.1796,
      "step": 104
    },
    {
      "epoch": 0.620384047267356,
      "grad_norm": 9.120264053344727,
      "learning_rate": 5e-05,
      "loss": 4.0159,
      "step": 105
    },
    {
      "epoch": 0.6262924667651403,
      "grad_norm": 8.118267059326172,
      "learning_rate": 4.917330276168208e-05,
      "loss": 3.1347,
      "step": 106
    },
    {
      "epoch": 0.6322008862629247,
      "grad_norm": 9.054883003234863,
      "learning_rate": 4.834683153413459e-05,
      "loss": 3.615,
      "step": 107
    },
    {
      "epoch": 0.638109305760709,
      "grad_norm": 9.330496788024902,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 3.7498,
      "step": 108
    },
    {
      "epoch": 0.6440177252584933,
      "grad_norm": 8.892244338989258,
      "learning_rate": 4.669547078371504e-05,
      "loss": 3.5159,
      "step": 109
    },
    {
      "epoch": 0.6499261447562777,
      "grad_norm": 9.172662734985352,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 3.5029,
      "step": 110
    },
    {
      "epoch": 0.6558345642540621,
      "grad_norm": 9.913756370544434,
      "learning_rate": 4.504772348747687e-05,
      "loss": 4.027,
      "step": 111
    },
    {
      "epoch": 0.6617429837518464,
      "grad_norm": 8.969059944152832,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 3.9222,
      "step": 112
    },
    {
      "epoch": 0.6676514032496307,
      "grad_norm": 10.866827011108398,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 3.9316,
      "step": 113
    },
    {
      "epoch": 0.6735598227474151,
      "grad_norm": 10.216789245605469,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 3.9624,
      "step": 114
    },
    {
      "epoch": 0.6794682422451994,
      "grad_norm": 9.727008819580078,
      "learning_rate": 4.17702704859633e-05,
      "loss": 3.9316,
      "step": 115
    },
    {
      "epoch": 0.6853766617429837,
      "grad_norm": 9.215411186218262,
      "learning_rate": 4.095597328339452e-05,
      "loss": 4.3148,
      "step": 116
    },
    {
      "epoch": 0.691285081240768,
      "grad_norm": 9.693857192993164,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 4.5526,
      "step": 117
    },
    {
      "epoch": 0.6971935007385525,
      "grad_norm": 9.990143775939941,
      "learning_rate": 3.933501846281267e-05,
      "loss": 4.3155,
      "step": 118
    },
    {
      "epoch": 0.7031019202363368,
      "grad_norm": 10.458694458007812,
      "learning_rate": 3.852880399766243e-05,
      "loss": 4.4618,
      "step": 119
    },
    {
      "epoch": 0.7090103397341211,
      "grad_norm": 10.072065353393555,
      "learning_rate": 3.772572564296005e-05,
      "loss": 4.8876,
      "step": 120
    },
    {
      "epoch": 0.7149187592319055,
      "grad_norm": 10.353560447692871,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 4.8817,
      "step": 121
    },
    {
      "epoch": 0.7208271787296898,
      "grad_norm": 10.958467483520508,
      "learning_rate": 3.612985456190778e-05,
      "loss": 4.8819,
      "step": 122
    },
    {
      "epoch": 0.7267355982274741,
      "grad_norm": 11.746684074401855,
      "learning_rate": 3.533749813077677e-05,
      "loss": 5.1958,
      "step": 123
    },
    {
      "epoch": 0.7326440177252584,
      "grad_norm": 12.395604133605957,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 6.0213,
      "step": 124
    },
    {
      "epoch": 0.7385524372230429,
      "grad_norm": 12.523317337036133,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 6.1967,
      "step": 125
    },
    {
      "epoch": 0.7444608567208272,
      "grad_norm": 13.590256690979004,
      "learning_rate": 3.298534127791785e-05,
      "loss": 6.2423,
      "step": 126
    },
    {
      "epoch": 0.7503692762186115,
      "grad_norm": 7.27439022064209,
      "learning_rate": 3.221030765387417e-05,
      "loss": 3.9273,
      "step": 127
    },
    {
      "epoch": 0.7562776957163959,
      "grad_norm": 7.474854469299316,
      "learning_rate": 3.144013755408895e-05,
      "loss": 3.4382,
      "step": 128
    },
    {
      "epoch": 0.7621861152141802,
      "grad_norm": 7.360347270965576,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 3.1805,
      "step": 129
    },
    {
      "epoch": 0.7680945347119645,
      "grad_norm": 7.347513198852539,
      "learning_rate": 2.991522876735154e-05,
      "loss": 3.2443,
      "step": 130
    },
    {
      "epoch": 0.7740029542097489,
      "grad_norm": 8.223475456237793,
      "learning_rate": 2.916090697523549e-05,
      "loss": 3.4577,
      "step": 131
    },
    {
      "epoch": 0.7799113737075333,
      "grad_norm": 7.665858268737793,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 3.3044,
      "step": 132
    },
    {
      "epoch": 0.7858197932053176,
      "grad_norm": 9.941719055175781,
      "learning_rate": 2.766955965735968e-05,
      "loss": 3.4403,
      "step": 133
    },
    {
      "epoch": 0.7917282127031019,
      "grad_norm": 9.350205421447754,
      "learning_rate": 2.693294185106562e-05,
      "loss": 3.4916,
      "step": 134
    },
    {
      "epoch": 0.7976366322008862,
      "grad_norm": 10.959920883178711,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 3.9573,
      "step": 135
    },
    {
      "epoch": 0.8035450516986706,
      "grad_norm": 9.262216567993164,
      "learning_rate": 2.547882480847461e-05,
      "loss": 3.4248,
      "step": 136
    },
    {
      "epoch": 0.8094534711964549,
      "grad_norm": 9.246731758117676,
      "learning_rate": 2.476172311325783e-05,
      "loss": 3.2714,
      "step": 137
    },
    {
      "epoch": 0.8153618906942393,
      "grad_norm": 8.525459289550781,
      "learning_rate": 2.405152131093926e-05,
      "loss": 3.3397,
      "step": 138
    },
    {
      "epoch": 0.8212703101920237,
      "grad_norm": 7.684481620788574,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 3.2098,
      "step": 139
    },
    {
      "epoch": 0.827178729689808,
      "grad_norm": 8.508077621459961,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 3.3596,
      "step": 140
    },
    {
      "epoch": 0.8330871491875923,
      "grad_norm": 8.784626960754395,
      "learning_rate": 2.196424713241637e-05,
      "loss": 3.2147,
      "step": 141
    },
    {
      "epoch": 0.8389955686853766,
      "grad_norm": 9.27164363861084,
      "learning_rate": 2.128356686585282e-05,
      "loss": 3.775,
      "step": 142
    },
    {
      "epoch": 0.844903988183161,
      "grad_norm": 8.394728660583496,
      "learning_rate": 2.061073738537635e-05,
      "loss": 3.3113,
      "step": 143
    },
    {
      "epoch": 0.8508124076809453,
      "grad_norm": 10.800779342651367,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 3.5034,
      "step": 144
    },
    {
      "epoch": 0.8567208271787297,
      "grad_norm": 7.826393127441406,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.8883,
      "step": 145
    },
    {
      "epoch": 0.8626292466765141,
      "grad_norm": 7.5174174308776855,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 3.2072,
      "step": 146
    },
    {
      "epoch": 0.8685376661742984,
      "grad_norm": 8.999770164489746,
      "learning_rate": 1.800157297483417e-05,
      "loss": 3.523,
      "step": 147
    },
    {
      "epoch": 0.8744460856720827,
      "grad_norm": 8.300889015197754,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 3.5789,
      "step": 148
    },
    {
      "epoch": 0.880354505169867,
      "grad_norm": 8.262804985046387,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 3.3697,
      "step": 149
    },
    {
      "epoch": 0.8862629246676514,
      "grad_norm": 7.75667142868042,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 3.104,
      "step": 150
    },
    {
      "epoch": 0.8862629246676514,
      "eval_loss": 0.9513868689537048,
      "eval_runtime": 21.051,
      "eval_samples_per_second": 13.586,
      "eval_steps_per_second": 3.42,
      "step": 150
    },
    {
      "epoch": 0.8921713441654358,
      "grad_norm": 10.413788795471191,
      "learning_rate": 1.553232954407171e-05,
      "loss": 3.8108,
      "step": 151
    },
    {
      "epoch": 0.8980797636632201,
      "grad_norm": 8.449130058288574,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 3.4101,
      "step": 152
    },
    {
      "epoch": 0.9039881831610044,
      "grad_norm": 8.775529861450195,
      "learning_rate": 1.435357758543015e-05,
      "loss": 3.804,
      "step": 153
    },
    {
      "epoch": 0.9098966026587888,
      "grad_norm": 8.538846015930176,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 3.6037,
      "step": 154
    },
    {
      "epoch": 0.9158050221565731,
      "grad_norm": 9.621621131896973,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 4.0096,
      "step": 155
    },
    {
      "epoch": 0.9217134416543574,
      "grad_norm": 9.71113109588623,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 4.2912,
      "step": 156
    },
    {
      "epoch": 0.9276218611521418,
      "grad_norm": 8.819686889648438,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 3.9396,
      "step": 157
    },
    {
      "epoch": 0.9335302806499262,
      "grad_norm": 9.197346687316895,
      "learning_rate": 1.157994445715706e-05,
      "loss": 4.073,
      "step": 158
    },
    {
      "epoch": 0.9394387001477105,
      "grad_norm": 10.014103889465332,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 3.7543,
      "step": 159
    },
    {
      "epoch": 0.9453471196454948,
      "grad_norm": 10.633454322814941,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 3.9223,
      "step": 160
    },
    {
      "epoch": 0.9512555391432792,
      "grad_norm": 10.449409484863281,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 4.5343,
      "step": 161
    },
    {
      "epoch": 0.9571639586410635,
      "grad_norm": 11.074226379394531,
      "learning_rate": 9.549150281252633e-06,
      "loss": 4.9203,
      "step": 162
    },
    {
      "epoch": 0.9630723781388478,
      "grad_norm": 12.032450675964355,
      "learning_rate": 9.068759265665384e-06,
      "loss": 5.7197,
      "step": 163
    },
    {
      "epoch": 0.9689807976366323,
      "grad_norm": 9.91462230682373,
      "learning_rate": 8.599558442598998e-06,
      "loss": 4.7608,
      "step": 164
    },
    {
      "epoch": 0.9748892171344166,
      "grad_norm": 12.094534873962402,
      "learning_rate": 8.141676086873572e-06,
      "loss": 5.7761,
      "step": 165
    },
    {
      "epoch": 0.9807976366322009,
      "grad_norm": 12.708998680114746,
      "learning_rate": 7.695237378953223e-06,
      "loss": 6.1031,
      "step": 166
    },
    {
      "epoch": 0.9867060561299852,
      "grad_norm": 12.794193267822266,
      "learning_rate": 7.260364370723044e-06,
      "loss": 5.4127,
      "step": 167
    },
    {
      "epoch": 0.9926144756277696,
      "grad_norm": 15.079773902893066,
      "learning_rate": 6.837175952121306e-06,
      "loss": 6.2951,
      "step": 168
    },
    {
      "epoch": 0.9985228951255539,
      "grad_norm": 8.368268013000488,
      "learning_rate": 6.425787818636131e-06,
      "loss": 3.4468,
      "step": 169
    },
    {
      "epoch": 1.0044313146233383,
      "grad_norm": 7.612218856811523,
      "learning_rate": 6.026312439675552e-06,
      "loss": 3.2699,
      "step": 170
    },
    {
      "epoch": 1.0103397341211227,
      "grad_norm": 6.2261128425598145,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 2.5886,
      "step": 171
    },
    {
      "epoch": 1.016248153618907,
      "grad_norm": 6.192266941070557,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.3112,
      "step": 172
    },
    {
      "epoch": 1.0221565731166913,
      "grad_norm": 6.399194717407227,
      "learning_rate": 4.900438493352055e-06,
      "loss": 2.2511,
      "step": 173
    },
    {
      "epoch": 1.0280649926144756,
      "grad_norm": 6.307614803314209,
      "learning_rate": 4.549673247541875e-06,
      "loss": 1.9278,
      "step": 174
    },
    {
      "epoch": 1.03397341211226,
      "grad_norm": 5.532915115356445,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 1.985,
      "step": 175
    },
    {
      "epoch": 1.0398818316100442,
      "grad_norm": 6.029874324798584,
      "learning_rate": 3.885512251130763e-06,
      "loss": 1.7499,
      "step": 176
    },
    {
      "epoch": 1.0457902511078287,
      "grad_norm": 5.670559883117676,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 1.8178,
      "step": 177
    },
    {
      "epoch": 1.051698670605613,
      "grad_norm": 6.167243480682373,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.0016,
      "step": 178
    },
    {
      "epoch": 1.0576070901033974,
      "grad_norm": 6.098389625549316,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 1.7977,
      "step": 179
    },
    {
      "epoch": 1.0635155096011817,
      "grad_norm": 5.410711288452148,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 1.6749,
      "step": 180
    },
    {
      "epoch": 1.069423929098966,
      "grad_norm": 5.890036106109619,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 1.8947,
      "step": 181
    },
    {
      "epoch": 1.0753323485967503,
      "grad_norm": 5.733892917633057,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 1.7251,
      "step": 182
    },
    {
      "epoch": 1.0812407680945346,
      "grad_norm": 5.511055946350098,
      "learning_rate": 1.962316193157593e-06,
      "loss": 1.7154,
      "step": 183
    },
    {
      "epoch": 1.0871491875923192,
      "grad_norm": 6.739435195922852,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 1.8798,
      "step": 184
    },
    {
      "epoch": 1.0930576070901035,
      "grad_norm": 6.21235466003418,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 1.6744,
      "step": 185
    },
    {
      "epoch": 1.0989660265878878,
      "grad_norm": 6.193417072296143,
      "learning_rate": 1.333670137599713e-06,
      "loss": 1.8161,
      "step": 186
    },
    {
      "epoch": 1.104874446085672,
      "grad_norm": 6.969656944274902,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.2349,
      "step": 187
    },
    {
      "epoch": 1.1107828655834564,
      "grad_norm": 5.748712539672852,
      "learning_rate": 9.810017062595322e-07,
      "loss": 1.9601,
      "step": 188
    },
    {
      "epoch": 1.1166912850812407,
      "grad_norm": 10.47679328918457,
      "learning_rate": 8.247462563808817e-07,
      "loss": 1.8083,
      "step": 189
    },
    {
      "epoch": 1.122599704579025,
      "grad_norm": 5.984971523284912,
      "learning_rate": 6.819348298638839e-07,
      "loss": 1.809,
      "step": 190
    },
    {
      "epoch": 1.1285081240768093,
      "grad_norm": 6.719799041748047,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.5313,
      "step": 191
    },
    {
      "epoch": 1.1344165435745939,
      "grad_norm": 5.560427665710449,
      "learning_rate": 4.367965336512403e-07,
      "loss": 1.7861,
      "step": 192
    },
    {
      "epoch": 1.1403249630723782,
      "grad_norm": 5.918384552001953,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.2358,
      "step": 193
    },
    {
      "epoch": 1.1462333825701625,
      "grad_norm": 6.6042914390563965,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.0036,
      "step": 194
    },
    {
      "epoch": 1.1521418020679468,
      "grad_norm": 6.131277084350586,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.082,
      "step": 195
    },
    {
      "epoch": 1.1580502215657311,
      "grad_norm": 6.451679706573486,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.4433,
      "step": 196
    },
    {
      "epoch": 1.1639586410635154,
      "grad_norm": 6.990377426147461,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.4368,
      "step": 197
    },
    {
      "epoch": 1.1698670605613,
      "grad_norm": 7.195181846618652,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 2.7055,
      "step": 198
    },
    {
      "epoch": 1.1757754800590843,
      "grad_norm": 6.5114288330078125,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.2431,
      "step": 199
    },
    {
      "epoch": 1.1816838995568686,
      "grad_norm": 7.174546241760254,
      "learning_rate": 0.0,
      "loss": 2.5577,
      "step": 200
    },
    {
      "epoch": 1.1816838995568686,
      "eval_loss": 0.9565478563308716,
      "eval_runtime": 21.0392,
      "eval_samples_per_second": 13.594,
      "eval_steps_per_second": 3.422,
      "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": 1
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.8616003626519757e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}