{
  "best_metric": 2.509678840637207,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.13717421124828533,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006858710562414266,
      "grad_norm": 68.98727416992188,
      "learning_rate": 1e-05,
      "loss": 12.5759,
      "step": 1
    },
    {
      "epoch": 0.0006858710562414266,
      "eval_loss": 4.443713188171387,
      "eval_runtime": 27.8601,
      "eval_samples_per_second": 88.155,
      "eval_steps_per_second": 22.039,
      "step": 1
    },
    {
      "epoch": 0.0013717421124828531,
      "grad_norm": 69.77474975585938,
      "learning_rate": 2e-05,
      "loss": 13.0793,
      "step": 2
    },
    {
      "epoch": 0.00205761316872428,
      "grad_norm": 63.65254211425781,
      "learning_rate": 3e-05,
      "loss": 13.3288,
      "step": 3
    },
    {
      "epoch": 0.0027434842249657062,
      "grad_norm": 71.66799926757812,
      "learning_rate": 4e-05,
      "loss": 13.4134,
      "step": 4
    },
    {
      "epoch": 0.003429355281207133,
      "grad_norm": 66.94480895996094,
      "learning_rate": 5e-05,
      "loss": 13.0442,
      "step": 5
    },
    {
      "epoch": 0.00411522633744856,
      "grad_norm": 67.95450592041016,
      "learning_rate": 6e-05,
      "loss": 12.2776,
      "step": 6
    },
    {
      "epoch": 0.004801097393689987,
      "grad_norm": 66.10748291015625,
      "learning_rate": 7e-05,
      "loss": 12.905,
      "step": 7
    },
    {
      "epoch": 0.0054869684499314125,
      "grad_norm": 72.70169067382812,
      "learning_rate": 8e-05,
      "loss": 13.1853,
      "step": 8
    },
    {
      "epoch": 0.006172839506172839,
      "grad_norm": 69.311279296875,
      "learning_rate": 9e-05,
      "loss": 12.9915,
      "step": 9
    },
    {
      "epoch": 0.006858710562414266,
      "grad_norm": 78.70733642578125,
      "learning_rate": 0.0001,
      "loss": 12.1994,
      "step": 10
    },
    {
      "epoch": 0.0075445816186556925,
      "grad_norm": 69.87124633789062,
      "learning_rate": 9.999316524962345e-05,
      "loss": 12.1302,
      "step": 11
    },
    {
      "epoch": 0.00823045267489712,
      "grad_norm": 67.85890197753906,
      "learning_rate": 9.997266286704631e-05,
      "loss": 12.686,
      "step": 12
    },
    {
      "epoch": 0.008916323731138546,
      "grad_norm": 62.897117614746094,
      "learning_rate": 9.993849845741524e-05,
      "loss": 12.1832,
      "step": 13
    },
    {
      "epoch": 0.009602194787379973,
      "grad_norm": 54.73257064819336,
      "learning_rate": 9.989068136093873e-05,
      "loss": 12.493,
      "step": 14
    },
    {
      "epoch": 0.0102880658436214,
      "grad_norm": 56.899253845214844,
      "learning_rate": 9.98292246503335e-05,
      "loss": 12.3362,
      "step": 15
    },
    {
      "epoch": 0.010973936899862825,
      "grad_norm": 48.628639221191406,
      "learning_rate": 9.975414512725057e-05,
      "loss": 12.4438,
      "step": 16
    },
    {
      "epoch": 0.011659807956104253,
      "grad_norm": 42.66781234741211,
      "learning_rate": 9.966546331768191e-05,
      "loss": 12.0002,
      "step": 17
    },
    {
      "epoch": 0.012345679012345678,
      "grad_norm": 44.45660400390625,
      "learning_rate": 9.956320346634876e-05,
      "loss": 12.2869,
      "step": 18
    },
    {
      "epoch": 0.013031550068587106,
      "grad_norm": 42.242130279541016,
      "learning_rate": 9.944739353007344e-05,
      "loss": 12.4483,
      "step": 19
    },
    {
      "epoch": 0.013717421124828532,
      "grad_norm": 57.27316665649414,
      "learning_rate": 9.931806517013612e-05,
      "loss": 12.6701,
      "step": 20
    },
    {
      "epoch": 0.01440329218106996,
      "grad_norm": 52.8787841796875,
      "learning_rate": 9.917525374361912e-05,
      "loss": 11.8724,
      "step": 21
    },
    {
      "epoch": 0.015089163237311385,
      "grad_norm": 60.196449279785156,
      "learning_rate": 9.901899829374047e-05,
      "loss": 12.1529,
      "step": 22
    },
    {
      "epoch": 0.01577503429355281,
      "grad_norm": 74.11493682861328,
      "learning_rate": 9.884934153917997e-05,
      "loss": 12.1641,
      "step": 23
    },
    {
      "epoch": 0.01646090534979424,
      "grad_norm": 64.55845642089844,
      "learning_rate": 9.86663298624003e-05,
      "loss": 12.5898,
      "step": 24
    },
    {
      "epoch": 0.017146776406035666,
      "grad_norm": 62.46492385864258,
      "learning_rate": 9.847001329696653e-05,
      "loss": 11.8134,
      "step": 25
    },
    {
      "epoch": 0.01783264746227709,
      "grad_norm": 72.96575164794922,
      "learning_rate": 9.826044551386744e-05,
      "loss": 11.6495,
      "step": 26
    },
    {
      "epoch": 0.018518518518518517,
      "grad_norm": 92.07255554199219,
      "learning_rate": 9.803768380684242e-05,
      "loss": 11.5376,
      "step": 27
    },
    {
      "epoch": 0.019204389574759947,
      "grad_norm": 63.74907302856445,
      "learning_rate": 9.780178907671789e-05,
      "loss": 11.6011,
      "step": 28
    },
    {
      "epoch": 0.019890260631001373,
      "grad_norm": 85.46188354492188,
      "learning_rate": 9.755282581475769e-05,
      "loss": 12.1512,
      "step": 29
    },
    {
      "epoch": 0.0205761316872428,
      "grad_norm": 65.44583892822266,
      "learning_rate": 9.729086208503174e-05,
      "loss": 11.8294,
      "step": 30
    },
    {
      "epoch": 0.021262002743484224,
      "grad_norm": 76.8721694946289,
      "learning_rate": 9.701596950580806e-05,
      "loss": 11.5529,
      "step": 31
    },
    {
      "epoch": 0.02194787379972565,
      "grad_norm": 75.14340209960938,
      "learning_rate": 9.672822322997305e-05,
      "loss": 11.1312,
      "step": 32
    },
    {
      "epoch": 0.02263374485596708,
      "grad_norm": 55.78513717651367,
      "learning_rate": 9.642770192448536e-05,
      "loss": 12.0024,
      "step": 33
    },
    {
      "epoch": 0.023319615912208505,
      "grad_norm": 66.52703094482422,
      "learning_rate": 9.611448774886924e-05,
      "loss": 11.6276,
      "step": 34
    },
    {
      "epoch": 0.02400548696844993,
      "grad_norm": 74.23927307128906,
      "learning_rate": 9.578866633275288e-05,
      "loss": 10.9297,
      "step": 35
    },
    {
      "epoch": 0.024691358024691357,
      "grad_norm": 82.38177490234375,
      "learning_rate": 9.545032675245813e-05,
      "loss": 11.1206,
      "step": 36
    },
    {
      "epoch": 0.025377229080932786,
      "grad_norm": 106.8994369506836,
      "learning_rate": 9.509956150664796e-05,
      "loss": 12.6163,
      "step": 37
    },
    {
      "epoch": 0.02606310013717421,
      "grad_norm": 85.90508270263672,
      "learning_rate": 9.473646649103818e-05,
      "loss": 11.6453,
      "step": 38
    },
    {
      "epoch": 0.026748971193415638,
      "grad_norm": 78.86741638183594,
      "learning_rate": 9.43611409721806e-05,
      "loss": 11.6472,
      "step": 39
    },
    {
      "epoch": 0.027434842249657063,
      "grad_norm": 83.07723236083984,
      "learning_rate": 9.397368756032445e-05,
      "loss": 12.1297,
      "step": 40
    },
    {
      "epoch": 0.028120713305898493,
      "grad_norm": 80.04178619384766,
      "learning_rate": 9.357421218136386e-05,
      "loss": 10.9839,
      "step": 41
    },
    {
      "epoch": 0.02880658436213992,
      "grad_norm": 76.00270080566406,
      "learning_rate": 9.316282404787871e-05,
      "loss": 10.4404,
      "step": 42
    },
    {
      "epoch": 0.029492455418381344,
      "grad_norm": 88.82781219482422,
      "learning_rate": 9.273963562927695e-05,
      "loss": 11.2721,
      "step": 43
    },
    {
      "epoch": 0.03017832647462277,
      "grad_norm": 71.48707580566406,
      "learning_rate": 9.230476262104677e-05,
      "loss": 11.7011,
      "step": 44
    },
    {
      "epoch": 0.030864197530864196,
      "grad_norm": 70.83800506591797,
      "learning_rate": 9.185832391312644e-05,
      "loss": 11.4524,
      "step": 45
    },
    {
      "epoch": 0.03155006858710562,
      "grad_norm": 69.23047637939453,
      "learning_rate": 9.140044155740101e-05,
      "loss": 10.0453,
      "step": 46
    },
    {
      "epoch": 0.03223593964334705,
      "grad_norm": 151.666015625,
      "learning_rate": 9.093124073433463e-05,
      "loss": 12.4546,
      "step": 47
    },
    {
      "epoch": 0.03292181069958848,
      "grad_norm": 65.79100799560547,
      "learning_rate": 9.045084971874738e-05,
      "loss": 11.5173,
      "step": 48
    },
    {
      "epoch": 0.033607681755829906,
      "grad_norm": 162.29966735839844,
      "learning_rate": 8.995939984474624e-05,
      "loss": 11.8805,
      "step": 49
    },
    {
      "epoch": 0.03429355281207133,
      "grad_norm": 148.68655395507812,
      "learning_rate": 8.945702546981969e-05,
      "loss": 14.867,
      "step": 50
    },
    {
      "epoch": 0.03429355281207133,
      "eval_loss": 2.7874114513397217,
      "eval_runtime": 27.7452,
      "eval_samples_per_second": 88.52,
      "eval_steps_per_second": 22.13,
      "step": 50
    },
    {
      "epoch": 0.03497942386831276,
      "grad_norm": 63.49116516113281,
      "learning_rate": 8.894386393810563e-05,
      "loss": 11.151,
      "step": 51
    },
    {
      "epoch": 0.03566529492455418,
      "grad_norm": 66.43113708496094,
      "learning_rate": 8.842005554284296e-05,
      "loss": 10.825,
      "step": 52
    },
    {
      "epoch": 0.03635116598079561,
      "grad_norm": 81.4939193725586,
      "learning_rate": 8.788574348801675e-05,
      "loss": 11.1428,
      "step": 53
    },
    {
      "epoch": 0.037037037037037035,
      "grad_norm": 74.94922637939453,
      "learning_rate": 8.73410738492077e-05,
      "loss": 10.8221,
      "step": 54
    },
    {
      "epoch": 0.03772290809327846,
      "grad_norm": 72.17744445800781,
      "learning_rate": 8.678619553365659e-05,
      "loss": 10.6662,
      "step": 55
    },
    {
      "epoch": 0.038408779149519894,
      "grad_norm": 74.77362823486328,
      "learning_rate": 8.622126023955446e-05,
      "loss": 10.4566,
      "step": 56
    },
    {
      "epoch": 0.03909465020576132,
      "grad_norm": 95.30804443359375,
      "learning_rate": 8.564642241456986e-05,
      "loss": 10.0424,
      "step": 57
    },
    {
      "epoch": 0.039780521262002745,
      "grad_norm": 75.75899505615234,
      "learning_rate": 8.506183921362443e-05,
      "loss": 10.4446,
      "step": 58
    },
    {
      "epoch": 0.04046639231824417,
      "grad_norm": 71.17166137695312,
      "learning_rate": 8.44676704559283e-05,
      "loss": 10.9478,
      "step": 59
    },
    {
      "epoch": 0.0411522633744856,
      "grad_norm": 91.02027130126953,
      "learning_rate": 8.386407858128706e-05,
      "loss": 10.3997,
      "step": 60
    },
    {
      "epoch": 0.04183813443072702,
      "grad_norm": 99.5153579711914,
      "learning_rate": 8.32512286056924e-05,
      "loss": 10.4099,
      "step": 61
    },
    {
      "epoch": 0.04252400548696845,
      "grad_norm": 78.62198638916016,
      "learning_rate": 8.262928807620843e-05,
      "loss": 10.6524,
      "step": 62
    },
    {
      "epoch": 0.043209876543209874,
      "grad_norm": 97.66352844238281,
      "learning_rate": 8.199842702516583e-05,
      "loss": 10.7963,
      "step": 63
    },
    {
      "epoch": 0.0438957475994513,
      "grad_norm": 97.64257049560547,
      "learning_rate": 8.135881792367686e-05,
      "loss": 11.2776,
      "step": 64
    },
    {
      "epoch": 0.04458161865569273,
      "grad_norm": 84.81942749023438,
      "learning_rate": 8.07106356344834e-05,
      "loss": 10.3551,
      "step": 65
    },
    {
      "epoch": 0.04526748971193416,
      "grad_norm": 78.4273681640625,
      "learning_rate": 8.005405736415126e-05,
      "loss": 10.1656,
      "step": 66
    },
    {
      "epoch": 0.045953360768175584,
      "grad_norm": 68.39305877685547,
      "learning_rate": 7.938926261462366e-05,
      "loss": 9.9817,
      "step": 67
    },
    {
      "epoch": 0.04663923182441701,
      "grad_norm": 89.85513305664062,
      "learning_rate": 7.871643313414718e-05,
      "loss": 10.3399,
      "step": 68
    },
    {
      "epoch": 0.047325102880658436,
      "grad_norm": 95.19136047363281,
      "learning_rate": 7.803575286758364e-05,
      "loss": 10.0599,
      "step": 69
    },
    {
      "epoch": 0.04801097393689986,
      "grad_norm": 112.87248992919922,
      "learning_rate": 7.734740790612136e-05,
      "loss": 10.3264,
      "step": 70
    },
    {
      "epoch": 0.04869684499314129,
      "grad_norm": 86.0616683959961,
      "learning_rate": 7.66515864363997e-05,
      "loss": 10.2898,
      "step": 71
    },
    {
      "epoch": 0.04938271604938271,
      "grad_norm": 81.79177856445312,
      "learning_rate": 7.594847868906076e-05,
      "loss": 10.8851,
      "step": 72
    },
    {
      "epoch": 0.050068587105624146,
      "grad_norm": 79.04595947265625,
      "learning_rate": 7.52382768867422e-05,
      "loss": 10.6459,
      "step": 73
    },
    {
      "epoch": 0.05075445816186557,
      "grad_norm": 103.68228149414062,
      "learning_rate": 7.452117519152542e-05,
      "loss": 10.215,
      "step": 74
    },
    {
      "epoch": 0.051440329218107,
      "grad_norm": 115.65605163574219,
      "learning_rate": 7.379736965185368e-05,
      "loss": 11.0884,
      "step": 75
    },
    {
      "epoch": 0.05212620027434842,
      "grad_norm": 129.08201599121094,
      "learning_rate": 7.30670581489344e-05,
      "loss": 10.3404,
      "step": 76
    },
    {
      "epoch": 0.05281207133058985,
      "grad_norm": 145.67327880859375,
      "learning_rate": 7.233044034264034e-05,
      "loss": 10.08,
      "step": 77
    },
    {
      "epoch": 0.053497942386831275,
      "grad_norm": 120.04740905761719,
      "learning_rate": 7.158771761692464e-05,
      "loss": 10.8512,
      "step": 78
    },
    {
      "epoch": 0.0541838134430727,
      "grad_norm": 154.80589294433594,
      "learning_rate": 7.083909302476453e-05,
      "loss": 10.9524,
      "step": 79
    },
    {
      "epoch": 0.05486968449931413,
      "grad_norm": 175.4161376953125,
      "learning_rate": 7.008477123264848e-05,
      "loss": 10.1077,
      "step": 80
    },
    {
      "epoch": 0.05555555555555555,
      "grad_norm": 159.41204833984375,
      "learning_rate": 6.932495846462261e-05,
      "loss": 10.1778,
      "step": 81
    },
    {
      "epoch": 0.056241426611796985,
      "grad_norm": 166.1090545654297,
      "learning_rate": 6.855986244591104e-05,
      "loss": 11.6773,
      "step": 82
    },
    {
      "epoch": 0.05692729766803841,
      "grad_norm": 230.90863037109375,
      "learning_rate": 6.778969234612584e-05,
      "loss": 10.9128,
      "step": 83
    },
    {
      "epoch": 0.05761316872427984,
      "grad_norm": 274.4411926269531,
      "learning_rate": 6.701465872208216e-05,
      "loss": 11.7554,
      "step": 84
    },
    {
      "epoch": 0.05829903978052126,
      "grad_norm": 464.356201171875,
      "learning_rate": 6.623497346023418e-05,
      "loss": 10.8449,
      "step": 85
    },
    {
      "epoch": 0.05898491083676269,
      "grad_norm": 213.29855346679688,
      "learning_rate": 6.545084971874738e-05,
      "loss": 11.4756,
      "step": 86
    },
    {
      "epoch": 0.059670781893004114,
      "grad_norm": 175.677978515625,
      "learning_rate": 6.466250186922325e-05,
      "loss": 10.1132,
      "step": 87
    },
    {
      "epoch": 0.06035665294924554,
      "grad_norm": 253.1256866455078,
      "learning_rate": 6.387014543809223e-05,
      "loss": 10.1074,
      "step": 88
    },
    {
      "epoch": 0.061042524005486966,
      "grad_norm": 177.0943603515625,
      "learning_rate": 6.307399704769099e-05,
      "loss": 11.273,
      "step": 89
    },
    {
      "epoch": 0.06172839506172839,
      "grad_norm": 188.79380798339844,
      "learning_rate": 6.227427435703997e-05,
      "loss": 10.4617,
      "step": 90
    },
    {
      "epoch": 0.062414266117969824,
      "grad_norm": 289.6875915527344,
      "learning_rate": 6.147119600233758e-05,
      "loss": 11.3646,
      "step": 91
    },
    {
      "epoch": 0.06310013717421124,
      "grad_norm": 157.2250518798828,
      "learning_rate": 6.066498153718735e-05,
      "loss": 9.5512,
      "step": 92
    },
    {
      "epoch": 0.06378600823045268,
      "grad_norm": 164.64764404296875,
      "learning_rate": 5.985585137257401e-05,
      "loss": 10.1047,
      "step": 93
    },
    {
      "epoch": 0.0644718792866941,
      "grad_norm": 159.32102966308594,
      "learning_rate": 5.90440267166055e-05,
      "loss": 11.0697,
      "step": 94
    },
    {
      "epoch": 0.06515775034293553,
      "grad_norm": 152.89402770996094,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 11.2579,
      "step": 95
    },
    {
      "epoch": 0.06584362139917696,
      "grad_norm": 142.66720581054688,
      "learning_rate": 5.74131823855921e-05,
      "loss": 11.081,
      "step": 96
    },
    {
      "epoch": 0.06652949245541838,
      "grad_norm": 159.0377197265625,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 11.5092,
      "step": 97
    },
    {
      "epoch": 0.06721536351165981,
      "grad_norm": 151.3229217529297,
      "learning_rate": 5.577423184847932e-05,
      "loss": 11.6836,
      "step": 98
    },
    {
      "epoch": 0.06790123456790123,
      "grad_norm": 194.11734008789062,
      "learning_rate": 5.495227651252315e-05,
      "loss": 11.8941,
      "step": 99
    },
    {
      "epoch": 0.06858710562414266,
      "grad_norm": 123.63009643554688,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 13.6284,
      "step": 100
    },
    {
      "epoch": 0.06858710562414266,
      "eval_loss": 2.7800729274749756,
      "eval_runtime": 27.8477,
      "eval_samples_per_second": 88.194,
      "eval_steps_per_second": 22.049,
      "step": 100
    },
    {
      "epoch": 0.06927297668038408,
      "grad_norm": 144.32546997070312,
      "learning_rate": 5.330452921628497e-05,
      "loss": 11.0133,
      "step": 101
    },
    {
      "epoch": 0.06995884773662552,
      "grad_norm": 127.10771179199219,
      "learning_rate": 5.247918773366112e-05,
      "loss": 10.5555,
      "step": 102
    },
    {
      "epoch": 0.07064471879286695,
      "grad_norm": 140.34780883789062,
      "learning_rate": 5.165316846586541e-05,
      "loss": 10.9478,
      "step": 103
    },
    {
      "epoch": 0.07133058984910837,
      "grad_norm": 124.15359497070312,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 10.733,
      "step": 104
    },
    {
      "epoch": 0.0720164609053498,
      "grad_norm": 130.2575225830078,
      "learning_rate": 5e-05,
      "loss": 11.1286,
      "step": 105
    },
    {
      "epoch": 0.07270233196159122,
      "grad_norm": 107.8804931640625,
      "learning_rate": 4.917330276168208e-05,
      "loss": 10.7003,
      "step": 106
    },
    {
      "epoch": 0.07338820301783265,
      "grad_norm": 88.30643463134766,
      "learning_rate": 4.834683153413459e-05,
      "loss": 10.4532,
      "step": 107
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 160.2826690673828,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 10.792,
      "step": 108
    },
    {
      "epoch": 0.0747599451303155,
      "grad_norm": 105.60736846923828,
      "learning_rate": 4.669547078371504e-05,
      "loss": 10.7673,
      "step": 109
    },
    {
      "epoch": 0.07544581618655692,
      "grad_norm": 108.36888122558594,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 10.9557,
      "step": 110
    },
    {
      "epoch": 0.07613168724279835,
      "grad_norm": 119.8900146484375,
      "learning_rate": 4.504772348747687e-05,
      "loss": 10.3313,
      "step": 111
    },
    {
      "epoch": 0.07681755829903979,
      "grad_norm": 105.52911376953125,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 10.5599,
      "step": 112
    },
    {
      "epoch": 0.0775034293552812,
      "grad_norm": 98.0126724243164,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 10.5653,
      "step": 113
    },
    {
      "epoch": 0.07818930041152264,
      "grad_norm": 106.5774154663086,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 11.0501,
      "step": 114
    },
    {
      "epoch": 0.07887517146776406,
      "grad_norm": 88.40042877197266,
      "learning_rate": 4.17702704859633e-05,
      "loss": 10.3608,
      "step": 115
    },
    {
      "epoch": 0.07956104252400549,
      "grad_norm": 89.58585357666016,
      "learning_rate": 4.095597328339452e-05,
      "loss": 10.4017,
      "step": 116
    },
    {
      "epoch": 0.08024691358024691,
      "grad_norm": 87.48372650146484,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 10.832,
      "step": 117
    },
    {
      "epoch": 0.08093278463648834,
      "grad_norm": 94.73457336425781,
      "learning_rate": 3.933501846281267e-05,
      "loss": 10.1883,
      "step": 118
    },
    {
      "epoch": 0.08161865569272976,
      "grad_norm": 83.19349670410156,
      "learning_rate": 3.852880399766243e-05,
      "loss": 10.3721,
      "step": 119
    },
    {
      "epoch": 0.0823045267489712,
      "grad_norm": 76.60916137695312,
      "learning_rate": 3.772572564296005e-05,
      "loss": 9.8807,
      "step": 120
    },
    {
      "epoch": 0.08299039780521263,
      "grad_norm": 82.24100494384766,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 9.8639,
      "step": 121
    },
    {
      "epoch": 0.08367626886145405,
      "grad_norm": 90.17107391357422,
      "learning_rate": 3.612985456190778e-05,
      "loss": 10.9134,
      "step": 122
    },
    {
      "epoch": 0.08436213991769548,
      "grad_norm": 82.33057403564453,
      "learning_rate": 3.533749813077677e-05,
      "loss": 10.452,
      "step": 123
    },
    {
      "epoch": 0.0850480109739369,
      "grad_norm": 70.53192138671875,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 10.7225,
      "step": 124
    },
    {
      "epoch": 0.08573388203017833,
      "grad_norm": 59.174747467041016,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 10.7619,
      "step": 125
    },
    {
      "epoch": 0.08641975308641975,
      "grad_norm": 72.57215118408203,
      "learning_rate": 3.298534127791785e-05,
      "loss": 10.1246,
      "step": 126
    },
    {
      "epoch": 0.08710562414266118,
      "grad_norm": 83.04279327392578,
      "learning_rate": 3.221030765387417e-05,
      "loss": 10.1285,
      "step": 127
    },
    {
      "epoch": 0.0877914951989026,
      "grad_norm": 62.89213180541992,
      "learning_rate": 3.144013755408895e-05,
      "loss": 10.2429,
      "step": 128
    },
    {
      "epoch": 0.08847736625514403,
      "grad_norm": 60.5438117980957,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 10.1504,
      "step": 129
    },
    {
      "epoch": 0.08916323731138547,
      "grad_norm": 69.0409927368164,
      "learning_rate": 2.991522876735154e-05,
      "loss": 9.4541,
      "step": 130
    },
    {
      "epoch": 0.08984910836762688,
      "grad_norm": 70.55886840820312,
      "learning_rate": 2.916090697523549e-05,
      "loss": 10.9784,
      "step": 131
    },
    {
      "epoch": 0.09053497942386832,
      "grad_norm": 92.44700622558594,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 10.361,
      "step": 132
    },
    {
      "epoch": 0.09122085048010974,
      "grad_norm": 80.08354949951172,
      "learning_rate": 2.766955965735968e-05,
      "loss": 10.7984,
      "step": 133
    },
    {
      "epoch": 0.09190672153635117,
      "grad_norm": 75.04423522949219,
      "learning_rate": 2.693294185106562e-05,
      "loss": 11.4142,
      "step": 134
    },
    {
      "epoch": 0.09259259259259259,
      "grad_norm": 73.25611877441406,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 10.552,
      "step": 135
    },
    {
      "epoch": 0.09327846364883402,
      "grad_norm": 62.592254638671875,
      "learning_rate": 2.547882480847461e-05,
      "loss": 9.9753,
      "step": 136
    },
    {
      "epoch": 0.09396433470507544,
      "grad_norm": 55.405303955078125,
      "learning_rate": 2.476172311325783e-05,
      "loss": 10.1547,
      "step": 137
    },
    {
      "epoch": 0.09465020576131687,
      "grad_norm": 72.78160858154297,
      "learning_rate": 2.405152131093926e-05,
      "loss": 11.4672,
      "step": 138
    },
    {
      "epoch": 0.0953360768175583,
      "grad_norm": 73.35169982910156,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 10.5994,
      "step": 139
    },
    {
      "epoch": 0.09602194787379972,
      "grad_norm": 72.02860260009766,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 11.1734,
      "step": 140
    },
    {
      "epoch": 0.09670781893004116,
      "grad_norm": 54.87779998779297,
      "learning_rate": 2.196424713241637e-05,
      "loss": 9.7345,
      "step": 141
    },
    {
      "epoch": 0.09739368998628258,
      "grad_norm": 68.5535888671875,
      "learning_rate": 2.128356686585282e-05,
      "loss": 8.9967,
      "step": 142
    },
    {
      "epoch": 0.09807956104252401,
      "grad_norm": 69.20039367675781,
      "learning_rate": 2.061073738537635e-05,
      "loss": 10.2656,
      "step": 143
    },
    {
      "epoch": 0.09876543209876543,
      "grad_norm": 55.695465087890625,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 9.837,
      "step": 144
    },
    {
      "epoch": 0.09945130315500686,
      "grad_norm": 77.79573822021484,
      "learning_rate": 1.928936436551661e-05,
      "loss": 9.434,
      "step": 145
    },
    {
      "epoch": 0.10013717421124829,
      "grad_norm": 77.5086898803711,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 11.135,
      "step": 146
    },
    {
      "epoch": 0.10082304526748971,
      "grad_norm": 71.69538879394531,
      "learning_rate": 1.800157297483417e-05,
      "loss": 10.8388,
      "step": 147
    },
    {
      "epoch": 0.10150891632373114,
      "grad_norm": 63.49129104614258,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 10.4073,
      "step": 148
    },
    {
      "epoch": 0.10219478737997256,
      "grad_norm": 70.04898071289062,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 9.7506,
      "step": 149
    },
    {
      "epoch": 0.102880658436214,
      "grad_norm": 75.61546325683594,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 10.8766,
      "step": 150
    },
    {
      "epoch": 0.102880658436214,
      "eval_loss": 2.5432915687561035,
      "eval_runtime": 27.7468,
      "eval_samples_per_second": 88.515,
      "eval_steps_per_second": 22.129,
      "step": 150
    },
    {
      "epoch": 0.10356652949245541,
      "grad_norm": 62.569976806640625,
      "learning_rate": 1.553232954407171e-05,
      "loss": 10.3179,
      "step": 151
    },
    {
      "epoch": 0.10425240054869685,
      "grad_norm": 57.56871032714844,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 10.3001,
      "step": 152
    },
    {
      "epoch": 0.10493827160493827,
      "grad_norm": 53.54316711425781,
      "learning_rate": 1.435357758543015e-05,
      "loss": 10.1136,
      "step": 153
    },
    {
      "epoch": 0.1056241426611797,
      "grad_norm": 53.32228469848633,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 10.0691,
      "step": 154
    },
    {
      "epoch": 0.10631001371742113,
      "grad_norm": 63.540740966796875,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 10.2592,
      "step": 155
    },
    {
      "epoch": 0.10699588477366255,
      "grad_norm": 49.91071319580078,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 10.2035,
      "step": 156
    },
    {
      "epoch": 0.10768175582990398,
      "grad_norm": 64.52957916259766,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 10.2365,
      "step": 157
    },
    {
      "epoch": 0.1083676268861454,
      "grad_norm": 54.689125061035156,
      "learning_rate": 1.157994445715706e-05,
      "loss": 10.1331,
      "step": 158
    },
    {
      "epoch": 0.10905349794238683,
      "grad_norm": 54.8894157409668,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 9.9868,
      "step": 159
    },
    {
      "epoch": 0.10973936899862825,
      "grad_norm": 53.94914627075195,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 10.0955,
      "step": 160
    },
    {
      "epoch": 0.11042524005486969,
      "grad_norm": 57.781578063964844,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 9.6592,
      "step": 161
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 50.867130279541016,
      "learning_rate": 9.549150281252633e-06,
      "loss": 9.7487,
      "step": 162
    },
    {
      "epoch": 0.11179698216735254,
      "grad_norm": 57.153541564941406,
      "learning_rate": 9.068759265665384e-06,
      "loss": 10.0747,
      "step": 163
    },
    {
      "epoch": 0.11248285322359397,
      "grad_norm": 54.89235305786133,
      "learning_rate": 8.599558442598998e-06,
      "loss": 10.6241,
      "step": 164
    },
    {
      "epoch": 0.11316872427983539,
      "grad_norm": 59.67436981201172,
      "learning_rate": 8.141676086873572e-06,
      "loss": 10.6134,
      "step": 165
    },
    {
      "epoch": 0.11385459533607682,
      "grad_norm": 59.92189407348633,
      "learning_rate": 7.695237378953223e-06,
      "loss": 10.1748,
      "step": 166
    },
    {
      "epoch": 0.11454046639231824,
      "grad_norm": 83.19001007080078,
      "learning_rate": 7.260364370723044e-06,
      "loss": 10.4407,
      "step": 167
    },
    {
      "epoch": 0.11522633744855967,
      "grad_norm": 65.23087310791016,
      "learning_rate": 6.837175952121306e-06,
      "loss": 10.4514,
      "step": 168
    },
    {
      "epoch": 0.11591220850480109,
      "grad_norm": 71.32051086425781,
      "learning_rate": 6.425787818636131e-06,
      "loss": 10.335,
      "step": 169
    },
    {
      "epoch": 0.11659807956104253,
      "grad_norm": 54.0788688659668,
      "learning_rate": 6.026312439675552e-06,
      "loss": 9.5638,
      "step": 170
    },
    {
      "epoch": 0.11728395061728394,
      "grad_norm": 57.08707046508789,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 10.5702,
      "step": 171
    },
    {
      "epoch": 0.11796982167352538,
      "grad_norm": 69.83554077148438,
      "learning_rate": 5.263533508961827e-06,
      "loss": 10.053,
      "step": 172
    },
    {
      "epoch": 0.11865569272976681,
      "grad_norm": 70.00753021240234,
      "learning_rate": 4.900438493352055e-06,
      "loss": 9.7216,
      "step": 173
    },
    {
      "epoch": 0.11934156378600823,
      "grad_norm": 59.393333435058594,
      "learning_rate": 4.549673247541875e-06,
      "loss": 10.1375,
      "step": 174
    },
    {
      "epoch": 0.12002743484224966,
      "grad_norm": 64.62028503417969,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 9.7999,
      "step": 175
    },
    {
      "epoch": 0.12071330589849108,
      "grad_norm": 56.879478454589844,
      "learning_rate": 3.885512251130763e-06,
      "loss": 10.1316,
      "step": 176
    },
    {
      "epoch": 0.12139917695473251,
      "grad_norm": 61.82935333251953,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 10.7256,
      "step": 177
    },
    {
      "epoch": 0.12208504801097393,
      "grad_norm": 52.658775329589844,
      "learning_rate": 3.271776770026963e-06,
      "loss": 10.3116,
      "step": 178
    },
    {
      "epoch": 0.12277091906721536,
      "grad_norm": 79.22455596923828,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 9.4347,
      "step": 179
    },
    {
      "epoch": 0.12345679012345678,
      "grad_norm": 72.32442474365234,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 10.17,
      "step": 180
    },
    {
      "epoch": 0.12414266117969822,
      "grad_norm": 69.06123352050781,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 10.9063,
      "step": 181
    },
    {
      "epoch": 0.12482853223593965,
      "grad_norm": 62.179603576660156,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 10.4259,
      "step": 182
    },
    {
      "epoch": 0.12551440329218108,
      "grad_norm": 70.77872467041016,
      "learning_rate": 1.962316193157593e-06,
      "loss": 10.6801,
      "step": 183
    },
    {
      "epoch": 0.1262002743484225,
      "grad_norm": 58.33908462524414,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 9.5376,
      "step": 184
    },
    {
      "epoch": 0.12688614540466392,
      "grad_norm": 61.758544921875,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 10.3152,
      "step": 185
    },
    {
      "epoch": 0.12757201646090535,
      "grad_norm": 66.19921112060547,
      "learning_rate": 1.333670137599713e-06,
      "loss": 10.1083,
      "step": 186
    },
    {
      "epoch": 0.12825788751714678,
      "grad_norm": 73.77542114257812,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 9.8743,
      "step": 187
    },
    {
      "epoch": 0.1289437585733882,
      "grad_norm": 66.90348815917969,
      "learning_rate": 9.810017062595322e-07,
      "loss": 9.6947,
      "step": 188
    },
    {
      "epoch": 0.12962962962962962,
      "grad_norm": 69.11670684814453,
      "learning_rate": 8.247462563808817e-07,
      "loss": 9.8188,
      "step": 189
    },
    {
      "epoch": 0.13031550068587106,
      "grad_norm": 62.114540100097656,
      "learning_rate": 6.819348298638839e-07,
      "loss": 9.4239,
      "step": 190
    },
    {
      "epoch": 0.1310013717421125,
      "grad_norm": 93.31256103515625,
      "learning_rate": 5.526064699265753e-07,
      "loss": 9.7525,
      "step": 191
    },
    {
      "epoch": 0.13168724279835392,
      "grad_norm": 74.23255157470703,
      "learning_rate": 4.367965336512403e-07,
      "loss": 9.6702,
      "step": 192
    },
    {
      "epoch": 0.13237311385459533,
      "grad_norm": 76.81401062011719,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 9.7011,
      "step": 193
    },
    {
      "epoch": 0.13305898491083676,
      "grad_norm": 70.80370330810547,
      "learning_rate": 2.458548727494292e-07,
      "loss": 9.7759,
      "step": 194
    },
    {
      "epoch": 0.1337448559670782,
      "grad_norm": 73.9662094116211,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 9.4364,
      "step": 195
    },
    {
      "epoch": 0.13443072702331962,
      "grad_norm": 76.17997741699219,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 8.9591,
      "step": 196
    },
    {
      "epoch": 0.13511659807956103,
      "grad_norm": 66.07431030273438,
      "learning_rate": 6.150154258476315e-08,
      "loss": 10.2433,
      "step": 197
    },
    {
      "epoch": 0.13580246913580246,
      "grad_norm": 78.14649963378906,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 9.7609,
      "step": 198
    },
    {
      "epoch": 0.1364883401920439,
      "grad_norm": 96.0922622680664,
      "learning_rate": 6.834750376549792e-09,
      "loss": 10.4252,
      "step": 199
    },
    {
      "epoch": 0.13717421124828533,
      "grad_norm": 80.69215393066406,
      "learning_rate": 0.0,
      "loss": 15.1446,
      "step": 200
    },
    {
      "epoch": 0.13717421124828533,
      "eval_loss": 2.509678840637207,
      "eval_runtime": 27.9296,
      "eval_samples_per_second": 87.935,
      "eval_steps_per_second": 21.984,
      "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": 1.49024359317504e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}