{
  "best_metric": 1.7213791608810425,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.43360433604336046,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002168021680216802,
      "grad_norm": 29.56697654724121,
      "learning_rate": 1e-05,
      "loss": 6.7917,
      "step": 1
    },
    {
      "epoch": 0.002168021680216802,
      "eval_loss": 2.9564599990844727,
      "eval_runtime": 9.3786,
      "eval_samples_per_second": 82.848,
      "eval_steps_per_second": 20.792,
      "step": 1
    },
    {
      "epoch": 0.004336043360433604,
      "grad_norm": 22.844648361206055,
      "learning_rate": 2e-05,
      "loss": 8.8215,
      "step": 2
    },
    {
      "epoch": 0.0065040650406504065,
      "grad_norm": 27.447298049926758,
      "learning_rate": 3e-05,
      "loss": 7.7656,
      "step": 3
    },
    {
      "epoch": 0.008672086720867209,
      "grad_norm": 25.369524002075195,
      "learning_rate": 4e-05,
      "loss": 9.5796,
      "step": 4
    },
    {
      "epoch": 0.01084010840108401,
      "grad_norm": 24.90519905090332,
      "learning_rate": 5e-05,
      "loss": 8.8202,
      "step": 5
    },
    {
      "epoch": 0.013008130081300813,
      "grad_norm": 23.38373565673828,
      "learning_rate": 6e-05,
      "loss": 8.6864,
      "step": 6
    },
    {
      "epoch": 0.015176151761517615,
      "grad_norm": 23.274526596069336,
      "learning_rate": 7e-05,
      "loss": 8.4283,
      "step": 7
    },
    {
      "epoch": 0.017344173441734417,
      "grad_norm": 25.13153076171875,
      "learning_rate": 8e-05,
      "loss": 9.4565,
      "step": 8
    },
    {
      "epoch": 0.01951219512195122,
      "grad_norm": 30.048437118530273,
      "learning_rate": 9e-05,
      "loss": 9.5829,
      "step": 9
    },
    {
      "epoch": 0.02168021680216802,
      "grad_norm": 27.801332473754883,
      "learning_rate": 0.0001,
      "loss": 9.0878,
      "step": 10
    },
    {
      "epoch": 0.023848238482384824,
      "grad_norm": 32.128700256347656,
      "learning_rate": 9.999316524962345e-05,
      "loss": 9.333,
      "step": 11
    },
    {
      "epoch": 0.026016260162601626,
      "grad_norm": 36.04731369018555,
      "learning_rate": 9.997266286704631e-05,
      "loss": 9.1451,
      "step": 12
    },
    {
      "epoch": 0.028184281842818428,
      "grad_norm": 36.079429626464844,
      "learning_rate": 9.993849845741524e-05,
      "loss": 9.3416,
      "step": 13
    },
    {
      "epoch": 0.03035230352303523,
      "grad_norm": 35.72296905517578,
      "learning_rate": 9.989068136093873e-05,
      "loss": 9.3537,
      "step": 14
    },
    {
      "epoch": 0.032520325203252036,
      "grad_norm": 39.040367126464844,
      "learning_rate": 9.98292246503335e-05,
      "loss": 9.0205,
      "step": 15
    },
    {
      "epoch": 0.034688346883468835,
      "grad_norm": 58.99087905883789,
      "learning_rate": 9.975414512725057e-05,
      "loss": 9.3102,
      "step": 16
    },
    {
      "epoch": 0.03685636856368564,
      "grad_norm": 45.97587585449219,
      "learning_rate": 9.966546331768191e-05,
      "loss": 9.2068,
      "step": 17
    },
    {
      "epoch": 0.03902439024390244,
      "grad_norm": 40.5340461730957,
      "learning_rate": 9.956320346634876e-05,
      "loss": 9.2059,
      "step": 18
    },
    {
      "epoch": 0.041192411924119245,
      "grad_norm": 44.311119079589844,
      "learning_rate": 9.944739353007344e-05,
      "loss": 8.7438,
      "step": 19
    },
    {
      "epoch": 0.04336043360433604,
      "grad_norm": 48.79315948486328,
      "learning_rate": 9.931806517013612e-05,
      "loss": 10.1446,
      "step": 20
    },
    {
      "epoch": 0.04552845528455285,
      "grad_norm": 44.125003814697266,
      "learning_rate": 9.917525374361912e-05,
      "loss": 9.969,
      "step": 21
    },
    {
      "epoch": 0.04769647696476965,
      "grad_norm": 41.132083892822266,
      "learning_rate": 9.901899829374047e-05,
      "loss": 9.857,
      "step": 22
    },
    {
      "epoch": 0.04986449864498645,
      "grad_norm": 41.87261962890625,
      "learning_rate": 9.884934153917997e-05,
      "loss": 10.0139,
      "step": 23
    },
    {
      "epoch": 0.05203252032520325,
      "grad_norm": 39.67081832885742,
      "learning_rate": 9.86663298624003e-05,
      "loss": 8.8703,
      "step": 24
    },
    {
      "epoch": 0.05420054200542006,
      "grad_norm": 44.639129638671875,
      "learning_rate": 9.847001329696653e-05,
      "loss": 9.3676,
      "step": 25
    },
    {
      "epoch": 0.056368563685636856,
      "grad_norm": 58.59317398071289,
      "learning_rate": 9.826044551386744e-05,
      "loss": 9.9945,
      "step": 26
    },
    {
      "epoch": 0.05853658536585366,
      "grad_norm": 58.96311569213867,
      "learning_rate": 9.803768380684242e-05,
      "loss": 9.7769,
      "step": 27
    },
    {
      "epoch": 0.06070460704607046,
      "grad_norm": 101.34839630126953,
      "learning_rate": 9.780178907671789e-05,
      "loss": 9.6022,
      "step": 28
    },
    {
      "epoch": 0.06287262872628727,
      "grad_norm": 55.3743782043457,
      "learning_rate": 9.755282581475769e-05,
      "loss": 9.5984,
      "step": 29
    },
    {
      "epoch": 0.06504065040650407,
      "grad_norm": 52.611061096191406,
      "learning_rate": 9.729086208503174e-05,
      "loss": 9.5768,
      "step": 30
    },
    {
      "epoch": 0.06720867208672086,
      "grad_norm": 63.68136215209961,
      "learning_rate": 9.701596950580806e-05,
      "loss": 10.0241,
      "step": 31
    },
    {
      "epoch": 0.06937669376693767,
      "grad_norm": 131.1375732421875,
      "learning_rate": 9.672822322997305e-05,
      "loss": 10.19,
      "step": 32
    },
    {
      "epoch": 0.07154471544715447,
      "grad_norm": 82.2738037109375,
      "learning_rate": 9.642770192448536e-05,
      "loss": 10.1478,
      "step": 33
    },
    {
      "epoch": 0.07371273712737128,
      "grad_norm": 61.93830490112305,
      "learning_rate": 9.611448774886924e-05,
      "loss": 10.0923,
      "step": 34
    },
    {
      "epoch": 0.07588075880758807,
      "grad_norm": 60.45297622680664,
      "learning_rate": 9.578866633275288e-05,
      "loss": 12.7574,
      "step": 35
    },
    {
      "epoch": 0.07804878048780488,
      "grad_norm": 105.29837036132812,
      "learning_rate": 9.545032675245813e-05,
      "loss": 10.0525,
      "step": 36
    },
    {
      "epoch": 0.08021680216802168,
      "grad_norm": 86.53067016601562,
      "learning_rate": 9.509956150664796e-05,
      "loss": 9.5383,
      "step": 37
    },
    {
      "epoch": 0.08238482384823849,
      "grad_norm": 118.15338134765625,
      "learning_rate": 9.473646649103818e-05,
      "loss": 10.3553,
      "step": 38
    },
    {
      "epoch": 0.08455284552845528,
      "grad_norm": 79.79243469238281,
      "learning_rate": 9.43611409721806e-05,
      "loss": 11.9819,
      "step": 39
    },
    {
      "epoch": 0.08672086720867209,
      "grad_norm": 101.70763397216797,
      "learning_rate": 9.397368756032445e-05,
      "loss": 12.6661,
      "step": 40
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 108.63827514648438,
      "learning_rate": 9.357421218136386e-05,
      "loss": 10.515,
      "step": 41
    },
    {
      "epoch": 0.0910569105691057,
      "grad_norm": 87.01688385009766,
      "learning_rate": 9.316282404787871e-05,
      "loss": 9.7758,
      "step": 42
    },
    {
      "epoch": 0.09322493224932249,
      "grad_norm": 131.51397705078125,
      "learning_rate": 9.273963562927695e-05,
      "loss": 7.3259,
      "step": 43
    },
    {
      "epoch": 0.0953929539295393,
      "grad_norm": 88.70328521728516,
      "learning_rate": 9.230476262104677e-05,
      "loss": 6.2769,
      "step": 44
    },
    {
      "epoch": 0.0975609756097561,
      "grad_norm": 123.07538604736328,
      "learning_rate": 9.185832391312644e-05,
      "loss": 5.2226,
      "step": 45
    },
    {
      "epoch": 0.0997289972899729,
      "grad_norm": 98.83578491210938,
      "learning_rate": 9.140044155740101e-05,
      "loss": 4.1133,
      "step": 46
    },
    {
      "epoch": 0.1018970189701897,
      "grad_norm": 183.61734008789062,
      "learning_rate": 9.093124073433463e-05,
      "loss": 3.8164,
      "step": 47
    },
    {
      "epoch": 0.1040650406504065,
      "grad_norm": 211.9551239013672,
      "learning_rate": 9.045084971874738e-05,
      "loss": 3.5105,
      "step": 48
    },
    {
      "epoch": 0.10623306233062331,
      "grad_norm": 154.25930786132812,
      "learning_rate": 8.995939984474624e-05,
      "loss": 3.2991,
      "step": 49
    },
    {
      "epoch": 0.10840108401084012,
      "grad_norm": 167.2089385986328,
      "learning_rate": 8.945702546981969e-05,
      "loss": 4.4288,
      "step": 50
    },
    {
      "epoch": 0.10840108401084012,
      "eval_loss": 2.758779287338257,
      "eval_runtime": 9.348,
      "eval_samples_per_second": 83.119,
      "eval_steps_per_second": 20.86,
      "step": 50
    },
    {
      "epoch": 0.11056910569105691,
      "grad_norm": 157.94134521484375,
      "learning_rate": 8.894386393810563e-05,
      "loss": 7.8429,
      "step": 51
    },
    {
      "epoch": 0.11273712737127371,
      "grad_norm": 81.84595489501953,
      "learning_rate": 8.842005554284296e-05,
      "loss": 12.011,
      "step": 52
    },
    {
      "epoch": 0.11490514905149052,
      "grad_norm": 71.08673095703125,
      "learning_rate": 8.788574348801675e-05,
      "loss": 14.1127,
      "step": 53
    },
    {
      "epoch": 0.11707317073170732,
      "grad_norm": 59.624229431152344,
      "learning_rate": 8.73410738492077e-05,
      "loss": 11.356,
      "step": 54
    },
    {
      "epoch": 0.11924119241192412,
      "grad_norm": 67.42726135253906,
      "learning_rate": 8.678619553365659e-05,
      "loss": 10.9877,
      "step": 55
    },
    {
      "epoch": 0.12140921409214092,
      "grad_norm": 40.2747688293457,
      "learning_rate": 8.622126023955446e-05,
      "loss": 10.6976,
      "step": 56
    },
    {
      "epoch": 0.12357723577235773,
      "grad_norm": 60.5155029296875,
      "learning_rate": 8.564642241456986e-05,
      "loss": 9.8698,
      "step": 57
    },
    {
      "epoch": 0.12574525745257453,
      "grad_norm": 35.934452056884766,
      "learning_rate": 8.506183921362443e-05,
      "loss": 9.3362,
      "step": 58
    },
    {
      "epoch": 0.12791327913279132,
      "grad_norm": 52.79769515991211,
      "learning_rate": 8.44676704559283e-05,
      "loss": 9.4859,
      "step": 59
    },
    {
      "epoch": 0.13008130081300814,
      "grad_norm": 69.04476165771484,
      "learning_rate": 8.386407858128706e-05,
      "loss": 9.3703,
      "step": 60
    },
    {
      "epoch": 0.13224932249322494,
      "grad_norm": 51.741085052490234,
      "learning_rate": 8.32512286056924e-05,
      "loss": 9.3188,
      "step": 61
    },
    {
      "epoch": 0.13441734417344173,
      "grad_norm": 41.74775314331055,
      "learning_rate": 8.262928807620843e-05,
      "loss": 8.6789,
      "step": 62
    },
    {
      "epoch": 0.13658536585365855,
      "grad_norm": 34.91529846191406,
      "learning_rate": 8.199842702516583e-05,
      "loss": 9.4133,
      "step": 63
    },
    {
      "epoch": 0.13875338753387534,
      "grad_norm": 33.73413848876953,
      "learning_rate": 8.135881792367686e-05,
      "loss": 8.4024,
      "step": 64
    },
    {
      "epoch": 0.14092140921409213,
      "grad_norm": 52.59294891357422,
      "learning_rate": 8.07106356344834e-05,
      "loss": 9.213,
      "step": 65
    },
    {
      "epoch": 0.14308943089430895,
      "grad_norm": 44.27796936035156,
      "learning_rate": 8.005405736415126e-05,
      "loss": 9.4206,
      "step": 66
    },
    {
      "epoch": 0.14525745257452574,
      "grad_norm": 32.0435905456543,
      "learning_rate": 7.938926261462366e-05,
      "loss": 9.1629,
      "step": 67
    },
    {
      "epoch": 0.14742547425474256,
      "grad_norm": 56.0610237121582,
      "learning_rate": 7.871643313414718e-05,
      "loss": 8.8188,
      "step": 68
    },
    {
      "epoch": 0.14959349593495935,
      "grad_norm": 31.955718994140625,
      "learning_rate": 7.803575286758364e-05,
      "loss": 10.0232,
      "step": 69
    },
    {
      "epoch": 0.15176151761517614,
      "grad_norm": 39.53376770019531,
      "learning_rate": 7.734740790612136e-05,
      "loss": 9.4435,
      "step": 70
    },
    {
      "epoch": 0.15392953929539296,
      "grad_norm": 28.377092361450195,
      "learning_rate": 7.66515864363997e-05,
      "loss": 9.36,
      "step": 71
    },
    {
      "epoch": 0.15609756097560976,
      "grad_norm": 64.54959106445312,
      "learning_rate": 7.594847868906076e-05,
      "loss": 9.6367,
      "step": 72
    },
    {
      "epoch": 0.15826558265582655,
      "grad_norm": 38.5235710144043,
      "learning_rate": 7.52382768867422e-05,
      "loss": 9.7493,
      "step": 73
    },
    {
      "epoch": 0.16043360433604337,
      "grad_norm": 45.679569244384766,
      "learning_rate": 7.452117519152542e-05,
      "loss": 9.6859,
      "step": 74
    },
    {
      "epoch": 0.16260162601626016,
      "grad_norm": 40.02254104614258,
      "learning_rate": 7.379736965185368e-05,
      "loss": 10.0519,
      "step": 75
    },
    {
      "epoch": 0.16476964769647698,
      "grad_norm": 38.8911018371582,
      "learning_rate": 7.30670581489344e-05,
      "loss": 9.3061,
      "step": 76
    },
    {
      "epoch": 0.16693766937669377,
      "grad_norm": 32.40317916870117,
      "learning_rate": 7.233044034264034e-05,
      "loss": 9.2322,
      "step": 77
    },
    {
      "epoch": 0.16910569105691056,
      "grad_norm": 37.332603454589844,
      "learning_rate": 7.158771761692464e-05,
      "loss": 9.3109,
      "step": 78
    },
    {
      "epoch": 0.17127371273712738,
      "grad_norm": 38.216148376464844,
      "learning_rate": 7.083909302476453e-05,
      "loss": 9.8138,
      "step": 79
    },
    {
      "epoch": 0.17344173441734417,
      "grad_norm": 39.74815368652344,
      "learning_rate": 7.008477123264848e-05,
      "loss": 8.7014,
      "step": 80
    },
    {
      "epoch": 0.17560975609756097,
      "grad_norm": 44.80895233154297,
      "learning_rate": 6.932495846462261e-05,
      "loss": 8.8614,
      "step": 81
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 40.72815704345703,
      "learning_rate": 6.855986244591104e-05,
      "loss": 9.1287,
      "step": 82
    },
    {
      "epoch": 0.17994579945799458,
      "grad_norm": 63.297245025634766,
      "learning_rate": 6.778969234612584e-05,
      "loss": 9.0044,
      "step": 83
    },
    {
      "epoch": 0.1821138211382114,
      "grad_norm": 49.13942337036133,
      "learning_rate": 6.701465872208216e-05,
      "loss": 10.1192,
      "step": 84
    },
    {
      "epoch": 0.1842818428184282,
      "grad_norm": 52.749820709228516,
      "learning_rate": 6.623497346023418e-05,
      "loss": 9.4146,
      "step": 85
    },
    {
      "epoch": 0.18644986449864498,
      "grad_norm": 53.51035690307617,
      "learning_rate": 6.545084971874738e-05,
      "loss": 8.7282,
      "step": 86
    },
    {
      "epoch": 0.1886178861788618,
      "grad_norm": 65.0464096069336,
      "learning_rate": 6.466250186922325e-05,
      "loss": 8.2676,
      "step": 87
    },
    {
      "epoch": 0.1907859078590786,
      "grad_norm": 99.9498291015625,
      "learning_rate": 6.387014543809223e-05,
      "loss": 9.0324,
      "step": 88
    },
    {
      "epoch": 0.19295392953929538,
      "grad_norm": 79.33631896972656,
      "learning_rate": 6.307399704769099e-05,
      "loss": 5.7613,
      "step": 89
    },
    {
      "epoch": 0.1951219512195122,
      "grad_norm": 96.63216400146484,
      "learning_rate": 6.227427435703997e-05,
      "loss": 6.6183,
      "step": 90
    },
    {
      "epoch": 0.197289972899729,
      "grad_norm": 89.3034439086914,
      "learning_rate": 6.147119600233758e-05,
      "loss": 4.3904,
      "step": 91
    },
    {
      "epoch": 0.1994579945799458,
      "grad_norm": 82.56645965576172,
      "learning_rate": 6.066498153718735e-05,
      "loss": 5.1966,
      "step": 92
    },
    {
      "epoch": 0.2016260162601626,
      "grad_norm": 72.83575439453125,
      "learning_rate": 5.985585137257401e-05,
      "loss": 2.9211,
      "step": 93
    },
    {
      "epoch": 0.2037940379403794,
      "grad_norm": 124.21688842773438,
      "learning_rate": 5.90440267166055e-05,
      "loss": 2.023,
      "step": 94
    },
    {
      "epoch": 0.20596205962059622,
      "grad_norm": 103.04071044921875,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.1432,
      "step": 95
    },
    {
      "epoch": 0.208130081300813,
      "grad_norm": 49.1472282409668,
      "learning_rate": 5.74131823855921e-05,
      "loss": 1.2012,
      "step": 96
    },
    {
      "epoch": 0.2102981029810298,
      "grad_norm": 93.72968292236328,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 1.6871,
      "step": 97
    },
    {
      "epoch": 0.21246612466124662,
      "grad_norm": 100.07096862792969,
      "learning_rate": 5.577423184847932e-05,
      "loss": 1.3877,
      "step": 98
    },
    {
      "epoch": 0.2146341463414634,
      "grad_norm": 39.16847610473633,
      "learning_rate": 5.495227651252315e-05,
      "loss": 0.7818,
      "step": 99
    },
    {
      "epoch": 0.21680216802168023,
      "grad_norm": 56.72313690185547,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 2.602,
      "step": 100
    },
    {
      "epoch": 0.21680216802168023,
      "eval_loss": 1.9244413375854492,
      "eval_runtime": 9.3416,
      "eval_samples_per_second": 83.177,
      "eval_steps_per_second": 20.874,
      "step": 100
    },
    {
      "epoch": 0.21897018970189702,
      "grad_norm": 35.613861083984375,
      "learning_rate": 5.330452921628497e-05,
      "loss": 4.964,
      "step": 101
    },
    {
      "epoch": 0.22113821138211381,
      "grad_norm": 40.78394317626953,
      "learning_rate": 5.247918773366112e-05,
      "loss": 11.7878,
      "step": 102
    },
    {
      "epoch": 0.22330623306233063,
      "grad_norm": 29.11028480529785,
      "learning_rate": 5.165316846586541e-05,
      "loss": 9.1479,
      "step": 103
    },
    {
      "epoch": 0.22547425474254743,
      "grad_norm": 35.1927604675293,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 10.0267,
      "step": 104
    },
    {
      "epoch": 0.22764227642276422,
      "grad_norm": 49.512245178222656,
      "learning_rate": 5e-05,
      "loss": 8.8112,
      "step": 105
    },
    {
      "epoch": 0.22981029810298104,
      "grad_norm": 41.58905029296875,
      "learning_rate": 4.917330276168208e-05,
      "loss": 8.0448,
      "step": 106
    },
    {
      "epoch": 0.23197831978319783,
      "grad_norm": 29.24210548400879,
      "learning_rate": 4.834683153413459e-05,
      "loss": 8.3196,
      "step": 107
    },
    {
      "epoch": 0.23414634146341465,
      "grad_norm": 28.055253982543945,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 8.8809,
      "step": 108
    },
    {
      "epoch": 0.23631436314363144,
      "grad_norm": 23.70207977294922,
      "learning_rate": 4.669547078371504e-05,
      "loss": 8.9278,
      "step": 109
    },
    {
      "epoch": 0.23848238482384823,
      "grad_norm": 39.66957092285156,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 9.1542,
      "step": 110
    },
    {
      "epoch": 0.24065040650406505,
      "grad_norm": 22.6226749420166,
      "learning_rate": 4.504772348747687e-05,
      "loss": 8.8319,
      "step": 111
    },
    {
      "epoch": 0.24281842818428184,
      "grad_norm": 32.19573211669922,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 9.4761,
      "step": 112
    },
    {
      "epoch": 0.24498644986449863,
      "grad_norm": 53.35695266723633,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 8.1584,
      "step": 113
    },
    {
      "epoch": 0.24715447154471545,
      "grad_norm": 22.227432250976562,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 9.1379,
      "step": 114
    },
    {
      "epoch": 0.24932249322493225,
      "grad_norm": 19.7288818359375,
      "learning_rate": 4.17702704859633e-05,
      "loss": 8.3286,
      "step": 115
    },
    {
      "epoch": 0.25149051490514907,
      "grad_norm": 20.1420841217041,
      "learning_rate": 4.095597328339452e-05,
      "loss": 8.4622,
      "step": 116
    },
    {
      "epoch": 0.25365853658536586,
      "grad_norm": 20.648754119873047,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 8.313,
      "step": 117
    },
    {
      "epoch": 0.25582655826558265,
      "grad_norm": 18.873445510864258,
      "learning_rate": 3.933501846281267e-05,
      "loss": 8.4576,
      "step": 118
    },
    {
      "epoch": 0.25799457994579944,
      "grad_norm": 19.276405334472656,
      "learning_rate": 3.852880399766243e-05,
      "loss": 8.8261,
      "step": 119
    },
    {
      "epoch": 0.2601626016260163,
      "grad_norm": 26.74322509765625,
      "learning_rate": 3.772572564296005e-05,
      "loss": 8.7568,
      "step": 120
    },
    {
      "epoch": 0.2623306233062331,
      "grad_norm": 32.53742980957031,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 9.0809,
      "step": 121
    },
    {
      "epoch": 0.26449864498644987,
      "grad_norm": 20.733869552612305,
      "learning_rate": 3.612985456190778e-05,
      "loss": 8.9164,
      "step": 122
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 25.21817970275879,
      "learning_rate": 3.533749813077677e-05,
      "loss": 9.0173,
      "step": 123
    },
    {
      "epoch": 0.26883468834688345,
      "grad_norm": 21.50628089904785,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 8.8179,
      "step": 124
    },
    {
      "epoch": 0.27100271002710025,
      "grad_norm": 20.020416259765625,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 9.306,
      "step": 125
    },
    {
      "epoch": 0.2731707317073171,
      "grad_norm": 20.90715789794922,
      "learning_rate": 3.298534127791785e-05,
      "loss": 9.0805,
      "step": 126
    },
    {
      "epoch": 0.2753387533875339,
      "grad_norm": 21.389572143554688,
      "learning_rate": 3.221030765387417e-05,
      "loss": 9.0048,
      "step": 127
    },
    {
      "epoch": 0.2775067750677507,
      "grad_norm": 26.47258758544922,
      "learning_rate": 3.144013755408895e-05,
      "loss": 8.9581,
      "step": 128
    },
    {
      "epoch": 0.27967479674796747,
      "grad_norm": 26.793968200683594,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 8.1696,
      "step": 129
    },
    {
      "epoch": 0.28184281842818426,
      "grad_norm": 23.339202880859375,
      "learning_rate": 2.991522876735154e-05,
      "loss": 8.7234,
      "step": 130
    },
    {
      "epoch": 0.2840108401084011,
      "grad_norm": 26.5540771484375,
      "learning_rate": 2.916090697523549e-05,
      "loss": 8.2496,
      "step": 131
    },
    {
      "epoch": 0.2861788617886179,
      "grad_norm": 26.76554298400879,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 8.164,
      "step": 132
    },
    {
      "epoch": 0.2883468834688347,
      "grad_norm": 25.66951560974121,
      "learning_rate": 2.766955965735968e-05,
      "loss": 8.5713,
      "step": 133
    },
    {
      "epoch": 0.2905149051490515,
      "grad_norm": 27.322519302368164,
      "learning_rate": 2.693294185106562e-05,
      "loss": 7.706,
      "step": 134
    },
    {
      "epoch": 0.2926829268292683,
      "grad_norm": 29.11598777770996,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 9.3715,
      "step": 135
    },
    {
      "epoch": 0.2948509485094851,
      "grad_norm": 29.22189712524414,
      "learning_rate": 2.547882480847461e-05,
      "loss": 8.8805,
      "step": 136
    },
    {
      "epoch": 0.2970189701897019,
      "grad_norm": 40.12459182739258,
      "learning_rate": 2.476172311325783e-05,
      "loss": 7.1181,
      "step": 137
    },
    {
      "epoch": 0.2991869918699187,
      "grad_norm": 37.60042190551758,
      "learning_rate": 2.405152131093926e-05,
      "loss": 6.6358,
      "step": 138
    },
    {
      "epoch": 0.3013550135501355,
      "grad_norm": 44.48047637939453,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 7.6617,
      "step": 139
    },
    {
      "epoch": 0.3035230352303523,
      "grad_norm": 69.62174224853516,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 5.7551,
      "step": 140
    },
    {
      "epoch": 0.3056910569105691,
      "grad_norm": 52.42491149902344,
      "learning_rate": 2.196424713241637e-05,
      "loss": 2.2201,
      "step": 141
    },
    {
      "epoch": 0.30785907859078593,
      "grad_norm": 51.486175537109375,
      "learning_rate": 2.128356686585282e-05,
      "loss": 3.3775,
      "step": 142
    },
    {
      "epoch": 0.3100271002710027,
      "grad_norm": 58.227561950683594,
      "learning_rate": 2.061073738537635e-05,
      "loss": 1.907,
      "step": 143
    },
    {
      "epoch": 0.3121951219512195,
      "grad_norm": 51.19715881347656,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 1.293,
      "step": 144
    },
    {
      "epoch": 0.3143631436314363,
      "grad_norm": 83.05162048339844,
      "learning_rate": 1.928936436551661e-05,
      "loss": 1.3919,
      "step": 145
    },
    {
      "epoch": 0.3165311653116531,
      "grad_norm": 46.72926712036133,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 0.5313,
      "step": 146
    },
    {
      "epoch": 0.31869918699186994,
      "grad_norm": 61.03974914550781,
      "learning_rate": 1.800157297483417e-05,
      "loss": 0.7563,
      "step": 147
    },
    {
      "epoch": 0.32086720867208673,
      "grad_norm": 50.94718551635742,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 0.8253,
      "step": 148
    },
    {
      "epoch": 0.3230352303523035,
      "grad_norm": 23.519289016723633,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 0.4298,
      "step": 149
    },
    {
      "epoch": 0.3252032520325203,
      "grad_norm": 37.908599853515625,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 1.921,
      "step": 150
    },
    {
      "epoch": 0.3252032520325203,
      "eval_loss": 1.788693904876709,
      "eval_runtime": 9.3835,
      "eval_samples_per_second": 82.805,
      "eval_steps_per_second": 20.781,
      "step": 150
    },
    {
      "epoch": 0.3273712737127371,
      "grad_norm": 23.393863677978516,
      "learning_rate": 1.553232954407171e-05,
      "loss": 6.4719,
      "step": 151
    },
    {
      "epoch": 0.32953929539295396,
      "grad_norm": 67.25944519042969,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 9.8484,
      "step": 152
    },
    {
      "epoch": 0.33170731707317075,
      "grad_norm": 22.97135353088379,
      "learning_rate": 1.435357758543015e-05,
      "loss": 9.9367,
      "step": 153
    },
    {
      "epoch": 0.33387533875338754,
      "grad_norm": 21.3618221282959,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 9.5086,
      "step": 154
    },
    {
      "epoch": 0.33604336043360433,
      "grad_norm": 24.72620391845703,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 8.4169,
      "step": 155
    },
    {
      "epoch": 0.3382113821138211,
      "grad_norm": 23.45319366455078,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 8.6031,
      "step": 156
    },
    {
      "epoch": 0.3403794037940379,
      "grad_norm": 29.1568660736084,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 8.9896,
      "step": 157
    },
    {
      "epoch": 0.34254742547425476,
      "grad_norm": 21.772184371948242,
      "learning_rate": 1.157994445715706e-05,
      "loss": 8.3909,
      "step": 158
    },
    {
      "epoch": 0.34471544715447155,
      "grad_norm": 17.402957916259766,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 8.1657,
      "step": 159
    },
    {
      "epoch": 0.34688346883468835,
      "grad_norm": 21.280426025390625,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 8.2439,
      "step": 160
    },
    {
      "epoch": 0.34905149051490514,
      "grad_norm": 18.396419525146484,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 8.3202,
      "step": 161
    },
    {
      "epoch": 0.35121951219512193,
      "grad_norm": 25.474815368652344,
      "learning_rate": 9.549150281252633e-06,
      "loss": 8.4802,
      "step": 162
    },
    {
      "epoch": 0.3533875338753388,
      "grad_norm": 22.321760177612305,
      "learning_rate": 9.068759265665384e-06,
      "loss": 8.9247,
      "step": 163
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 19.393081665039062,
      "learning_rate": 8.599558442598998e-06,
      "loss": 8.2353,
      "step": 164
    },
    {
      "epoch": 0.35772357723577236,
      "grad_norm": 22.461700439453125,
      "learning_rate": 8.141676086873572e-06,
      "loss": 8.3274,
      "step": 165
    },
    {
      "epoch": 0.35989159891598915,
      "grad_norm": 16.620500564575195,
      "learning_rate": 7.695237378953223e-06,
      "loss": 8.5916,
      "step": 166
    },
    {
      "epoch": 0.36205962059620594,
      "grad_norm": 21.414871215820312,
      "learning_rate": 7.260364370723044e-06,
      "loss": 8.1114,
      "step": 167
    },
    {
      "epoch": 0.3642276422764228,
      "grad_norm": 46.16633224487305,
      "learning_rate": 6.837175952121306e-06,
      "loss": 8.3523,
      "step": 168
    },
    {
      "epoch": 0.3663956639566396,
      "grad_norm": 28.88848114013672,
      "learning_rate": 6.425787818636131e-06,
      "loss": 7.8829,
      "step": 169
    },
    {
      "epoch": 0.3685636856368564,
      "grad_norm": 28.91385269165039,
      "learning_rate": 6.026312439675552e-06,
      "loss": 9.0534,
      "step": 170
    },
    {
      "epoch": 0.37073170731707317,
      "grad_norm": 20.6962833404541,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 9.1964,
      "step": 171
    },
    {
      "epoch": 0.37289972899728996,
      "grad_norm": 19.90764808654785,
      "learning_rate": 5.263533508961827e-06,
      "loss": 9.1433,
      "step": 172
    },
    {
      "epoch": 0.37506775067750675,
      "grad_norm": 22.188579559326172,
      "learning_rate": 4.900438493352055e-06,
      "loss": 9.7097,
      "step": 173
    },
    {
      "epoch": 0.3772357723577236,
      "grad_norm": 19.773141860961914,
      "learning_rate": 4.549673247541875e-06,
      "loss": 8.6201,
      "step": 174
    },
    {
      "epoch": 0.3794037940379404,
      "grad_norm": 29.198694229125977,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 8.9861,
      "step": 175
    },
    {
      "epoch": 0.3815718157181572,
      "grad_norm": 25.00130844116211,
      "learning_rate": 3.885512251130763e-06,
      "loss": 8.8726,
      "step": 176
    },
    {
      "epoch": 0.383739837398374,
      "grad_norm": 23.308883666992188,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 9.429,
      "step": 177
    },
    {
      "epoch": 0.38590785907859076,
      "grad_norm": 22.925567626953125,
      "learning_rate": 3.271776770026963e-06,
      "loss": 8.6373,
      "step": 178
    },
    {
      "epoch": 0.3880758807588076,
      "grad_norm": 23.13882827758789,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 9.5522,
      "step": 179
    },
    {
      "epoch": 0.3902439024390244,
      "grad_norm": 39.564605712890625,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 8.0607,
      "step": 180
    },
    {
      "epoch": 0.3924119241192412,
      "grad_norm": 23.98958396911621,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 9.0663,
      "step": 181
    },
    {
      "epoch": 0.394579945799458,
      "grad_norm": 27.768707275390625,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 7.6105,
      "step": 182
    },
    {
      "epoch": 0.3967479674796748,
      "grad_norm": 27.029319763183594,
      "learning_rate": 1.962316193157593e-06,
      "loss": 7.9193,
      "step": 183
    },
    {
      "epoch": 0.3989159891598916,
      "grad_norm": 26.111921310424805,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 8.5289,
      "step": 184
    },
    {
      "epoch": 0.4010840108401084,
      "grad_norm": 28.225522994995117,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 9.1447,
      "step": 185
    },
    {
      "epoch": 0.4032520325203252,
      "grad_norm": 41.012489318847656,
      "learning_rate": 1.333670137599713e-06,
      "loss": 9.9929,
      "step": 186
    },
    {
      "epoch": 0.405420054200542,
      "grad_norm": 30.45234489440918,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 8.7301,
      "step": 187
    },
    {
      "epoch": 0.4075880758807588,
      "grad_norm": 39.41157913208008,
      "learning_rate": 9.810017062595322e-07,
      "loss": 5.2084,
      "step": 188
    },
    {
      "epoch": 0.4097560975609756,
      "grad_norm": 44.882301330566406,
      "learning_rate": 8.247462563808817e-07,
      "loss": 4.658,
      "step": 189
    },
    {
      "epoch": 0.41192411924119243,
      "grad_norm": 67.9166488647461,
      "learning_rate": 6.819348298638839e-07,
      "loss": 6.5314,
      "step": 190
    },
    {
      "epoch": 0.4140921409214092,
      "grad_norm": 67.44029235839844,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.9418,
      "step": 191
    },
    {
      "epoch": 0.416260162601626,
      "grad_norm": 42.995975494384766,
      "learning_rate": 4.367965336512403e-07,
      "loss": 1.5539,
      "step": 192
    },
    {
      "epoch": 0.4184281842818428,
      "grad_norm": 72.60922241210938,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 3.2945,
      "step": 193
    },
    {
      "epoch": 0.4205962059620596,
      "grad_norm": 59.845577239990234,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.0104,
      "step": 194
    },
    {
      "epoch": 0.42276422764227645,
      "grad_norm": 49.002201080322266,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 0.894,
      "step": 195
    },
    {
      "epoch": 0.42493224932249324,
      "grad_norm": 50.669837951660156,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 1.0598,
      "step": 196
    },
    {
      "epoch": 0.42710027100271003,
      "grad_norm": 38.22815704345703,
      "learning_rate": 6.150154258476315e-08,
      "loss": 0.7984,
      "step": 197
    },
    {
      "epoch": 0.4292682926829268,
      "grad_norm": 47.77947235107422,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.0837,
      "step": 198
    },
    {
      "epoch": 0.4314363143631436,
      "grad_norm": 41.17764663696289,
      "learning_rate": 6.834750376549792e-09,
      "loss": 0.6525,
      "step": 199
    },
    {
      "epoch": 0.43360433604336046,
      "grad_norm": 48.83980941772461,
      "learning_rate": 0.0,
      "loss": 2.9829,
      "step": 200
    },
    {
      "epoch": 0.43360433604336046,
      "eval_loss": 1.7213791608810425,
      "eval_runtime": 9.3323,
      "eval_samples_per_second": 83.259,
      "eval_steps_per_second": 20.895,
      "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": 1.5703441863081984e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}