{
  "best_metric": 0.04161956161260605,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.0171155944459896,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 8.557797222994801e-05,
      "grad_norm": 13.020096778869629,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 3.3018,
      "step": 1
    },
    {
      "epoch": 8.557797222994801e-05,
      "eval_loss": 3.5091288089752197,
      "eval_runtime": 1592.4677,
      "eval_samples_per_second": 12.359,
      "eval_steps_per_second": 6.18,
      "step": 1
    },
    {
      "epoch": 0.00017115594445989602,
      "grad_norm": 13.982504844665527,
      "learning_rate": 6.666666666666667e-06,
      "loss": 3.6035,
      "step": 2
    },
    {
      "epoch": 0.000256733916689844,
      "grad_norm": 12.648855209350586,
      "learning_rate": 1e-05,
      "loss": 3.4461,
      "step": 3
    },
    {
      "epoch": 0.00034231188891979203,
      "grad_norm": 11.70174503326416,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 3.2101,
      "step": 4
    },
    {
      "epoch": 0.00042788986114974004,
      "grad_norm": 9.91709041595459,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 3.1165,
      "step": 5
    },
    {
      "epoch": 0.000513467833379688,
      "grad_norm": 8.46243667602539,
      "learning_rate": 2e-05,
      "loss": 2.8427,
      "step": 6
    },
    {
      "epoch": 0.0005990458056096361,
      "grad_norm": 8.12342357635498,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 2.3386,
      "step": 7
    },
    {
      "epoch": 0.0006846237778395841,
      "grad_norm": 7.001865863800049,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 2.1186,
      "step": 8
    },
    {
      "epoch": 0.0007702017500695321,
      "grad_norm": 6.231939315795898,
      "learning_rate": 3e-05,
      "loss": 1.6328,
      "step": 9
    },
    {
      "epoch": 0.0008557797222994801,
      "grad_norm": 6.940419673919678,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.1098,
      "step": 10
    },
    {
      "epoch": 0.0009413576945294281,
      "grad_norm": 6.4078369140625,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 0.8605,
      "step": 11
    },
    {
      "epoch": 0.001026935666759376,
      "grad_norm": 5.962477207183838,
      "learning_rate": 4e-05,
      "loss": 0.4664,
      "step": 12
    },
    {
      "epoch": 0.0011125136389893242,
      "grad_norm": 7.7827067375183105,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 0.2416,
      "step": 13
    },
    {
      "epoch": 0.0011980916112192721,
      "grad_norm": 8.21843147277832,
      "learning_rate": 4.666666666666667e-05,
      "loss": 0.1779,
      "step": 14
    },
    {
      "epoch": 0.0012836695834492202,
      "grad_norm": 4.3173723220825195,
      "learning_rate": 5e-05,
      "loss": 0.1195,
      "step": 15
    },
    {
      "epoch": 0.0013692475556791681,
      "grad_norm": 11.584345817565918,
      "learning_rate": 5.333333333333333e-05,
      "loss": 0.178,
      "step": 16
    },
    {
      "epoch": 0.0014548255279091162,
      "grad_norm": 6.5930047035217285,
      "learning_rate": 5.666666666666667e-05,
      "loss": 0.1796,
      "step": 17
    },
    {
      "epoch": 0.0015404035001390641,
      "grad_norm": 6.612515926361084,
      "learning_rate": 6e-05,
      "loss": 0.162,
      "step": 18
    },
    {
      "epoch": 0.0016259814723690123,
      "grad_norm": 16.772546768188477,
      "learning_rate": 6.333333333333333e-05,
      "loss": 0.3361,
      "step": 19
    },
    {
      "epoch": 0.0017115594445989602,
      "grad_norm": 3.768550395965576,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.1024,
      "step": 20
    },
    {
      "epoch": 0.0017971374168289083,
      "grad_norm": 5.523356914520264,
      "learning_rate": 7e-05,
      "loss": 0.24,
      "step": 21
    },
    {
      "epoch": 0.0018827153890588562,
      "grad_norm": 3.7887825965881348,
      "learning_rate": 7.333333333333333e-05,
      "loss": 0.0857,
      "step": 22
    },
    {
      "epoch": 0.0019682933612888043,
      "grad_norm": 18.40159034729004,
      "learning_rate": 7.666666666666667e-05,
      "loss": 0.5776,
      "step": 23
    },
    {
      "epoch": 0.002053871333518752,
      "grad_norm": 10.70200252532959,
      "learning_rate": 8e-05,
      "loss": 0.4233,
      "step": 24
    },
    {
      "epoch": 0.0021394493057487,
      "grad_norm": 6.764603614807129,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.091,
      "step": 25
    },
    {
      "epoch": 0.0022250272779786484,
      "grad_norm": 3.3309624195098877,
      "learning_rate": 8.666666666666667e-05,
      "loss": 0.1264,
      "step": 26
    },
    {
      "epoch": 0.0023106052502085963,
      "grad_norm": 1.2735038995742798,
      "learning_rate": 9e-05,
      "loss": 0.0888,
      "step": 27
    },
    {
      "epoch": 0.0023961832224385442,
      "grad_norm": 3.2824547290802,
      "learning_rate": 9.333333333333334e-05,
      "loss": 0.1087,
      "step": 28
    },
    {
      "epoch": 0.002481761194668492,
      "grad_norm": 2.485607862472534,
      "learning_rate": 9.666666666666667e-05,
      "loss": 0.0832,
      "step": 29
    },
    {
      "epoch": 0.0025673391668984405,
      "grad_norm": 3.024195432662964,
      "learning_rate": 0.0001,
      "loss": 0.1534,
      "step": 30
    },
    {
      "epoch": 0.0026529171391283884,
      "grad_norm": 1.252853512763977,
      "learning_rate": 9.999146252290264e-05,
      "loss": 0.0844,
      "step": 31
    },
    {
      "epoch": 0.0027384951113583363,
      "grad_norm": 2.0944342613220215,
      "learning_rate": 9.996585300715116e-05,
      "loss": 0.0752,
      "step": 32
    },
    {
      "epoch": 0.002824073083588284,
      "grad_norm": 1.5007476806640625,
      "learning_rate": 9.99231801983717e-05,
      "loss": 0.0694,
      "step": 33
    },
    {
      "epoch": 0.0029096510558182325,
      "grad_norm": 1.4180450439453125,
      "learning_rate": 9.986345866928941e-05,
      "loss": 0.0709,
      "step": 34
    },
    {
      "epoch": 0.0029952290280481804,
      "grad_norm": 1.172573208808899,
      "learning_rate": 9.978670881475172e-05,
      "loss": 0.0907,
      "step": 35
    },
    {
      "epoch": 0.0030808070002781283,
      "grad_norm": 1.704951286315918,
      "learning_rate": 9.96929568447637e-05,
      "loss": 0.0737,
      "step": 36
    },
    {
      "epoch": 0.0031663849725080766,
      "grad_norm": 0.877009391784668,
      "learning_rate": 9.958223477553714e-05,
      "loss": 0.0579,
      "step": 37
    },
    {
      "epoch": 0.0032519629447380245,
      "grad_norm": 1.010015606880188,
      "learning_rate": 9.94545804185573e-05,
      "loss": 0.0547,
      "step": 38
    },
    {
      "epoch": 0.0033375409169679724,
      "grad_norm": 0.6118600964546204,
      "learning_rate": 9.931003736767013e-05,
      "loss": 0.0676,
      "step": 39
    },
    {
      "epoch": 0.0034231188891979203,
      "grad_norm": 0.8259471654891968,
      "learning_rate": 9.91486549841951e-05,
      "loss": 0.0621,
      "step": 40
    },
    {
      "epoch": 0.0035086968614278687,
      "grad_norm": 1.026671051979065,
      "learning_rate": 9.89704883800683e-05,
      "loss": 0.0478,
      "step": 41
    },
    {
      "epoch": 0.0035942748336578166,
      "grad_norm": 1.778664469718933,
      "learning_rate": 9.877559839902184e-05,
      "loss": 0.0773,
      "step": 42
    },
    {
      "epoch": 0.0036798528058877645,
      "grad_norm": 1.4090718030929565,
      "learning_rate": 9.85640515958057e-05,
      "loss": 0.0609,
      "step": 43
    },
    {
      "epoch": 0.0037654307781177124,
      "grad_norm": 1.2930989265441895,
      "learning_rate": 9.833592021345937e-05,
      "loss": 0.0745,
      "step": 44
    },
    {
      "epoch": 0.0038510087503476607,
      "grad_norm": 0.5950098037719727,
      "learning_rate": 9.809128215864097e-05,
      "loss": 0.0446,
      "step": 45
    },
    {
      "epoch": 0.003936586722577609,
      "grad_norm": 2.021824598312378,
      "learning_rate": 9.783022097502204e-05,
      "loss": 0.089,
      "step": 46
    },
    {
      "epoch": 0.004022164694807557,
      "grad_norm": 0.6578583717346191,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.0357,
      "step": 47
    },
    {
      "epoch": 0.004107742667037504,
      "grad_norm": 1.012323021888733,
      "learning_rate": 9.725919140804099e-05,
      "loss": 0.0606,
      "step": 48
    },
    {
      "epoch": 0.004193320639267453,
      "grad_norm": 1.2615344524383545,
      "learning_rate": 9.694941803075283e-05,
      "loss": 0.0438,
      "step": 49
    },
    {
      "epoch": 0.0042788986114974,
      "grad_norm": 1.3731071949005127,
      "learning_rate": 9.662361147021779e-05,
      "loss": 0.0568,
      "step": 50
    },
    {
      "epoch": 0.0042788986114974,
      "eval_loss": 0.2535214424133301,
      "eval_runtime": 1602.7735,
      "eval_samples_per_second": 12.279,
      "eval_steps_per_second": 6.14,
      "step": 50
    },
    {
      "epoch": 0.0043644765837273485,
      "grad_norm": 2.4292540550231934,
      "learning_rate": 9.628188298907782e-05,
      "loss": 0.3446,
      "step": 51
    },
    {
      "epoch": 0.004450054555957297,
      "grad_norm": 2.532665729522705,
      "learning_rate": 9.592434928729616e-05,
      "loss": 0.3271,
      "step": 52
    },
    {
      "epoch": 0.004535632528187244,
      "grad_norm": 1.4810094833374023,
      "learning_rate": 9.555113246230442e-05,
      "loss": 0.1112,
      "step": 53
    },
    {
      "epoch": 0.004621210500417193,
      "grad_norm": 2.738724946975708,
      "learning_rate": 9.516235996730645e-05,
      "loss": 0.1618,
      "step": 54
    },
    {
      "epoch": 0.004706788472647141,
      "grad_norm": 1.8832288980484009,
      "learning_rate": 9.475816456775313e-05,
      "loss": 0.1699,
      "step": 55
    },
    {
      "epoch": 0.0047923664448770885,
      "grad_norm": 2.132007360458374,
      "learning_rate": 9.43386842960031e-05,
      "loss": 0.1674,
      "step": 56
    },
    {
      "epoch": 0.004877944417107037,
      "grad_norm": 1.064928650856018,
      "learning_rate": 9.39040624041849e-05,
      "loss": 0.0901,
      "step": 57
    },
    {
      "epoch": 0.004963522389336984,
      "grad_norm": 1.07744300365448,
      "learning_rate": 9.345444731527642e-05,
      "loss": 0.1,
      "step": 58
    },
    {
      "epoch": 0.005049100361566933,
      "grad_norm": 1.2358143329620361,
      "learning_rate": 9.298999257241863e-05,
      "loss": 0.1119,
      "step": 59
    },
    {
      "epoch": 0.005134678333796881,
      "grad_norm": 1.293792724609375,
      "learning_rate": 9.251085678648072e-05,
      "loss": 0.1038,
      "step": 60
    },
    {
      "epoch": 0.005220256306026828,
      "grad_norm": 1.1207275390625,
      "learning_rate": 9.201720358189464e-05,
      "loss": 0.0922,
      "step": 61
    },
    {
      "epoch": 0.005305834278256777,
      "grad_norm": 1.7835878133773804,
      "learning_rate": 9.150920154077754e-05,
      "loss": 0.1157,
      "step": 62
    },
    {
      "epoch": 0.005391412250486725,
      "grad_norm": 0.7669386863708496,
      "learning_rate": 9.098702414536107e-05,
      "loss": 0.0712,
      "step": 63
    },
    {
      "epoch": 0.0054769902227166725,
      "grad_norm": 0.8523790240287781,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.07,
      "step": 64
    },
    {
      "epoch": 0.005562568194946621,
      "grad_norm": 0.8387200832366943,
      "learning_rate": 8.9900861364012e-05,
      "loss": 0.0788,
      "step": 65
    },
    {
      "epoch": 0.005648146167176568,
      "grad_norm": 0.5394012928009033,
      "learning_rate": 8.933724690167417e-05,
      "loss": 0.0598,
      "step": 66
    },
    {
      "epoch": 0.005733724139406517,
      "grad_norm": 0.5872977375984192,
      "learning_rate": 8.876019880555649e-05,
      "loss": 0.051,
      "step": 67
    },
    {
      "epoch": 0.005819302111636465,
      "grad_norm": 0.6690393686294556,
      "learning_rate": 8.816991413705516e-05,
      "loss": 0.0532,
      "step": 68
    },
    {
      "epoch": 0.0059048800838664125,
      "grad_norm": 0.7872377038002014,
      "learning_rate": 8.756659447784368e-05,
      "loss": 0.0512,
      "step": 69
    },
    {
      "epoch": 0.005990458056096361,
      "grad_norm": 0.693568766117096,
      "learning_rate": 8.695044586103296e-05,
      "loss": 0.0546,
      "step": 70
    },
    {
      "epoch": 0.006076036028326309,
      "grad_norm": 0.7384729981422424,
      "learning_rate": 8.632167870081121e-05,
      "loss": 0.0547,
      "step": 71
    },
    {
      "epoch": 0.006161614000556257,
      "grad_norm": 1.0216960906982422,
      "learning_rate": 8.568050772058762e-05,
      "loss": 0.0484,
      "step": 72
    },
    {
      "epoch": 0.006247191972786205,
      "grad_norm": 0.39973679184913635,
      "learning_rate": 8.502715187966455e-05,
      "loss": 0.0232,
      "step": 73
    },
    {
      "epoch": 0.006332769945016153,
      "grad_norm": 0.4455772042274475,
      "learning_rate": 8.436183429846313e-05,
      "loss": 0.0365,
      "step": 74
    },
    {
      "epoch": 0.006418347917246101,
      "grad_norm": 0.8289114832878113,
      "learning_rate": 8.368478218232787e-05,
      "loss": 0.0517,
      "step": 75
    },
    {
      "epoch": 0.006503925889476049,
      "grad_norm": 0.7261016368865967,
      "learning_rate": 8.299622674393614e-05,
      "loss": 0.0399,
      "step": 76
    },
    {
      "epoch": 0.0065895038617059965,
      "grad_norm": 1.4494256973266602,
      "learning_rate": 8.229640312433937e-05,
      "loss": 0.1227,
      "step": 77
    },
    {
      "epoch": 0.006675081833935945,
      "grad_norm": 1.0130767822265625,
      "learning_rate": 8.158555031266254e-05,
      "loss": 0.084,
      "step": 78
    },
    {
      "epoch": 0.006760659806165893,
      "grad_norm": 0.8755025863647461,
      "learning_rate": 8.086391106448965e-05,
      "loss": 0.0748,
      "step": 79
    },
    {
      "epoch": 0.006846237778395841,
      "grad_norm": 0.6819718480110168,
      "learning_rate": 8.013173181896283e-05,
      "loss": 0.0536,
      "step": 80
    },
    {
      "epoch": 0.006931815750625789,
      "grad_norm": 0.4696519076824188,
      "learning_rate": 7.938926261462366e-05,
      "loss": 0.0461,
      "step": 81
    },
    {
      "epoch": 0.007017393722855737,
      "grad_norm": 0.4552120268344879,
      "learning_rate": 7.863675700402526e-05,
      "loss": 0.0516,
      "step": 82
    },
    {
      "epoch": 0.007102971695085685,
      "grad_norm": 0.5942057967185974,
      "learning_rate": 7.787447196714427e-05,
      "loss": 0.0448,
      "step": 83
    },
    {
      "epoch": 0.007188549667315633,
      "grad_norm": 0.4820343255996704,
      "learning_rate": 7.710266782362247e-05,
      "loss": 0.0497,
      "step": 84
    },
    {
      "epoch": 0.007274127639545581,
      "grad_norm": 0.4750823676586151,
      "learning_rate": 7.63216081438678e-05,
      "loss": 0.0569,
      "step": 85
    },
    {
      "epoch": 0.007359705611775529,
      "grad_norm": 0.5007278323173523,
      "learning_rate": 7.553155965904535e-05,
      "loss": 0.053,
      "step": 86
    },
    {
      "epoch": 0.007445283584005477,
      "grad_norm": 0.6972682476043701,
      "learning_rate": 7.473279216998895e-05,
      "loss": 0.0592,
      "step": 87
    },
    {
      "epoch": 0.007530861556235425,
      "grad_norm": 0.42754825949668884,
      "learning_rate": 7.392557845506432e-05,
      "loss": 0.0512,
      "step": 88
    },
    {
      "epoch": 0.007616439528465373,
      "grad_norm": 0.5751356482505798,
      "learning_rate": 7.311019417701566e-05,
      "loss": 0.0558,
      "step": 89
    },
    {
      "epoch": 0.007702017500695321,
      "grad_norm": 0.6590269207954407,
      "learning_rate": 7.228691778882693e-05,
      "loss": 0.0548,
      "step": 90
    },
    {
      "epoch": 0.007787595472925269,
      "grad_norm": 0.6272323131561279,
      "learning_rate": 7.145603043863045e-05,
      "loss": 0.0504,
      "step": 91
    },
    {
      "epoch": 0.007873173445155217,
      "grad_norm": 0.5708761215209961,
      "learning_rate": 7.061781587369519e-05,
      "loss": 0.047,
      "step": 92
    },
    {
      "epoch": 0.007958751417385165,
      "grad_norm": 0.7379321455955505,
      "learning_rate": 6.977256034352712e-05,
      "loss": 0.0663,
      "step": 93
    },
    {
      "epoch": 0.008044329389615114,
      "grad_norm": 1.128997564315796,
      "learning_rate": 6.892055250211552e-05,
      "loss": 0.0725,
      "step": 94
    },
    {
      "epoch": 0.008129907361845061,
      "grad_norm": 0.6566733717918396,
      "learning_rate": 6.806208330935766e-05,
      "loss": 0.0436,
      "step": 95
    },
    {
      "epoch": 0.008215485334075009,
      "grad_norm": 0.4307779371738434,
      "learning_rate": 6.719744593169641e-05,
      "loss": 0.0482,
      "step": 96
    },
    {
      "epoch": 0.008301063306304956,
      "grad_norm": 0.7917209267616272,
      "learning_rate": 6.632693564200416e-05,
      "loss": 0.0312,
      "step": 97
    },
    {
      "epoch": 0.008386641278534905,
      "grad_norm": 0.6638113260269165,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.0591,
      "step": 98
    },
    {
      "epoch": 0.008472219250764853,
      "grad_norm": 0.5659735798835754,
      "learning_rate": 6.456948734446624e-05,
      "loss": 0.0691,
      "step": 99
    },
    {
      "epoch": 0.0085577972229948,
      "grad_norm": 0.8545747995376587,
      "learning_rate": 6.368314950360415e-05,
      "loss": 0.04,
      "step": 100
    },
    {
      "epoch": 0.0085577972229948,
      "eval_loss": 0.07757702469825745,
      "eval_runtime": 1604.3633,
      "eval_samples_per_second": 12.267,
      "eval_steps_per_second": 6.134,
      "step": 100
    },
    {
      "epoch": 0.00864337519522475,
      "grad_norm": 0.9993880987167358,
      "learning_rate": 6.279213887972179e-05,
      "loss": 0.086,
      "step": 101
    },
    {
      "epoch": 0.008728953167454697,
      "grad_norm": 1.159125804901123,
      "learning_rate": 6.189675975213094e-05,
      "loss": 0.084,
      "step": 102
    },
    {
      "epoch": 0.008814531139684645,
      "grad_norm": 0.4134770631790161,
      "learning_rate": 6.099731789198344e-05,
      "loss": 0.042,
      "step": 103
    },
    {
      "epoch": 0.008900109111914594,
      "grad_norm": 1.0623034238815308,
      "learning_rate": 6.009412045785051e-05,
      "loss": 0.0795,
      "step": 104
    },
    {
      "epoch": 0.008985687084144541,
      "grad_norm": 0.6767681241035461,
      "learning_rate": 5.918747589082853e-05,
      "loss": 0.0425,
      "step": 105
    },
    {
      "epoch": 0.009071265056374489,
      "grad_norm": 0.7142776250839233,
      "learning_rate": 5.82776938092065e-05,
      "loss": 0.0535,
      "step": 106
    },
    {
      "epoch": 0.009156843028604438,
      "grad_norm": 0.4814240634441376,
      "learning_rate": 5.736508490273188e-05,
      "loss": 0.0478,
      "step": 107
    },
    {
      "epoch": 0.009242421000834385,
      "grad_norm": 0.42527705430984497,
      "learning_rate": 5.644996082651017e-05,
      "loss": 0.042,
      "step": 108
    },
    {
      "epoch": 0.009327998973064333,
      "grad_norm": 0.7720891833305359,
      "learning_rate": 5.553263409457504e-05,
      "loss": 0.0798,
      "step": 109
    },
    {
      "epoch": 0.009413576945294282,
      "grad_norm": 0.8637982606887817,
      "learning_rate": 5.4613417973165106e-05,
      "loss": 0.0495,
      "step": 110
    },
    {
      "epoch": 0.00949915491752423,
      "grad_norm": 0.5723322033882141,
      "learning_rate": 5.3692626373743706e-05,
      "loss": 0.0454,
      "step": 111
    },
    {
      "epoch": 0.009584732889754177,
      "grad_norm": 0.3151332139968872,
      "learning_rate": 5.27705737457985e-05,
      "loss": 0.0311,
      "step": 112
    },
    {
      "epoch": 0.009670310861984126,
      "grad_norm": 0.5164255499839783,
      "learning_rate": 5.184757496945726e-05,
      "loss": 0.0498,
      "step": 113
    },
    {
      "epoch": 0.009755888834214074,
      "grad_norm": 0.6250416040420532,
      "learning_rate": 5.092394524795649e-05,
      "loss": 0.0428,
      "step": 114
    },
    {
      "epoch": 0.009841466806444021,
      "grad_norm": 0.9330922365188599,
      "learning_rate": 5e-05,
      "loss": 0.0583,
      "step": 115
    },
    {
      "epoch": 0.009927044778673969,
      "grad_norm": 0.4849117398262024,
      "learning_rate": 4.907605475204352e-05,
      "loss": 0.0445,
      "step": 116
    },
    {
      "epoch": 0.010012622750903918,
      "grad_norm": 0.43486136198043823,
      "learning_rate": 4.8152425030542766e-05,
      "loss": 0.0297,
      "step": 117
    },
    {
      "epoch": 0.010098200723133865,
      "grad_norm": 0.35892051458358765,
      "learning_rate": 4.72294262542015e-05,
      "loss": 0.0324,
      "step": 118
    },
    {
      "epoch": 0.010183778695363813,
      "grad_norm": 0.890055775642395,
      "learning_rate": 4.6307373626256306e-05,
      "loss": 0.0361,
      "step": 119
    },
    {
      "epoch": 0.010269356667593762,
      "grad_norm": 0.8304891586303711,
      "learning_rate": 4.5386582026834906e-05,
      "loss": 0.036,
      "step": 120
    },
    {
      "epoch": 0.01035493463982371,
      "grad_norm": 0.7514469623565674,
      "learning_rate": 4.446736590542497e-05,
      "loss": 0.0541,
      "step": 121
    },
    {
      "epoch": 0.010440512612053657,
      "grad_norm": 1.3039171695709229,
      "learning_rate": 4.3550039173489845e-05,
      "loss": 0.0553,
      "step": 122
    },
    {
      "epoch": 0.010526090584283606,
      "grad_norm": 1.1074358224868774,
      "learning_rate": 4.2634915097268115e-05,
      "loss": 0.0578,
      "step": 123
    },
    {
      "epoch": 0.010611668556513553,
      "grad_norm": 0.46257588267326355,
      "learning_rate": 4.1722306190793495e-05,
      "loss": 0.0214,
      "step": 124
    },
    {
      "epoch": 0.010697246528743501,
      "grad_norm": 0.32463976740837097,
      "learning_rate": 4.0812524109171476e-05,
      "loss": 0.0165,
      "step": 125
    },
    {
      "epoch": 0.01078282450097345,
      "grad_norm": 0.5336787104606628,
      "learning_rate": 3.99058795421495e-05,
      "loss": 0.0467,
      "step": 126
    },
    {
      "epoch": 0.010868402473203398,
      "grad_norm": 0.5016961693763733,
      "learning_rate": 3.9002682108016585e-05,
      "loss": 0.0514,
      "step": 127
    },
    {
      "epoch": 0.010953980445433345,
      "grad_norm": 0.3432494103908539,
      "learning_rate": 3.8103240247869075e-05,
      "loss": 0.0376,
      "step": 128
    },
    {
      "epoch": 0.011039558417663294,
      "grad_norm": 0.4820862114429474,
      "learning_rate": 3.720786112027822e-05,
      "loss": 0.0597,
      "step": 129
    },
    {
      "epoch": 0.011125136389893242,
      "grad_norm": 0.671643078327179,
      "learning_rate": 3.631685049639586e-05,
      "loss": 0.0729,
      "step": 130
    },
    {
      "epoch": 0.01121071436212319,
      "grad_norm": 0.5347462296485901,
      "learning_rate": 3.543051265553377e-05,
      "loss": 0.0525,
      "step": 131
    },
    {
      "epoch": 0.011296292334353137,
      "grad_norm": 1.0950725078582764,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.0368,
      "step": 132
    },
    {
      "epoch": 0.011381870306583086,
      "grad_norm": 0.4670153260231018,
      "learning_rate": 3.367306435799584e-05,
      "loss": 0.0474,
      "step": 133
    },
    {
      "epoch": 0.011467448278813033,
      "grad_norm": 0.4681526720523834,
      "learning_rate": 3.2802554068303596e-05,
      "loss": 0.0542,
      "step": 134
    },
    {
      "epoch": 0.01155302625104298,
      "grad_norm": 0.5600285530090332,
      "learning_rate": 3.1937916690642356e-05,
      "loss": 0.0402,
      "step": 135
    },
    {
      "epoch": 0.01163860422327293,
      "grad_norm": 0.48455125093460083,
      "learning_rate": 3.107944749788449e-05,
      "loss": 0.036,
      "step": 136
    },
    {
      "epoch": 0.011724182195502877,
      "grad_norm": 0.39404425024986267,
      "learning_rate": 3.0227439656472877e-05,
      "loss": 0.0412,
      "step": 137
    },
    {
      "epoch": 0.011809760167732825,
      "grad_norm": 0.4101657569408417,
      "learning_rate": 2.9382184126304834e-05,
      "loss": 0.0339,
      "step": 138
    },
    {
      "epoch": 0.011895338139962774,
      "grad_norm": 0.4828612506389618,
      "learning_rate": 2.8543969561369556e-05,
      "loss": 0.0467,
      "step": 139
    },
    {
      "epoch": 0.011980916112192722,
      "grad_norm": 1.045760989189148,
      "learning_rate": 2.771308221117309e-05,
      "loss": 0.0765,
      "step": 140
    },
    {
      "epoch": 0.012066494084422669,
      "grad_norm": 0.2914433479309082,
      "learning_rate": 2.688980582298435e-05,
      "loss": 0.0373,
      "step": 141
    },
    {
      "epoch": 0.012152072056652618,
      "grad_norm": 0.38904571533203125,
      "learning_rate": 2.607442154493568e-05,
      "loss": 0.0552,
      "step": 142
    },
    {
      "epoch": 0.012237650028882566,
      "grad_norm": 0.3400273621082306,
      "learning_rate": 2.5267207830011068e-05,
      "loss": 0.0261,
      "step": 143
    },
    {
      "epoch": 0.012323228001112513,
      "grad_norm": 0.2726489007472992,
      "learning_rate": 2.446844034095466e-05,
      "loss": 0.0198,
      "step": 144
    },
    {
      "epoch": 0.012408805973342462,
      "grad_norm": 0.2568134367465973,
      "learning_rate": 2.3678391856132204e-05,
      "loss": 0.0307,
      "step": 145
    },
    {
      "epoch": 0.01249438394557241,
      "grad_norm": 0.47352564334869385,
      "learning_rate": 2.2897332176377528e-05,
      "loss": 0.0287,
      "step": 146
    },
    {
      "epoch": 0.012579961917802357,
      "grad_norm": 0.5932596921920776,
      "learning_rate": 2.2125528032855724e-05,
      "loss": 0.0461,
      "step": 147
    },
    {
      "epoch": 0.012665539890032306,
      "grad_norm": 0.4587341845035553,
      "learning_rate": 2.136324299597474e-05,
      "loss": 0.0338,
      "step": 148
    },
    {
      "epoch": 0.012751117862262254,
      "grad_norm": 0.5234631896018982,
      "learning_rate": 2.061073738537635e-05,
      "loss": 0.0197,
      "step": 149
    },
    {
      "epoch": 0.012836695834492201,
      "grad_norm": 0.5870745778083801,
      "learning_rate": 1.9868268181037185e-05,
      "loss": 0.0295,
      "step": 150
    },
    {
      "epoch": 0.012836695834492201,
      "eval_loss": 0.07830318063497543,
      "eval_runtime": 1605.9776,
      "eval_samples_per_second": 12.255,
      "eval_steps_per_second": 6.128,
      "step": 150
    },
    {
      "epoch": 0.012922273806722149,
      "grad_norm": 0.8338009715080261,
      "learning_rate": 1.9136088935510362e-05,
      "loss": 0.0953,
      "step": 151
    },
    {
      "epoch": 0.013007851778952098,
      "grad_norm": 1.0214147567749023,
      "learning_rate": 1.8414449687337464e-05,
      "loss": 0.092,
      "step": 152
    },
    {
      "epoch": 0.013093429751182046,
      "grad_norm": 1.2112953662872314,
      "learning_rate": 1.7703596875660645e-05,
      "loss": 0.141,
      "step": 153
    },
    {
      "epoch": 0.013179007723411993,
      "grad_norm": 1.1335638761520386,
      "learning_rate": 1.700377325606388e-05,
      "loss": 0.1125,
      "step": 154
    },
    {
      "epoch": 0.013264585695641942,
      "grad_norm": 0.7966949939727783,
      "learning_rate": 1.631521781767214e-05,
      "loss": 0.0562,
      "step": 155
    },
    {
      "epoch": 0.01335016366787189,
      "grad_norm": 0.9197629690170288,
      "learning_rate": 1.5638165701536868e-05,
      "loss": 0.0866,
      "step": 156
    },
    {
      "epoch": 0.013435741640101837,
      "grad_norm": 0.6064220666885376,
      "learning_rate": 1.4972848120335453e-05,
      "loss": 0.0479,
      "step": 157
    },
    {
      "epoch": 0.013521319612331786,
      "grad_norm": 0.2772460877895355,
      "learning_rate": 1.4319492279412388e-05,
      "loss": 0.0219,
      "step": 158
    },
    {
      "epoch": 0.013606897584561734,
      "grad_norm": 0.36105939745903015,
      "learning_rate": 1.3678321299188801e-05,
      "loss": 0.0405,
      "step": 159
    },
    {
      "epoch": 0.013692475556791681,
      "grad_norm": 0.28316134214401245,
      "learning_rate": 1.3049554138967051e-05,
      "loss": 0.0354,
      "step": 160
    },
    {
      "epoch": 0.01377805352902163,
      "grad_norm": 0.4974602460861206,
      "learning_rate": 1.2433405522156332e-05,
      "loss": 0.0393,
      "step": 161
    },
    {
      "epoch": 0.013863631501251578,
      "grad_norm": 0.8241268992424011,
      "learning_rate": 1.183008586294485e-05,
      "loss": 0.0663,
      "step": 162
    },
    {
      "epoch": 0.013949209473481525,
      "grad_norm": 0.7367086410522461,
      "learning_rate": 1.1239801194443506e-05,
      "loss": 0.0417,
      "step": 163
    },
    {
      "epoch": 0.014034787445711475,
      "grad_norm": 0.6524820923805237,
      "learning_rate": 1.066275309832584e-05,
      "loss": 0.059,
      "step": 164
    },
    {
      "epoch": 0.014120365417941422,
      "grad_norm": 0.48198258876800537,
      "learning_rate": 1.0099138635988026e-05,
      "loss": 0.0432,
      "step": 165
    },
    {
      "epoch": 0.01420594339017137,
      "grad_norm": 0.6708958148956299,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.056,
      "step": 166
    },
    {
      "epoch": 0.014291521362401319,
      "grad_norm": 0.6288459897041321,
      "learning_rate": 9.012975854638949e-06,
      "loss": 0.0465,
      "step": 167
    },
    {
      "epoch": 0.014377099334631266,
      "grad_norm": 0.47757601737976074,
      "learning_rate": 8.490798459222476e-06,
      "loss": 0.0461,
      "step": 168
    },
    {
      "epoch": 0.014462677306861214,
      "grad_norm": 0.47537004947662354,
      "learning_rate": 7.982796418105371e-06,
      "loss": 0.0444,
      "step": 169
    },
    {
      "epoch": 0.014548255279091161,
      "grad_norm": 0.27161651849746704,
      "learning_rate": 7.489143213519301e-06,
      "loss": 0.0266,
      "step": 170
    },
    {
      "epoch": 0.01463383325132111,
      "grad_norm": 0.3608476221561432,
      "learning_rate": 7.010007427581378e-06,
      "loss": 0.0257,
      "step": 171
    },
    {
      "epoch": 0.014719411223551058,
      "grad_norm": 0.34825852513313293,
      "learning_rate": 6.5455526847235825e-06,
      "loss": 0.0376,
      "step": 172
    },
    {
      "epoch": 0.014804989195781005,
      "grad_norm": 0.38921675086021423,
      "learning_rate": 6.0959375958151045e-06,
      "loss": 0.0284,
      "step": 173
    },
    {
      "epoch": 0.014890567168010954,
      "grad_norm": 0.3063510060310364,
      "learning_rate": 5.6613157039969055e-06,
      "loss": 0.0246,
      "step": 174
    },
    {
      "epoch": 0.014976145140240902,
      "grad_norm": 0.3037574291229248,
      "learning_rate": 5.241835432246889e-06,
      "loss": 0.0357,
      "step": 175
    },
    {
      "epoch": 0.01506172311247085,
      "grad_norm": 0.36166590452194214,
      "learning_rate": 4.837640032693558e-06,
      "loss": 0.0404,
      "step": 176
    },
    {
      "epoch": 0.015147301084700799,
      "grad_norm": 0.4001633822917938,
      "learning_rate": 4.448867537695578e-06,
      "loss": 0.0462,
      "step": 177
    },
    {
      "epoch": 0.015232879056930746,
      "grad_norm": 0.5758128762245178,
      "learning_rate": 4.075650712703849e-06,
      "loss": 0.0447,
      "step": 178
    },
    {
      "epoch": 0.015318457029160694,
      "grad_norm": 0.4326450228691101,
      "learning_rate": 3.71811701092219e-06,
      "loss": 0.0437,
      "step": 179
    },
    {
      "epoch": 0.015404035001390643,
      "grad_norm": 0.3488145172595978,
      "learning_rate": 3.376388529782215e-06,
      "loss": 0.0392,
      "step": 180
    },
    {
      "epoch": 0.01548961297362059,
      "grad_norm": 0.4991649091243744,
      "learning_rate": 3.0505819692471792e-06,
      "loss": 0.04,
      "step": 181
    },
    {
      "epoch": 0.015575190945850538,
      "grad_norm": 0.3567849397659302,
      "learning_rate": 2.7408085919590264e-06,
      "loss": 0.0447,
      "step": 182
    },
    {
      "epoch": 0.015660768918080485,
      "grad_norm": 0.5800079107284546,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 0.0476,
      "step": 183
    },
    {
      "epoch": 0.015746346890310434,
      "grad_norm": 0.29148969054222107,
      "learning_rate": 2.1697790249779636e-06,
      "loss": 0.031,
      "step": 184
    },
    {
      "epoch": 0.015831924862540384,
      "grad_norm": 0.3116988241672516,
      "learning_rate": 1.908717841359048e-06,
      "loss": 0.0313,
      "step": 185
    },
    {
      "epoch": 0.01591750283477033,
      "grad_norm": 0.46554750204086304,
      "learning_rate": 1.6640797865406288e-06,
      "loss": 0.0401,
      "step": 186
    },
    {
      "epoch": 0.01600308080700028,
      "grad_norm": 0.3885302245616913,
      "learning_rate": 1.4359484041943038e-06,
      "loss": 0.0311,
      "step": 187
    },
    {
      "epoch": 0.016088658779230228,
      "grad_norm": 0.38669371604919434,
      "learning_rate": 1.2244016009781701e-06,
      "loss": 0.042,
      "step": 188
    },
    {
      "epoch": 0.016174236751460173,
      "grad_norm": 0.3348408639431,
      "learning_rate": 1.0295116199317057e-06,
      "loss": 0.0239,
      "step": 189
    },
    {
      "epoch": 0.016259814723690123,
      "grad_norm": 0.666803777217865,
      "learning_rate": 8.513450158049108e-07,
      "loss": 0.0413,
      "step": 190
    },
    {
      "epoch": 0.016345392695920072,
      "grad_norm": 0.5500365495681763,
      "learning_rate": 6.899626323298713e-07,
      "loss": 0.0516,
      "step": 191
    },
    {
      "epoch": 0.016430970668150018,
      "grad_norm": 0.5613580346107483,
      "learning_rate": 5.454195814427021e-07,
      "loss": 0.0627,
      "step": 192
    },
    {
      "epoch": 0.016516548640379967,
      "grad_norm": 0.2784683406352997,
      "learning_rate": 4.177652244628627e-07,
      "loss": 0.0199,
      "step": 193
    },
    {
      "epoch": 0.016602126612609912,
      "grad_norm": 0.6985279321670532,
      "learning_rate": 3.0704315523631953e-07,
      "loss": 0.054,
      "step": 194
    },
    {
      "epoch": 0.01668770458483986,
      "grad_norm": 0.7367770671844482,
      "learning_rate": 2.1329118524827662e-07,
      "loss": 0.0643,
      "step": 195
    },
    {
      "epoch": 0.01677328255706981,
      "grad_norm": 0.35487690567970276,
      "learning_rate": 1.3654133071059893e-07,
      "loss": 0.0291,
      "step": 196
    },
    {
      "epoch": 0.016858860529299757,
      "grad_norm": 0.5007591843605042,
      "learning_rate": 7.681980162830282e-08,
      "loss": 0.03,
      "step": 197
    },
    {
      "epoch": 0.016944438501529706,
      "grad_norm": 0.3655904233455658,
      "learning_rate": 3.4146992848854695e-08,
      "loss": 0.0475,
      "step": 198
    },
    {
      "epoch": 0.017030016473759655,
      "grad_norm": 0.7564741373062134,
      "learning_rate": 8.537477097364522e-09,
      "loss": 0.04,
      "step": 199
    },
    {
      "epoch": 0.0171155944459896,
      "grad_norm": 1.8403810262680054,
      "learning_rate": 0.0,
      "loss": 0.1042,
      "step": 200
    },
    {
      "epoch": 0.0171155944459896,
      "eval_loss": 0.04161956161260605,
      "eval_runtime": 1606.9293,
      "eval_samples_per_second": 12.248,
      "eval_steps_per_second": 6.124,
      "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": 3.017027157491712e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}