{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.979702300405954,
  "eval_steps": 500,
  "global_step": 460,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010825439783491205,
      "grad_norm": 5.808831848901163,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 0.8566,
      "step": 1
    },
    {
      "epoch": 0.02165087956698241,
      "grad_norm": 5.8131650809808635,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 0.8603,
      "step": 2
    },
    {
      "epoch": 0.03247631935047361,
      "grad_norm": 5.758169356290023,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.8631,
      "step": 3
    },
    {
      "epoch": 0.04330175913396482,
      "grad_norm": 4.238914874845898,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.8176,
      "step": 4
    },
    {
      "epoch": 0.05412719891745602,
      "grad_norm": 2.235103598722892,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.7658,
      "step": 5
    },
    {
      "epoch": 0.06495263870094722,
      "grad_norm": 4.07899438012489,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.7574,
      "step": 6
    },
    {
      "epoch": 0.07577807848443843,
      "grad_norm": 4.324129456803467,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.7776,
      "step": 7
    },
    {
      "epoch": 0.08660351826792964,
      "grad_norm": 4.2756579015034015,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.7457,
      "step": 8
    },
    {
      "epoch": 0.09742895805142084,
      "grad_norm": 3.6399981122437852,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.7141,
      "step": 9
    },
    {
      "epoch": 0.10825439783491204,
      "grad_norm": 2.114757736524999,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.6986,
      "step": 10
    },
    {
      "epoch": 0.11907983761840325,
      "grad_norm": 2.407779675631392,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.6642,
      "step": 11
    },
    {
      "epoch": 0.12990527740189445,
      "grad_norm": 1.9690081792484082,
      "learning_rate": 2.0869565217391306e-05,
      "loss": 0.6566,
      "step": 12
    },
    {
      "epoch": 0.14073071718538566,
      "grad_norm": 1.2791289873731233,
      "learning_rate": 2.2608695652173914e-05,
      "loss": 0.636,
      "step": 13
    },
    {
      "epoch": 0.15155615696887687,
      "grad_norm": 1.4272246324351143,
      "learning_rate": 2.4347826086956526e-05,
      "loss": 0.6427,
      "step": 14
    },
    {
      "epoch": 0.16238159675236807,
      "grad_norm": 1.1897907235661358,
      "learning_rate": 2.608695652173913e-05,
      "loss": 0.6195,
      "step": 15
    },
    {
      "epoch": 0.17320703653585928,
      "grad_norm": 0.9139563980188694,
      "learning_rate": 2.782608695652174e-05,
      "loss": 0.6092,
      "step": 16
    },
    {
      "epoch": 0.18403247631935046,
      "grad_norm": 1.0660755035768057,
      "learning_rate": 2.956521739130435e-05,
      "loss": 0.6006,
      "step": 17
    },
    {
      "epoch": 0.19485791610284167,
      "grad_norm": 1.1289299633501881,
      "learning_rate": 3.130434782608696e-05,
      "loss": 0.5868,
      "step": 18
    },
    {
      "epoch": 0.20568335588633288,
      "grad_norm": 1.005093467371569,
      "learning_rate": 3.304347826086957e-05,
      "loss": 0.5769,
      "step": 19
    },
    {
      "epoch": 0.2165087956698241,
      "grad_norm": 1.0253086560111697,
      "learning_rate": 3.478260869565218e-05,
      "loss": 0.5798,
      "step": 20
    },
    {
      "epoch": 0.2273342354533153,
      "grad_norm": 1.4416719038843315,
      "learning_rate": 3.6521739130434786e-05,
      "loss": 0.5738,
      "step": 21
    },
    {
      "epoch": 0.2381596752368065,
      "grad_norm": 0.9795141368299348,
      "learning_rate": 3.8260869565217395e-05,
      "loss": 0.5713,
      "step": 22
    },
    {
      "epoch": 0.2489851150202977,
      "grad_norm": 0.9113506379565265,
      "learning_rate": 4e-05,
      "loss": 0.5636,
      "step": 23
    },
    {
      "epoch": 0.2598105548037889,
      "grad_norm": 1.3018876548232534,
      "learning_rate": 4.173913043478261e-05,
      "loss": 0.5703,
      "step": 24
    },
    {
      "epoch": 0.2706359945872801,
      "grad_norm": 1.0977569575275883,
      "learning_rate": 4.347826086956522e-05,
      "loss": 0.5608,
      "step": 25
    },
    {
      "epoch": 0.2814614343707713,
      "grad_norm": 1.3185220107226532,
      "learning_rate": 4.521739130434783e-05,
      "loss": 0.5383,
      "step": 26
    },
    {
      "epoch": 0.2922868741542625,
      "grad_norm": 1.4506983344294453,
      "learning_rate": 4.6956521739130444e-05,
      "loss": 0.5481,
      "step": 27
    },
    {
      "epoch": 0.30311231393775373,
      "grad_norm": 0.7383393635142479,
      "learning_rate": 4.869565217391305e-05,
      "loss": 0.5463,
      "step": 28
    },
    {
      "epoch": 0.31393775372124494,
      "grad_norm": 2.4652927181675315,
      "learning_rate": 5.043478260869566e-05,
      "loss": 0.5542,
      "step": 29
    },
    {
      "epoch": 0.32476319350473615,
      "grad_norm": 1.1746605790954319,
      "learning_rate": 5.217391304347826e-05,
      "loss": 0.5401,
      "step": 30
    },
    {
      "epoch": 0.33558863328822736,
      "grad_norm": 2.892014348770916,
      "learning_rate": 5.391304347826087e-05,
      "loss": 0.5552,
      "step": 31
    },
    {
      "epoch": 0.34641407307171856,
      "grad_norm": 2.5911400202995685,
      "learning_rate": 5.565217391304348e-05,
      "loss": 0.5547,
      "step": 32
    },
    {
      "epoch": 0.3572395128552097,
      "grad_norm": 1.688863282311232,
      "learning_rate": 5.739130434782609e-05,
      "loss": 0.5499,
      "step": 33
    },
    {
      "epoch": 0.3680649526387009,
      "grad_norm": 1.8689766358229456,
      "learning_rate": 5.91304347826087e-05,
      "loss": 0.55,
      "step": 34
    },
    {
      "epoch": 0.37889039242219213,
      "grad_norm": 1.561406905112785,
      "learning_rate": 6.086956521739131e-05,
      "loss": 0.5347,
      "step": 35
    },
    {
      "epoch": 0.38971583220568334,
      "grad_norm": 1.689162114377315,
      "learning_rate": 6.260869565217392e-05,
      "loss": 0.5264,
      "step": 36
    },
    {
      "epoch": 0.40054127198917455,
      "grad_norm": 1.444322528581452,
      "learning_rate": 6.434782608695653e-05,
      "loss": 0.5362,
      "step": 37
    },
    {
      "epoch": 0.41136671177266576,
      "grad_norm": 1.2509308676002273,
      "learning_rate": 6.608695652173914e-05,
      "loss": 0.5238,
      "step": 38
    },
    {
      "epoch": 0.42219215155615697,
      "grad_norm": 1.88113743058469,
      "learning_rate": 6.782608695652175e-05,
      "loss": 0.5311,
      "step": 39
    },
    {
      "epoch": 0.4330175913396482,
      "grad_norm": 1.5022638539817577,
      "learning_rate": 6.956521739130436e-05,
      "loss": 0.523,
      "step": 40
    },
    {
      "epoch": 0.4438430311231394,
      "grad_norm": 1.205391078251559,
      "learning_rate": 7.130434782608696e-05,
      "loss": 0.5188,
      "step": 41
    },
    {
      "epoch": 0.4546684709066306,
      "grad_norm": 1.8357524846560098,
      "learning_rate": 7.304347826086957e-05,
      "loss": 0.5239,
      "step": 42
    },
    {
      "epoch": 0.4654939106901218,
      "grad_norm": 1.3953240643163467,
      "learning_rate": 7.478260869565218e-05,
      "loss": 0.5174,
      "step": 43
    },
    {
      "epoch": 0.476319350473613,
      "grad_norm": 1.2828206105300195,
      "learning_rate": 7.652173913043479e-05,
      "loss": 0.5147,
      "step": 44
    },
    {
      "epoch": 0.4871447902571042,
      "grad_norm": 1.53155664347226,
      "learning_rate": 7.82608695652174e-05,
      "loss": 0.5243,
      "step": 45
    },
    {
      "epoch": 0.4979702300405954,
      "grad_norm": 2.0731025195597055,
      "learning_rate": 8e-05,
      "loss": 0.5176,
      "step": 46
    },
    {
      "epoch": 0.5087956698240866,
      "grad_norm": 1.297417565546252,
      "learning_rate": 7.999884833286186e-05,
      "loss": 0.5124,
      "step": 47
    },
    {
      "epoch": 0.5196211096075778,
      "grad_norm": 2.0613915660565474,
      "learning_rate": 7.999539339776424e-05,
      "loss": 0.5211,
      "step": 48
    },
    {
      "epoch": 0.530446549391069,
      "grad_norm": 1.9265188351114966,
      "learning_rate": 7.998963539365397e-05,
      "loss": 0.5144,
      "step": 49
    },
    {
      "epoch": 0.5412719891745602,
      "grad_norm": 1.1654357885632856,
      "learning_rate": 7.998157465209622e-05,
      "loss": 0.5084,
      "step": 50
    },
    {
      "epoch": 0.5520974289580515,
      "grad_norm": 3.550729874275142,
      "learning_rate": 7.997121163725553e-05,
      "loss": 0.52,
      "step": 51
    },
    {
      "epoch": 0.5629228687415426,
      "grad_norm": 2.538469141985988,
      "learning_rate": 7.995854694586911e-05,
      "loss": 0.527,
      "step": 52
    },
    {
      "epoch": 0.5737483085250338,
      "grad_norm": 2.3230308550181977,
      "learning_rate": 7.994358130721238e-05,
      "loss": 0.5125,
      "step": 53
    },
    {
      "epoch": 0.584573748308525,
      "grad_norm": 2.054696895438045,
      "learning_rate": 7.992631558305709e-05,
      "loss": 0.5239,
      "step": 54
    },
    {
      "epoch": 0.5953991880920162,
      "grad_norm": 1.6692562446454873,
      "learning_rate": 7.990675076762158e-05,
      "loss": 0.5054,
      "step": 55
    },
    {
      "epoch": 0.6062246278755075,
      "grad_norm": 1.6216758113364556,
      "learning_rate": 7.988488798751358e-05,
      "loss": 0.5106,
      "step": 56
    },
    {
      "epoch": 0.6170500676589986,
      "grad_norm": 1.2978731074509846,
      "learning_rate": 7.986072850166538e-05,
      "loss": 0.5065,
      "step": 57
    },
    {
      "epoch": 0.6278755074424899,
      "grad_norm": 1.5652885560156862,
      "learning_rate": 7.983427370126126e-05,
      "loss": 0.5062,
      "step": 58
    },
    {
      "epoch": 0.638700947225981,
      "grad_norm": 1.0902975831854205,
      "learning_rate": 7.980552510965744e-05,
      "loss": 0.4905,
      "step": 59
    },
    {
      "epoch": 0.6495263870094723,
      "grad_norm": 1.7790014999967663,
      "learning_rate": 7.977448438229435e-05,
      "loss": 0.5013,
      "step": 60
    },
    {
      "epoch": 0.6603518267929634,
      "grad_norm": 1.1023742875049254,
      "learning_rate": 7.974115330660125e-05,
      "loss": 0.5107,
      "step": 61
    },
    {
      "epoch": 0.6711772665764547,
      "grad_norm": 1.6569283779226378,
      "learning_rate": 7.970553380189338e-05,
      "loss": 0.5004,
      "step": 62
    },
    {
      "epoch": 0.6820027063599459,
      "grad_norm": 1.1231904813496703,
      "learning_rate": 7.966762791926138e-05,
      "loss": 0.5112,
      "step": 63
    },
    {
      "epoch": 0.6928281461434371,
      "grad_norm": 1.3259929729824094,
      "learning_rate": 7.962743784145323e-05,
      "loss": 0.5033,
      "step": 64
    },
    {
      "epoch": 0.7036535859269283,
      "grad_norm": 1.5905266367527235,
      "learning_rate": 7.958496588274853e-05,
      "loss": 0.4994,
      "step": 65
    },
    {
      "epoch": 0.7144790257104194,
      "grad_norm": 1.3556731999712488,
      "learning_rate": 7.954021448882522e-05,
      "loss": 0.5014,
      "step": 66
    },
    {
      "epoch": 0.7253044654939107,
      "grad_norm": 1.3046471245382867,
      "learning_rate": 7.94931862366188e-05,
      "loss": 0.4988,
      "step": 67
    },
    {
      "epoch": 0.7361299052774019,
      "grad_norm": 1.1190463003547946,
      "learning_rate": 7.944388383417388e-05,
      "loss": 0.5078,
      "step": 68
    },
    {
      "epoch": 0.7469553450608931,
      "grad_norm": 0.9403691436043241,
      "learning_rate": 7.939231012048833e-05,
      "loss": 0.497,
      "step": 69
    },
    {
      "epoch": 0.7577807848443843,
      "grad_norm": 1.2389417771472975,
      "learning_rate": 7.93384680653497e-05,
      "loss": 0.4826,
      "step": 70
    },
    {
      "epoch": 0.7686062246278755,
      "grad_norm": 1.4827310111058076,
      "learning_rate": 7.928236076916423e-05,
      "loss": 0.4856,
      "step": 71
    },
    {
      "epoch": 0.7794316644113667,
      "grad_norm": 1.0266845521857724,
      "learning_rate": 7.922399146277846e-05,
      "loss": 0.4953,
      "step": 72
    },
    {
      "epoch": 0.790257104194858,
      "grad_norm": 1.5291553894877223,
      "learning_rate": 7.916336350729293e-05,
      "loss": 0.5043,
      "step": 73
    },
    {
      "epoch": 0.8010825439783491,
      "grad_norm": 1.8339943841994815,
      "learning_rate": 7.910048039386885e-05,
      "loss": 0.4982,
      "step": 74
    },
    {
      "epoch": 0.8119079837618404,
      "grad_norm": 0.9242986864591416,
      "learning_rate": 7.903534574352698e-05,
      "loss": 0.4924,
      "step": 75
    },
    {
      "epoch": 0.8227334235453315,
      "grad_norm": 2.1145786188124056,
      "learning_rate": 7.896796330693917e-05,
      "loss": 0.4941,
      "step": 76
    },
    {
      "epoch": 0.8335588633288228,
      "grad_norm": 1.4704175209963846,
      "learning_rate": 7.889833696421229e-05,
      "loss": 0.4905,
      "step": 77
    },
    {
      "epoch": 0.8443843031123139,
      "grad_norm": 1.642621714596807,
      "learning_rate": 7.882647072466488e-05,
      "loss": 0.4949,
      "step": 78
    },
    {
      "epoch": 0.8552097428958051,
      "grad_norm": 1.6556981762458693,
      "learning_rate": 7.875236872659626e-05,
      "loss": 0.4956,
      "step": 79
    },
    {
      "epoch": 0.8660351826792964,
      "grad_norm": 1.2644916442646428,
      "learning_rate": 7.867603523704825e-05,
      "loss": 0.4891,
      "step": 80
    },
    {
      "epoch": 0.8768606224627875,
      "grad_norm": 1.2082452092510187,
      "learning_rate": 7.859747465155941e-05,
      "loss": 0.4983,
      "step": 81
    },
    {
      "epoch": 0.8876860622462788,
      "grad_norm": 1.7000626612415155,
      "learning_rate": 7.851669149391198e-05,
      "loss": 0.4853,
      "step": 82
    },
    {
      "epoch": 0.8985115020297699,
      "grad_norm": 0.9763637711052663,
      "learning_rate": 7.843369041587135e-05,
      "loss": 0.4899,
      "step": 83
    },
    {
      "epoch": 0.9093369418132612,
      "grad_norm": 1.8287685880691984,
      "learning_rate": 7.834847619691823e-05,
      "loss": 0.4921,
      "step": 84
    },
    {
      "epoch": 0.9201623815967523,
      "grad_norm": 1.14896663354127,
      "learning_rate": 7.82610537439734e-05,
      "loss": 0.491,
      "step": 85
    },
    {
      "epoch": 0.9309878213802436,
      "grad_norm": 1.3967407140368298,
      "learning_rate": 7.817142809111516e-05,
      "loss": 0.4927,
      "step": 86
    },
    {
      "epoch": 0.9418132611637348,
      "grad_norm": 1.1259520998597332,
      "learning_rate": 7.807960439928948e-05,
      "loss": 0.4901,
      "step": 87
    },
    {
      "epoch": 0.952638700947226,
      "grad_norm": 1.1407620174965618,
      "learning_rate": 7.798558795601278e-05,
      "loss": 0.4727,
      "step": 88
    },
    {
      "epoch": 0.9634641407307172,
      "grad_norm": 1.3691640356867825,
      "learning_rate": 7.788938417506746e-05,
      "loss": 0.4909,
      "step": 89
    },
    {
      "epoch": 0.9742895805142084,
      "grad_norm": 1.0738641081237286,
      "learning_rate": 7.779099859619017e-05,
      "loss": 0.486,
      "step": 90
    },
    {
      "epoch": 0.9851150202976996,
      "grad_norm": 1.4806199400962288,
      "learning_rate": 7.769043688475283e-05,
      "loss": 0.4874,
      "step": 91
    },
    {
      "epoch": 0.9959404600811907,
      "grad_norm": 1.1066743457942432,
      "learning_rate": 7.758770483143634e-05,
      "loss": 0.4911,
      "step": 92
    },
    {
      "epoch": 1.006765899864682,
      "grad_norm": 1.4950362727418445,
      "learning_rate": 7.748280835189721e-05,
      "loss": 0.7639,
      "step": 93
    },
    {
      "epoch": 1.0175913396481733,
      "grad_norm": 1.7271457672573984,
      "learning_rate": 7.737575348642685e-05,
      "loss": 0.4898,
      "step": 94
    },
    {
      "epoch": 1.0284167794316643,
      "grad_norm": 1.3798243699861912,
      "learning_rate": 7.726654639960379e-05,
      "loss": 0.4714,
      "step": 95
    },
    {
      "epoch": 1.0392422192151556,
      "grad_norm": 1.0339630311729218,
      "learning_rate": 7.715519337993869e-05,
      "loss": 0.4714,
      "step": 96
    },
    {
      "epoch": 1.0500676589986468,
      "grad_norm": 1.4631122991865602,
      "learning_rate": 7.704170083951221e-05,
      "loss": 0.4684,
      "step": 97
    },
    {
      "epoch": 1.060893098782138,
      "grad_norm": 1.0376053642460665,
      "learning_rate": 7.692607531360583e-05,
      "loss": 0.4741,
      "step": 98
    },
    {
      "epoch": 1.0717185385656292,
      "grad_norm": 1.4499129358788265,
      "learning_rate": 7.680832346032547e-05,
      "loss": 0.4738,
      "step": 99
    },
    {
      "epoch": 1.0825439783491204,
      "grad_norm": 0.9495063833446171,
      "learning_rate": 7.668845206021812e-05,
      "loss": 0.4652,
      "step": 100
    },
    {
      "epoch": 1.0933694181326117,
      "grad_norm": 1.3963825686162348,
      "learning_rate": 7.65664680158814e-05,
      "loss": 0.4699,
      "step": 101
    },
    {
      "epoch": 1.104194857916103,
      "grad_norm": 1.1180437569666657,
      "learning_rate": 7.644237835156608e-05,
      "loss": 0.4684,
      "step": 102
    },
    {
      "epoch": 1.115020297699594,
      "grad_norm": 1.0782036495140483,
      "learning_rate": 7.631619021277156e-05,
      "loss": 0.4678,
      "step": 103
    },
    {
      "epoch": 1.1258457374830853,
      "grad_norm": 1.1233442468927228,
      "learning_rate": 7.618791086583451e-05,
      "loss": 0.4596,
      "step": 104
    },
    {
      "epoch": 1.1366711772665765,
      "grad_norm": 0.8610664763636607,
      "learning_rate": 7.605754769751033e-05,
      "loss": 0.4602,
      "step": 105
    },
    {
      "epoch": 1.1474966170500678,
      "grad_norm": 0.9718815741535982,
      "learning_rate": 7.592510821454785e-05,
      "loss": 0.4664,
      "step": 106
    },
    {
      "epoch": 1.1583220568335588,
      "grad_norm": 0.9229650455953634,
      "learning_rate": 7.579060004325712e-05,
      "loss": 0.4588,
      "step": 107
    },
    {
      "epoch": 1.16914749661705,
      "grad_norm": 1.3303689079136494,
      "learning_rate": 7.565403092907015e-05,
      "loss": 0.4567,
      "step": 108
    },
    {
      "epoch": 1.1799729364005414,
      "grad_norm": 0.6851116777600401,
      "learning_rate": 7.551540873609502e-05,
      "loss": 0.4586,
      "step": 109
    },
    {
      "epoch": 1.1907983761840324,
      "grad_norm": 0.6799287282219464,
      "learning_rate": 7.53747414466629e-05,
      "loss": 0.4618,
      "step": 110
    },
    {
      "epoch": 1.2016238159675237,
      "grad_norm": 0.8466498529896461,
      "learning_rate": 7.523203716086855e-05,
      "loss": 0.462,
      "step": 111
    },
    {
      "epoch": 1.212449255751015,
      "grad_norm": 1.117632861579111,
      "learning_rate": 7.508730409610377e-05,
      "loss": 0.4569,
      "step": 112
    },
    {
      "epoch": 1.2232746955345062,
      "grad_norm": 1.23247753401417,
      "learning_rate": 7.494055058658431e-05,
      "loss": 0.4522,
      "step": 113
    },
    {
      "epoch": 1.2341001353179972,
      "grad_norm": 0.566030648541157,
      "learning_rate": 7.479178508286987e-05,
      "loss": 0.4564,
      "step": 114
    },
    {
      "epoch": 1.2449255751014885,
      "grad_norm": 1.058284756916062,
      "learning_rate": 7.464101615137756e-05,
      "loss": 0.4638,
      "step": 115
    },
    {
      "epoch": 1.2557510148849798,
      "grad_norm": 1.1303301791005846,
      "learning_rate": 7.448825247388854e-05,
      "loss": 0.4564,
      "step": 116
    },
    {
      "epoch": 1.266576454668471,
      "grad_norm": 0.7983550091865995,
      "learning_rate": 7.433350284704822e-05,
      "loss": 0.4509,
      "step": 117
    },
    {
      "epoch": 1.277401894451962,
      "grad_norm": 0.7407469605628687,
      "learning_rate": 7.417677618185955e-05,
      "loss": 0.4515,
      "step": 118
    },
    {
      "epoch": 1.2882273342354533,
      "grad_norm": 1.2023281817243203,
      "learning_rate": 7.401808150317005e-05,
      "loss": 0.4594,
      "step": 119
    },
    {
      "epoch": 1.2990527740189446,
      "grad_norm": 1.189203037612418,
      "learning_rate": 7.385742794915205e-05,
      "loss": 0.4488,
      "step": 120
    },
    {
      "epoch": 1.3098782138024356,
      "grad_norm": 0.7803369478691763,
      "learning_rate": 7.369482477077647e-05,
      "loss": 0.4538,
      "step": 121
    },
    {
      "epoch": 1.320703653585927,
      "grad_norm": 0.7956495942708723,
      "learning_rate": 7.353028133128018e-05,
      "loss": 0.4508,
      "step": 122
    },
    {
      "epoch": 1.3315290933694182,
      "grad_norm": 0.7174722493029648,
      "learning_rate": 7.336380710562678e-05,
      "loss": 0.4541,
      "step": 123
    },
    {
      "epoch": 1.3423545331529092,
      "grad_norm": 1.232102677986553,
      "learning_rate": 7.319541167996102e-05,
      "loss": 0.4516,
      "step": 124
    },
    {
      "epoch": 1.3531799729364005,
      "grad_norm": 1.0800843680461558,
      "learning_rate": 7.302510475105679e-05,
      "loss": 0.4501,
      "step": 125
    },
    {
      "epoch": 1.3640054127198917,
      "grad_norm": 0.5894796400207796,
      "learning_rate": 7.285289612575876e-05,
      "loss": 0.4483,
      "step": 126
    },
    {
      "epoch": 1.374830852503383,
      "grad_norm": 1.1029047763873172,
      "learning_rate": 7.267879572041768e-05,
      "loss": 0.4419,
      "step": 127
    },
    {
      "epoch": 1.3856562922868743,
      "grad_norm": 1.0991239671136233,
      "learning_rate": 7.250281356031932e-05,
      "loss": 0.4444,
      "step": 128
    },
    {
      "epoch": 1.3964817320703653,
      "grad_norm": 0.6037885510659295,
      "learning_rate": 7.232495977910722e-05,
      "loss": 0.4429,
      "step": 129
    },
    {
      "epoch": 1.4073071718538566,
      "grad_norm": 0.8683412309547249,
      "learning_rate": 7.214524461819913e-05,
      "loss": 0.4474,
      "step": 130
    },
    {
      "epoch": 1.4181326116373478,
      "grad_norm": 1.0334636806232549,
      "learning_rate": 7.196367842619732e-05,
      "loss": 0.4488,
      "step": 131
    },
    {
      "epoch": 1.4289580514208389,
      "grad_norm": 0.6466646925111661,
      "learning_rate": 7.178027165829259e-05,
      "loss": 0.4491,
      "step": 132
    },
    {
      "epoch": 1.4397834912043301,
      "grad_norm": 0.5372317843885743,
      "learning_rate": 7.159503487566236e-05,
      "loss": 0.4429,
      "step": 133
    },
    {
      "epoch": 1.4506089309878214,
      "grad_norm": 0.8834867060322021,
      "learning_rate": 7.140797874486237e-05,
      "loss": 0.4452,
      "step": 134
    },
    {
      "epoch": 1.4614343707713127,
      "grad_norm": 1.2604508969646684,
      "learning_rate": 7.121911403721254e-05,
      "loss": 0.4468,
      "step": 135
    },
    {
      "epoch": 1.472259810554804,
      "grad_norm": 0.880677105803717,
      "learning_rate": 7.10284516281768e-05,
      "loss": 0.444,
      "step": 136
    },
    {
      "epoch": 1.483085250338295,
      "grad_norm": 1.3371545062212136,
      "learning_rate": 7.083600249673665e-05,
      "loss": 0.4433,
      "step": 137
    },
    {
      "epoch": 1.4939106901217862,
      "grad_norm": 0.7924984552860372,
      "learning_rate": 7.064177772475912e-05,
      "loss": 0.4475,
      "step": 138
    },
    {
      "epoch": 1.5047361299052775,
      "grad_norm": 0.9612014933614488,
      "learning_rate": 7.04457884963586e-05,
      "loss": 0.4468,
      "step": 139
    },
    {
      "epoch": 1.5155615696887685,
      "grad_norm": 0.9392541492862665,
      "learning_rate": 7.024804609725275e-05,
      "loss": 0.4464,
      "step": 140
    },
    {
      "epoch": 1.5263870094722598,
      "grad_norm": 0.9060567297197456,
      "learning_rate": 7.004856191411271e-05,
      "loss": 0.4432,
      "step": 141
    },
    {
      "epoch": 1.537212449255751,
      "grad_norm": 1.1369625648430384,
      "learning_rate": 6.984734743390743e-05,
      "loss": 0.4461,
      "step": 142
    },
    {
      "epoch": 1.548037889039242,
      "grad_norm": 0.7931097580196472,
      "learning_rate": 6.964441424324211e-05,
      "loss": 0.4498,
      "step": 143
    },
    {
      "epoch": 1.5588633288227334,
      "grad_norm": 0.6650867960347666,
      "learning_rate": 6.94397740276911e-05,
      "loss": 0.4413,
      "step": 144
    },
    {
      "epoch": 1.5696887686062246,
      "grad_norm": 0.6886753142567965,
      "learning_rate": 6.923343857112497e-05,
      "loss": 0.4387,
      "step": 145
    },
    {
      "epoch": 1.5805142083897157,
      "grad_norm": 0.805662884063706,
      "learning_rate": 6.902541975503196e-05,
      "loss": 0.4397,
      "step": 146
    },
    {
      "epoch": 1.5913396481732072,
      "grad_norm": 1.0626477516142214,
      "learning_rate": 6.881572955783382e-05,
      "loss": 0.4368,
      "step": 147
    },
    {
      "epoch": 1.6021650879566982,
      "grad_norm": 1.5615470459694212,
      "learning_rate": 6.860438005419598e-05,
      "loss": 0.4574,
      "step": 148
    },
    {
      "epoch": 1.6129905277401895,
      "grad_norm": 0.5178061477784288,
      "learning_rate": 6.839138341433236e-05,
      "loss": 0.4374,
      "step": 149
    },
    {
      "epoch": 1.6238159675236807,
      "grad_norm": 1.546889550067696,
      "learning_rate": 6.81767519033045e-05,
      "loss": 0.4497,
      "step": 150
    },
    {
      "epoch": 1.6346414073071718,
      "grad_norm": 0.9730507254509676,
      "learning_rate": 6.796049788031528e-05,
      "loss": 0.4449,
      "step": 151
    },
    {
      "epoch": 1.645466847090663,
      "grad_norm": 0.9982813699267655,
      "learning_rate": 6.774263379799732e-05,
      "loss": 0.4432,
      "step": 152
    },
    {
      "epoch": 1.6562922868741543,
      "grad_norm": 0.9179776054358382,
      "learning_rate": 6.752317220169581e-05,
      "loss": 0.4549,
      "step": 153
    },
    {
      "epoch": 1.6671177266576453,
      "grad_norm": 0.7434050468751098,
      "learning_rate": 6.730212572874618e-05,
      "loss": 0.4448,
      "step": 154
    },
    {
      "epoch": 1.6779431664411368,
      "grad_norm": 0.8630643785039801,
      "learning_rate": 6.707950710774636e-05,
      "loss": 0.4389,
      "step": 155
    },
    {
      "epoch": 1.6887686062246279,
      "grad_norm": 0.95594617069569,
      "learning_rate": 6.685532915782388e-05,
      "loss": 0.446,
      "step": 156
    },
    {
      "epoch": 1.699594046008119,
      "grad_norm": 0.6787389488446376,
      "learning_rate": 6.662960478789762e-05,
      "loss": 0.4308,
      "step": 157
    },
    {
      "epoch": 1.7104194857916104,
      "grad_norm": 0.7138902871017276,
      "learning_rate": 6.640234699593457e-05,
      "loss": 0.4422,
      "step": 158
    },
    {
      "epoch": 1.7212449255751014,
      "grad_norm": 0.8555364605092094,
      "learning_rate": 6.617356886820125e-05,
      "loss": 0.4489,
      "step": 159
    },
    {
      "epoch": 1.7320703653585927,
      "grad_norm": 0.6416736344584283,
      "learning_rate": 6.594328357851024e-05,
      "loss": 0.4421,
      "step": 160
    },
    {
      "epoch": 1.742895805142084,
      "grad_norm": 0.5280156168745573,
      "learning_rate": 6.571150438746157e-05,
      "loss": 0.4388,
      "step": 161
    },
    {
      "epoch": 1.753721244925575,
      "grad_norm": 0.6725595610717778,
      "learning_rate": 6.547824464167913e-05,
      "loss": 0.4445,
      "step": 162
    },
    {
      "epoch": 1.7645466847090663,
      "grad_norm": 0.866322013820141,
      "learning_rate": 6.524351777304212e-05,
      "loss": 0.4377,
      "step": 163
    },
    {
      "epoch": 1.7753721244925575,
      "grad_norm": 1.0791676369113699,
      "learning_rate": 6.500733729791156e-05,
      "loss": 0.436,
      "step": 164
    },
    {
      "epoch": 1.7861975642760486,
      "grad_norm": 0.8392320511506657,
      "learning_rate": 6.476971681635209e-05,
      "loss": 0.4422,
      "step": 165
    },
    {
      "epoch": 1.79702300405954,
      "grad_norm": 0.9498667235195808,
      "learning_rate": 6.453067001134866e-05,
      "loss": 0.4363,
      "step": 166
    },
    {
      "epoch": 1.8078484438430311,
      "grad_norm": 0.9454272098217824,
      "learning_rate": 6.429021064801879e-05,
      "loss": 0.4413,
      "step": 167
    },
    {
      "epoch": 1.8186738836265224,
      "grad_norm": 1.0615616521628106,
      "learning_rate": 6.404835257281981e-05,
      "loss": 0.4381,
      "step": 168
    },
    {
      "epoch": 1.8294993234100136,
      "grad_norm": 0.9856953495735615,
      "learning_rate": 6.38051097127516e-05,
      "loss": 0.4464,
      "step": 169
    },
    {
      "epoch": 1.8403247631935047,
      "grad_norm": 0.753057646053996,
      "learning_rate": 6.356049607455456e-05,
      "loss": 0.4363,
      "step": 170
    },
    {
      "epoch": 1.851150202976996,
      "grad_norm": 0.39232654080746043,
      "learning_rate": 6.331452574390314e-05,
      "loss": 0.4312,
      "step": 171
    },
    {
      "epoch": 1.8619756427604872,
      "grad_norm": 0.4651286070389448,
      "learning_rate": 6.30672128845947e-05,
      "loss": 0.4326,
      "step": 172
    },
    {
      "epoch": 1.8728010825439783,
      "grad_norm": 0.498700372148241,
      "learning_rate": 6.281857173773384e-05,
      "loss": 0.4266,
      "step": 173
    },
    {
      "epoch": 1.8836265223274695,
      "grad_norm": 0.6143971888783442,
      "learning_rate": 6.25686166209125e-05,
      "loss": 0.4394,
      "step": 174
    },
    {
      "epoch": 1.8944519621109608,
      "grad_norm": 0.5823010909272562,
      "learning_rate": 6.231736192738535e-05,
      "loss": 0.4334,
      "step": 175
    },
    {
      "epoch": 1.9052774018944518,
      "grad_norm": 0.5433244104111088,
      "learning_rate": 6.206482212524109e-05,
      "loss": 0.4279,
      "step": 176
    },
    {
      "epoch": 1.9161028416779433,
      "grad_norm": 0.40429580977808727,
      "learning_rate": 6.181101175656932e-05,
      "loss": 0.4294,
      "step": 177
    },
    {
      "epoch": 1.9269282814614344,
      "grad_norm": 0.3710586154986815,
      "learning_rate": 6.155594543662304e-05,
      "loss": 0.4395,
      "step": 178
    },
    {
      "epoch": 1.9377537212449256,
      "grad_norm": 0.3830066254564287,
      "learning_rate": 6.12996378529772e-05,
      "loss": 0.4333,
      "step": 179
    },
    {
      "epoch": 1.9485791610284169,
      "grad_norm": 0.3863659341469274,
      "learning_rate": 6.104210376468284e-05,
      "loss": 0.4358,
      "step": 180
    },
    {
      "epoch": 1.959404600811908,
      "grad_norm": 0.35342627993699427,
      "learning_rate": 6.078335800141735e-05,
      "loss": 0.4341,
      "step": 181
    },
    {
      "epoch": 1.9702300405953992,
      "grad_norm": 0.36966775051107076,
      "learning_rate": 6.052341546263032e-05,
      "loss": 0.4306,
      "step": 182
    },
    {
      "epoch": 1.9810554803788905,
      "grad_norm": 0.4038581298736109,
      "learning_rate": 6.026229111668572e-05,
      "loss": 0.4378,
      "step": 183
    },
    {
      "epoch": 1.9918809201623815,
      "grad_norm": 0.4488772095787638,
      "learning_rate": 6.000000000000001e-05,
      "loss": 0.4389,
      "step": 184
    },
    {
      "epoch": 2.002706359945873,
      "grad_norm": 0.8295155001148825,
      "learning_rate": 5.973655721617613e-05,
      "loss": 0.6879,
      "step": 185
    },
    {
      "epoch": 2.013531799729364,
      "grad_norm": 1.4737016899362347,
      "learning_rate": 5.947197793513394e-05,
      "loss": 0.407,
      "step": 186
    },
    {
      "epoch": 2.024357239512855,
      "grad_norm": 0.5993476508246326,
      "learning_rate": 5.920627739223663e-05,
      "loss": 0.4022,
      "step": 187
    },
    {
      "epoch": 2.0351826792963466,
      "grad_norm": 0.7123790524835452,
      "learning_rate": 5.8939470887413364e-05,
      "loss": 0.4034,
      "step": 188
    },
    {
      "epoch": 2.0460081190798376,
      "grad_norm": 1.0489551042023235,
      "learning_rate": 5.867157378427835e-05,
      "loss": 0.398,
      "step": 189
    },
    {
      "epoch": 2.0568335588633286,
      "grad_norm": 1.1280435275179306,
      "learning_rate": 5.840260150924609e-05,
      "loss": 0.4039,
      "step": 190
    },
    {
      "epoch": 2.06765899864682,
      "grad_norm": 0.8086213856542805,
      "learning_rate": 5.81325695506431e-05,
      "loss": 0.4069,
      "step": 191
    },
    {
      "epoch": 2.078484438430311,
      "grad_norm": 0.7971526749207055,
      "learning_rate": 5.7861493457816005e-05,
      "loss": 0.4108,
      "step": 192
    },
    {
      "epoch": 2.089309878213802,
      "grad_norm": 0.6150880177459436,
      "learning_rate": 5.7589388840236235e-05,
      "loss": 0.4023,
      "step": 193
    },
    {
      "epoch": 2.1001353179972937,
      "grad_norm": 0.6200048701618767,
      "learning_rate": 5.73162713666011e-05,
      "loss": 0.3947,
      "step": 194
    },
    {
      "epoch": 2.1109607577807847,
      "grad_norm": 0.9512484994979096,
      "learning_rate": 5.704215676393154e-05,
      "loss": 0.4043,
      "step": 195
    },
    {
      "epoch": 2.121786197564276,
      "grad_norm": 0.8033132486864113,
      "learning_rate": 5.676706081666657e-05,
      "loss": 0.4023,
      "step": 196
    },
    {
      "epoch": 2.1326116373477673,
      "grad_norm": 0.6857766246265746,
      "learning_rate": 5.6490999365754304e-05,
      "loss": 0.3989,
      "step": 197
    },
    {
      "epoch": 2.1434370771312583,
      "grad_norm": 0.7160087962374165,
      "learning_rate": 5.621398830773979e-05,
      "loss": 0.3974,
      "step": 198
    },
    {
      "epoch": 2.15426251691475,
      "grad_norm": 0.7255913343428894,
      "learning_rate": 5.593604359384967e-05,
      "loss": 0.4046,
      "step": 199
    },
    {
      "epoch": 2.165087956698241,
      "grad_norm": 1.0668814854383228,
      "learning_rate": 5.5657181229073584e-05,
      "loss": 0.4069,
      "step": 200
    },
    {
      "epoch": 2.175913396481732,
      "grad_norm": 0.6695749773505052,
      "learning_rate": 5.5377417271242605e-05,
      "loss": 0.4004,
      "step": 201
    },
    {
      "epoch": 2.1867388362652234,
      "grad_norm": 0.41018811941016275,
      "learning_rate": 5.5096767830104574e-05,
      "loss": 0.3973,
      "step": 202
    },
    {
      "epoch": 2.1975642760487144,
      "grad_norm": 0.7331180833734219,
      "learning_rate": 5.481524906639645e-05,
      "loss": 0.4049,
      "step": 203
    },
    {
      "epoch": 2.208389715832206,
      "grad_norm": 0.7311488343422455,
      "learning_rate": 5.453287719091366e-05,
      "loss": 0.4039,
      "step": 204
    },
    {
      "epoch": 2.219215155615697,
      "grad_norm": 0.46217024270353824,
      "learning_rate": 5.424966846357668e-05,
      "loss": 0.4001,
      "step": 205
    },
    {
      "epoch": 2.230040595399188,
      "grad_norm": 0.3580272770419428,
      "learning_rate": 5.3965639192494746e-05,
      "loss": 0.3995,
      "step": 206
    },
    {
      "epoch": 2.2408660351826795,
      "grad_norm": 0.5415105544180893,
      "learning_rate": 5.368080573302676e-05,
      "loss": 0.4047,
      "step": 207
    },
    {
      "epoch": 2.2516914749661705,
      "grad_norm": 0.5146452956341113,
      "learning_rate": 5.339518448683945e-05,
      "loss": 0.3959,
      "step": 208
    },
    {
      "epoch": 2.2625169147496615,
      "grad_norm": 0.40046443729910486,
      "learning_rate": 5.3108791900963e-05,
      "loss": 0.3997,
      "step": 209
    },
    {
      "epoch": 2.273342354533153,
      "grad_norm": 0.3552796868979913,
      "learning_rate": 5.282164446684389e-05,
      "loss": 0.3954,
      "step": 210
    },
    {
      "epoch": 2.284167794316644,
      "grad_norm": 0.495387489856506,
      "learning_rate": 5.2533758719395294e-05,
      "loss": 0.3972,
      "step": 211
    },
    {
      "epoch": 2.2949932341001356,
      "grad_norm": 0.37686968387016173,
      "learning_rate": 5.2245151236044974e-05,
      "loss": 0.3976,
      "step": 212
    },
    {
      "epoch": 2.3058186738836266,
      "grad_norm": 0.2823917242370533,
      "learning_rate": 5.195583863578064e-05,
      "loss": 0.4049,
      "step": 213
    },
    {
      "epoch": 2.3166441136671176,
      "grad_norm": 0.33559837716294566,
      "learning_rate": 5.1665837578193e-05,
      "loss": 0.3992,
      "step": 214
    },
    {
      "epoch": 2.3274695534506087,
      "grad_norm": 0.2837563379510337,
      "learning_rate": 5.137516476251648e-05,
      "loss": 0.3997,
      "step": 215
    },
    {
      "epoch": 2.3382949932341,
      "grad_norm": 0.28846039259956097,
      "learning_rate": 5.108383692666756e-05,
      "loss": 0.4054,
      "step": 216
    },
    {
      "epoch": 2.349120433017591,
      "grad_norm": 0.3171321353936286,
      "learning_rate": 5.0791870846280974e-05,
      "loss": 0.3992,
      "step": 217
    },
    {
      "epoch": 2.3599458728010827,
      "grad_norm": 0.2678755670767341,
      "learning_rate": 5.0499283333743754e-05,
      "loss": 0.4077,
      "step": 218
    },
    {
      "epoch": 2.3707713125845737,
      "grad_norm": 0.2737614063807259,
      "learning_rate": 5.020609123722705e-05,
      "loss": 0.4037,
      "step": 219
    },
    {
      "epoch": 2.381596752368065,
      "grad_norm": 0.3048714730712856,
      "learning_rate": 4.9912311439716e-05,
      "loss": 0.3911,
      "step": 220
    },
    {
      "epoch": 2.3924221921515563,
      "grad_norm": 0.2574922530022283,
      "learning_rate": 4.961796085803754e-05,
      "loss": 0.3987,
      "step": 221
    },
    {
      "epoch": 2.4032476319350473,
      "grad_norm": 0.32375321489222875,
      "learning_rate": 4.932305644188627e-05,
      "loss": 0.4056,
      "step": 222
    },
    {
      "epoch": 2.414073071718539,
      "grad_norm": 0.3940310241959818,
      "learning_rate": 4.902761517284843e-05,
      "loss": 0.4041,
      "step": 223
    },
    {
      "epoch": 2.42489851150203,
      "grad_norm": 0.30709183482329716,
      "learning_rate": 4.8731654063424066e-05,
      "loss": 0.4012,
      "step": 224
    },
    {
      "epoch": 2.435723951285521,
      "grad_norm": 0.24828710130084936,
      "learning_rate": 4.843519015604739e-05,
      "loss": 0.3994,
      "step": 225
    },
    {
      "epoch": 2.4465493910690124,
      "grad_norm": 0.3039622222887511,
      "learning_rate": 4.8138240522105365e-05,
      "loss": 0.4019,
      "step": 226
    },
    {
      "epoch": 2.4573748308525034,
      "grad_norm": 0.3003924416251595,
      "learning_rate": 4.784082226095474e-05,
      "loss": 0.3962,
      "step": 227
    },
    {
      "epoch": 2.4682002706359945,
      "grad_norm": 0.22755737327455813,
      "learning_rate": 4.7542952498937437e-05,
      "loss": 0.3983,
      "step": 228
    },
    {
      "epoch": 2.479025710419486,
      "grad_norm": 0.25304360527606307,
      "learning_rate": 4.7244648388394235e-05,
      "loss": 0.3942,
      "step": 229
    },
    {
      "epoch": 2.489851150202977,
      "grad_norm": 0.24578325368030968,
      "learning_rate": 4.694592710667723e-05,
      "loss": 0.3963,
      "step": 230
    },
    {
      "epoch": 2.500676589986468,
      "grad_norm": 0.24077179007380192,
      "learning_rate": 4.664680585516057e-05,
      "loss": 0.4052,
      "step": 231
    },
    {
      "epoch": 2.5115020297699595,
      "grad_norm": 0.25770064300310863,
      "learning_rate": 4.634730185825006e-05,
      "loss": 0.4023,
      "step": 232
    },
    {
      "epoch": 2.5223274695534506,
      "grad_norm": 0.29543906900824984,
      "learning_rate": 4.604743236239124e-05,
      "loss": 0.3974,
      "step": 233
    },
    {
      "epoch": 2.533152909336942,
      "grad_norm": 0.23991016795181042,
      "learning_rate": 4.574721463507633e-05,
      "loss": 0.3999,
      "step": 234
    },
    {
      "epoch": 2.543978349120433,
      "grad_norm": 0.21566542112560377,
      "learning_rate": 4.5446665963849874e-05,
      "loss": 0.3931,
      "step": 235
    },
    {
      "epoch": 2.554803788903924,
      "grad_norm": 0.26245544883616334,
      "learning_rate": 4.514580365531325e-05,
      "loss": 0.3921,
      "step": 236
    },
    {
      "epoch": 2.565629228687415,
      "grad_norm": 0.2509466224574822,
      "learning_rate": 4.484464503412817e-05,
      "loss": 0.3989,
      "step": 237
    },
    {
      "epoch": 2.5764546684709067,
      "grad_norm": 0.2921963850485716,
      "learning_rate": 4.4543207442019e-05,
      "loss": 0.3983,
      "step": 238
    },
    {
      "epoch": 2.5872801082543977,
      "grad_norm": 0.21232868508333116,
      "learning_rate": 4.424150823677419e-05,
      "loss": 0.3951,
      "step": 239
    },
    {
      "epoch": 2.598105548037889,
      "grad_norm": 0.2846484943408765,
      "learning_rate": 4.3939564791246744e-05,
      "loss": 0.4034,
      "step": 240
    },
    {
      "epoch": 2.60893098782138,
      "grad_norm": 0.29357221150822416,
      "learning_rate": 4.3637394492353886e-05,
      "loss": 0.4052,
      "step": 241
    },
    {
      "epoch": 2.6197564276048713,
      "grad_norm": 0.2532347044145875,
      "learning_rate": 4.333501474007574e-05,
      "loss": 0.4045,
      "step": 242
    },
    {
      "epoch": 2.6305818673883627,
      "grad_norm": 0.20455259571037776,
      "learning_rate": 4.303244294645354e-05,
      "loss": 0.3958,
      "step": 243
    },
    {
      "epoch": 2.641407307171854,
      "grad_norm": 0.23989966946640798,
      "learning_rate": 4.272969653458685e-05,
      "loss": 0.394,
      "step": 244
    },
    {
      "epoch": 2.6522327469553453,
      "grad_norm": 0.2255061069896111,
      "learning_rate": 4.242679293763035e-05,
      "loss": 0.3942,
      "step": 245
    },
    {
      "epoch": 2.6630581867388363,
      "grad_norm": 0.23515116200339933,
      "learning_rate": 4.2123749597789984e-05,
      "loss": 0.3956,
      "step": 246
    },
    {
      "epoch": 2.6738836265223274,
      "grad_norm": 0.2248941135472504,
      "learning_rate": 4.1820583965318555e-05,
      "loss": 0.4011,
      "step": 247
    },
    {
      "epoch": 2.6847090663058184,
      "grad_norm": 0.24317542119733318,
      "learning_rate": 4.151731349751086e-05,
      "loss": 0.3964,
      "step": 248
    },
    {
      "epoch": 2.69553450608931,
      "grad_norm": 0.2182498690125644,
      "learning_rate": 4.12139556576985e-05,
      "loss": 0.3924,
      "step": 249
    },
    {
      "epoch": 2.706359945872801,
      "grad_norm": 0.19757577350714361,
      "learning_rate": 4.091052791424423e-05,
      "loss": 0.4033,
      "step": 250
    },
    {
      "epoch": 2.7171853856562924,
      "grad_norm": 0.24674384659213253,
      "learning_rate": 4.060704773953611e-05,
      "loss": 0.3998,
      "step": 251
    },
    {
      "epoch": 2.7280108254397835,
      "grad_norm": 0.25492479480718416,
      "learning_rate": 4.0303532608981336e-05,
      "loss": 0.4035,
      "step": 252
    },
    {
      "epoch": 2.7388362652232745,
      "grad_norm": 0.2061832272126445,
      "learning_rate": 4e-05,
      "loss": 0.4019,
      "step": 253
    },
    {
      "epoch": 2.749661705006766,
      "grad_norm": 0.20527089406948276,
      "learning_rate": 3.969646739101868e-05,
      "loss": 0.3954,
      "step": 254
    },
    {
      "epoch": 2.760487144790257,
      "grad_norm": 0.17828129068585608,
      "learning_rate": 3.93929522604639e-05,
      "loss": 0.3921,
      "step": 255
    },
    {
      "epoch": 2.7713125845737485,
      "grad_norm": 0.22611834558135363,
      "learning_rate": 3.908947208575577e-05,
      "loss": 0.3903,
      "step": 256
    },
    {
      "epoch": 2.7821380243572396,
      "grad_norm": 0.2132214449799281,
      "learning_rate": 3.878604434230151e-05,
      "loss": 0.4042,
      "step": 257
    },
    {
      "epoch": 2.7929634641407306,
      "grad_norm": 0.18422878450192567,
      "learning_rate": 3.848268650248915e-05,
      "loss": 0.3973,
      "step": 258
    },
    {
      "epoch": 2.803788903924222,
      "grad_norm": 0.199127350255564,
      "learning_rate": 3.8179416034681465e-05,
      "loss": 0.3954,
      "step": 259
    },
    {
      "epoch": 2.814614343707713,
      "grad_norm": 0.196931992775811,
      "learning_rate": 3.7876250402210036e-05,
      "loss": 0.4007,
      "step": 260
    },
    {
      "epoch": 2.825439783491204,
      "grad_norm": 0.22253615210660102,
      "learning_rate": 3.757320706236966e-05,
      "loss": 0.3902,
      "step": 261
    },
    {
      "epoch": 2.8362652232746957,
      "grad_norm": 0.21034998071613925,
      "learning_rate": 3.727030346541317e-05,
      "loss": 0.3951,
      "step": 262
    },
    {
      "epoch": 2.8470906630581867,
      "grad_norm": 0.220038673465254,
      "learning_rate": 3.696755705354647e-05,
      "loss": 0.3995,
      "step": 263
    },
    {
      "epoch": 2.8579161028416777,
      "grad_norm": 0.2376263555554069,
      "learning_rate": 3.6664985259924265e-05,
      "loss": 0.3955,
      "step": 264
    },
    {
      "epoch": 2.8687415426251692,
      "grad_norm": 0.18498115246118843,
      "learning_rate": 3.636260550764613e-05,
      "loss": 0.3994,
      "step": 265
    },
    {
      "epoch": 2.8795669824086603,
      "grad_norm": 0.2281213349380945,
      "learning_rate": 3.6060435208753256e-05,
      "loss": 0.3976,
      "step": 266
    },
    {
      "epoch": 2.8903924221921518,
      "grad_norm": 0.21257471713489817,
      "learning_rate": 3.575849176322582e-05,
      "loss": 0.3935,
      "step": 267
    },
    {
      "epoch": 2.901217861975643,
      "grad_norm": 0.17668776804131658,
      "learning_rate": 3.5456792557981e-05,
      "loss": 0.3958,
      "step": 268
    },
    {
      "epoch": 2.912043301759134,
      "grad_norm": 0.23579452170282567,
      "learning_rate": 3.5155354965871846e-05,
      "loss": 0.397,
      "step": 269
    },
    {
      "epoch": 2.9228687415426253,
      "grad_norm": 0.21220661458040213,
      "learning_rate": 3.485419634468676e-05,
      "loss": 0.398,
      "step": 270
    },
    {
      "epoch": 2.9336941813261164,
      "grad_norm": 0.29156175148563857,
      "learning_rate": 3.4553334036150146e-05,
      "loss": 0.3954,
      "step": 271
    },
    {
      "epoch": 2.944519621109608,
      "grad_norm": 0.21342855706614183,
      "learning_rate": 3.425278536492368e-05,
      "loss": 0.4014,
      "step": 272
    },
    {
      "epoch": 2.955345060893099,
      "grad_norm": 0.19007967113000737,
      "learning_rate": 3.395256763760877e-05,
      "loss": 0.3987,
      "step": 273
    },
    {
      "epoch": 2.96617050067659,
      "grad_norm": 0.20030092145985207,
      "learning_rate": 3.3652698141749946e-05,
      "loss": 0.3969,
      "step": 274
    },
    {
      "epoch": 2.976995940460081,
      "grad_norm": 0.1671102845524452,
      "learning_rate": 3.3353194144839444e-05,
      "loss": 0.4047,
      "step": 275
    },
    {
      "epoch": 2.9878213802435725,
      "grad_norm": 0.22912012233424828,
      "learning_rate": 3.305407289332279e-05,
      "loss": 0.3926,
      "step": 276
    },
    {
      "epoch": 2.9986468200270635,
      "grad_norm": 0.2683137652561869,
      "learning_rate": 3.2755351611605765e-05,
      "loss": 0.566,
      "step": 277
    },
    {
      "epoch": 3.009472259810555,
      "grad_norm": 0.2952413857194597,
      "learning_rate": 3.245704750106257e-05,
      "loss": 0.455,
      "step": 278
    },
    {
      "epoch": 3.020297699594046,
      "grad_norm": 0.23270668650047643,
      "learning_rate": 3.215917773904528e-05,
      "loss": 0.3727,
      "step": 279
    },
    {
      "epoch": 3.031123139377537,
      "grad_norm": 0.24171506781246388,
      "learning_rate": 3.1861759477894656e-05,
      "loss": 0.3665,
      "step": 280
    },
    {
      "epoch": 3.0419485791610286,
      "grad_norm": 0.2680860291808305,
      "learning_rate": 3.156480984395263e-05,
      "loss": 0.3666,
      "step": 281
    },
    {
      "epoch": 3.0527740189445196,
      "grad_norm": 0.19524723806636465,
      "learning_rate": 3.126834593657595e-05,
      "loss": 0.3574,
      "step": 282
    },
    {
      "epoch": 3.0635994587280106,
      "grad_norm": 0.23365726223954875,
      "learning_rate": 3.097238482715158e-05,
      "loss": 0.3691,
      "step": 283
    },
    {
      "epoch": 3.074424898511502,
      "grad_norm": 0.20482362620161818,
      "learning_rate": 3.0676943558113745e-05,
      "loss": 0.3689,
      "step": 284
    },
    {
      "epoch": 3.085250338294993,
      "grad_norm": 0.21174717829037049,
      "learning_rate": 3.0382039141962473e-05,
      "loss": 0.3631,
      "step": 285
    },
    {
      "epoch": 3.096075778078484,
      "grad_norm": 0.19387440535216213,
      "learning_rate": 3.008768856028401e-05,
      "loss": 0.3666,
      "step": 286
    },
    {
      "epoch": 3.1069012178619757,
      "grad_norm": 0.21151253274662332,
      "learning_rate": 2.9793908762772954e-05,
      "loss": 0.3667,
      "step": 287
    },
    {
      "epoch": 3.1177266576454667,
      "grad_norm": 0.19220716557476908,
      "learning_rate": 2.950071666625625e-05,
      "loss": 0.3676,
      "step": 288
    },
    {
      "epoch": 3.1285520974289582,
      "grad_norm": 0.18887116061755296,
      "learning_rate": 2.9208129153719026e-05,
      "loss": 0.3712,
      "step": 289
    },
    {
      "epoch": 3.1393775372124493,
      "grad_norm": 0.217906436813338,
      "learning_rate": 2.8916163073332462e-05,
      "loss": 0.368,
      "step": 290
    },
    {
      "epoch": 3.1502029769959403,
      "grad_norm": 0.16785424264087018,
      "learning_rate": 2.8624835237483535e-05,
      "loss": 0.3545,
      "step": 291
    },
    {
      "epoch": 3.161028416779432,
      "grad_norm": 0.21288405997782842,
      "learning_rate": 2.8334162421807002e-05,
      "loss": 0.3603,
      "step": 292
    },
    {
      "epoch": 3.171853856562923,
      "grad_norm": 0.17657164097795794,
      "learning_rate": 2.8044161364219378e-05,
      "loss": 0.3636,
      "step": 293
    },
    {
      "epoch": 3.182679296346414,
      "grad_norm": 0.17503579725776347,
      "learning_rate": 2.775484876395504e-05,
      "loss": 0.3669,
      "step": 294
    },
    {
      "epoch": 3.1935047361299054,
      "grad_norm": 0.1661974981289015,
      "learning_rate": 2.7466241280604716e-05,
      "loss": 0.3617,
      "step": 295
    },
    {
      "epoch": 3.2043301759133964,
      "grad_norm": 0.1804231214261922,
      "learning_rate": 2.717835553315612e-05,
      "loss": 0.3665,
      "step": 296
    },
    {
      "epoch": 3.2151556156968875,
      "grad_norm": 0.1627404959706761,
      "learning_rate": 2.6891208099037006e-05,
      "loss": 0.3719,
      "step": 297
    },
    {
      "epoch": 3.225981055480379,
      "grad_norm": 0.16858290476423804,
      "learning_rate": 2.6604815513160556e-05,
      "loss": 0.373,
      "step": 298
    },
    {
      "epoch": 3.23680649526387,
      "grad_norm": 0.17315102971170435,
      "learning_rate": 2.6319194266973256e-05,
      "loss": 0.3628,
      "step": 299
    },
    {
      "epoch": 3.2476319350473615,
      "grad_norm": 0.17092619054701708,
      "learning_rate": 2.6034360807505267e-05,
      "loss": 0.3583,
      "step": 300
    },
    {
      "epoch": 3.2584573748308525,
      "grad_norm": 0.16860391372808461,
      "learning_rate": 2.575033153642334e-05,
      "loss": 0.3627,
      "step": 301
    },
    {
      "epoch": 3.2692828146143436,
      "grad_norm": 0.18423357259943915,
      "learning_rate": 2.5467122809086355e-05,
      "loss": 0.37,
      "step": 302
    },
    {
      "epoch": 3.280108254397835,
      "grad_norm": 0.1549373166740434,
      "learning_rate": 2.518475093360356e-05,
      "loss": 0.3676,
      "step": 303
    },
    {
      "epoch": 3.290933694181326,
      "grad_norm": 0.20520475892662024,
      "learning_rate": 2.490323216989543e-05,
      "loss": 0.3671,
      "step": 304
    },
    {
      "epoch": 3.301759133964817,
      "grad_norm": 0.1462251183862993,
      "learning_rate": 2.46225827287574e-05,
      "loss": 0.3665,
      "step": 305
    },
    {
      "epoch": 3.3125845737483086,
      "grad_norm": 0.2037257968520142,
      "learning_rate": 2.4342818770926437e-05,
      "loss": 0.3669,
      "step": 306
    },
    {
      "epoch": 3.3234100135317997,
      "grad_norm": 0.15979516331505322,
      "learning_rate": 2.4063956406150345e-05,
      "loss": 0.3717,
      "step": 307
    },
    {
      "epoch": 3.3342354533152907,
      "grad_norm": 0.18421360813445462,
      "learning_rate": 2.3786011692260216e-05,
      "loss": 0.3659,
      "step": 308
    },
    {
      "epoch": 3.345060893098782,
      "grad_norm": 0.18353862952549732,
      "learning_rate": 2.35090006342457e-05,
      "loss": 0.3638,
      "step": 309
    },
    {
      "epoch": 3.3558863328822732,
      "grad_norm": 0.16761186075113543,
      "learning_rate": 2.323293918333345e-05,
      "loss": 0.3704,
      "step": 310
    },
    {
      "epoch": 3.3667117726657647,
      "grad_norm": 0.2012259488253085,
      "learning_rate": 2.295784323606848e-05,
      "loss": 0.3594,
      "step": 311
    },
    {
      "epoch": 3.3775372124492558,
      "grad_norm": 0.14417434763773077,
      "learning_rate": 2.2683728633398926e-05,
      "loss": 0.3604,
      "step": 312
    },
    {
      "epoch": 3.388362652232747,
      "grad_norm": 0.19765429801608228,
      "learning_rate": 2.2410611159763765e-05,
      "loss": 0.363,
      "step": 313
    },
    {
      "epoch": 3.3991880920162383,
      "grad_norm": 0.14242034553646193,
      "learning_rate": 2.2138506542184e-05,
      "loss": 0.3608,
      "step": 314
    },
    {
      "epoch": 3.4100135317997293,
      "grad_norm": 0.19492907121639994,
      "learning_rate": 2.1867430449356917e-05,
      "loss": 0.3672,
      "step": 315
    },
    {
      "epoch": 3.420838971583221,
      "grad_norm": 0.18616307600245358,
      "learning_rate": 2.1597398490753917e-05,
      "loss": 0.3559,
      "step": 316
    },
    {
      "epoch": 3.431664411366712,
      "grad_norm": 0.17108649074089605,
      "learning_rate": 2.1328426215721655e-05,
      "loss": 0.3619,
      "step": 317
    },
    {
      "epoch": 3.442489851150203,
      "grad_norm": 0.1745701339642718,
      "learning_rate": 2.1060529112586642e-05,
      "loss": 0.3694,
      "step": 318
    },
    {
      "epoch": 3.453315290933694,
      "grad_norm": 0.17738755753222712,
      "learning_rate": 2.0793722607763383e-05,
      "loss": 0.3655,
      "step": 319
    },
    {
      "epoch": 3.4641407307171854,
      "grad_norm": 0.17061926888972667,
      "learning_rate": 2.0528022064866048e-05,
      "loss": 0.3753,
      "step": 320
    },
    {
      "epoch": 3.4749661705006765,
      "grad_norm": 0.16060632755656637,
      "learning_rate": 2.0263442783823884e-05,
      "loss": 0.3582,
      "step": 321
    },
    {
      "epoch": 3.485791610284168,
      "grad_norm": 0.1603561152232048,
      "learning_rate": 2.0000000000000012e-05,
      "loss": 0.3667,
      "step": 322
    },
    {
      "epoch": 3.496617050067659,
      "grad_norm": 0.16661893716863155,
      "learning_rate": 1.9737708883314287e-05,
      "loss": 0.3631,
      "step": 323
    },
    {
      "epoch": 3.50744248985115,
      "grad_norm": 0.16611613821795707,
      "learning_rate": 1.9476584537369705e-05,
      "loss": 0.368,
      "step": 324
    },
    {
      "epoch": 3.5182679296346415,
      "grad_norm": 0.1544782537793828,
      "learning_rate": 1.9216641998582666e-05,
      "loss": 0.359,
      "step": 325
    },
    {
      "epoch": 3.5290933694181326,
      "grad_norm": 0.15513098503272876,
      "learning_rate": 1.8957896235317153e-05,
      "loss": 0.3653,
      "step": 326
    },
    {
      "epoch": 3.539918809201624,
      "grad_norm": 0.13881993920195754,
      "learning_rate": 1.8700362147022823e-05,
      "loss": 0.3643,
      "step": 327
    },
    {
      "epoch": 3.550744248985115,
      "grad_norm": 0.16797414295499802,
      "learning_rate": 1.844405456337698e-05,
      "loss": 0.3599,
      "step": 328
    },
    {
      "epoch": 3.561569688768606,
      "grad_norm": 0.1290170002098324,
      "learning_rate": 1.8188988243430698e-05,
      "loss": 0.3663,
      "step": 329
    },
    {
      "epoch": 3.572395128552097,
      "grad_norm": 0.1449933017909107,
      "learning_rate": 1.793517787475891e-05,
      "loss": 0.3652,
      "step": 330
    },
    {
      "epoch": 3.5832205683355887,
      "grad_norm": 0.12623373795565543,
      "learning_rate": 1.7682638072614667e-05,
      "loss": 0.3673,
      "step": 331
    },
    {
      "epoch": 3.5940460081190797,
      "grad_norm": 0.13727234583503345,
      "learning_rate": 1.7431383379087523e-05,
      "loss": 0.3684,
      "step": 332
    },
    {
      "epoch": 3.604871447902571,
      "grad_norm": 0.1216442976874091,
      "learning_rate": 1.7181428262266172e-05,
      "loss": 0.3663,
      "step": 333
    },
    {
      "epoch": 3.6156968876860622,
      "grad_norm": 0.11801832138683788,
      "learning_rate": 1.6932787115405318e-05,
      "loss": 0.3551,
      "step": 334
    },
    {
      "epoch": 3.6265223274695533,
      "grad_norm": 0.11494702663483314,
      "learning_rate": 1.6685474256096865e-05,
      "loss": 0.364,
      "step": 335
    },
    {
      "epoch": 3.6373477672530448,
      "grad_norm": 0.12323916060969814,
      "learning_rate": 1.6439503925445453e-05,
      "loss": 0.3607,
      "step": 336
    },
    {
      "epoch": 3.648173207036536,
      "grad_norm": 0.1098349302717658,
      "learning_rate": 1.619489028724842e-05,
      "loss": 0.3571,
      "step": 337
    },
    {
      "epoch": 3.6589986468200273,
      "grad_norm": 0.1325820150308931,
      "learning_rate": 1.595164742718019e-05,
      "loss": 0.3649,
      "step": 338
    },
    {
      "epoch": 3.6698240866035183,
      "grad_norm": 0.12739598866297566,
      "learning_rate": 1.5709789351981214e-05,
      "loss": 0.365,
      "step": 339
    },
    {
      "epoch": 3.6806495263870094,
      "grad_norm": 0.1279574306960612,
      "learning_rate": 1.5469329988651344e-05,
      "loss": 0.3616,
      "step": 340
    },
    {
      "epoch": 3.6914749661705004,
      "grad_norm": 0.12740756778928286,
      "learning_rate": 1.5230283183647938e-05,
      "loss": 0.3675,
      "step": 341
    },
    {
      "epoch": 3.702300405953992,
      "grad_norm": 0.12199919817169146,
      "learning_rate": 1.4992662702088451e-05,
      "loss": 0.357,
      "step": 342
    },
    {
      "epoch": 3.713125845737483,
      "grad_norm": 0.13290437651272327,
      "learning_rate": 1.47564822269579e-05,
      "loss": 0.3643,
      "step": 343
    },
    {
      "epoch": 3.7239512855209744,
      "grad_norm": 0.12309727230864277,
      "learning_rate": 1.4521755358320868e-05,
      "loss": 0.3654,
      "step": 344
    },
    {
      "epoch": 3.7347767253044655,
      "grad_norm": 0.11580584203358321,
      "learning_rate": 1.4288495612538427e-05,
      "loss": 0.3603,
      "step": 345
    },
    {
      "epoch": 3.7456021650879565,
      "grad_norm": 0.12053339473193542,
      "learning_rate": 1.4056716421489762e-05,
      "loss": 0.3613,
      "step": 346
    },
    {
      "epoch": 3.756427604871448,
      "grad_norm": 0.13687114857273397,
      "learning_rate": 1.3826431131798757e-05,
      "loss": 0.3657,
      "step": 347
    },
    {
      "epoch": 3.767253044654939,
      "grad_norm": 0.10679108792867367,
      "learning_rate": 1.3597653004065436e-05,
      "loss": 0.3678,
      "step": 348
    },
    {
      "epoch": 3.7780784844384305,
      "grad_norm": 0.13191858463972334,
      "learning_rate": 1.337039521210238e-05,
      "loss": 0.3624,
      "step": 349
    },
    {
      "epoch": 3.7889039242219216,
      "grad_norm": 0.11520187391454531,
      "learning_rate": 1.3144670842176134e-05,
      "loss": 0.3571,
      "step": 350
    },
    {
      "epoch": 3.7997293640054126,
      "grad_norm": 0.1072027738008208,
      "learning_rate": 1.2920492892253642e-05,
      "loss": 0.3583,
      "step": 351
    },
    {
      "epoch": 3.8105548037889037,
      "grad_norm": 0.1348518526777805,
      "learning_rate": 1.2697874271253844e-05,
      "loss": 0.3674,
      "step": 352
    },
    {
      "epoch": 3.821380243572395,
      "grad_norm": 0.11201420019346077,
      "learning_rate": 1.247682779830421e-05,
      "loss": 0.373,
      "step": 353
    },
    {
      "epoch": 3.832205683355886,
      "grad_norm": 0.11768440300752973,
      "learning_rate": 1.2257366202002693e-05,
      "loss": 0.3655,
      "step": 354
    },
    {
      "epoch": 3.8430311231393777,
      "grad_norm": 0.11313665709627198,
      "learning_rate": 1.2039502119684725e-05,
      "loss": 0.3671,
      "step": 355
    },
    {
      "epoch": 3.8538565629228687,
      "grad_norm": 0.10418998401518839,
      "learning_rate": 1.1823248096695515e-05,
      "loss": 0.3618,
      "step": 356
    },
    {
      "epoch": 3.8646820027063598,
      "grad_norm": 0.110112732988295,
      "learning_rate": 1.160861658566764e-05,
      "loss": 0.3611,
      "step": 357
    },
    {
      "epoch": 3.8755074424898512,
      "grad_norm": 0.09942258198464687,
      "learning_rate": 1.1395619945804021e-05,
      "loss": 0.3586,
      "step": 358
    },
    {
      "epoch": 3.8863328822733423,
      "grad_norm": 0.10319257251599674,
      "learning_rate": 1.118427044216619e-05,
      "loss": 0.356,
      "step": 359
    },
    {
      "epoch": 3.8971583220568338,
      "grad_norm": 0.10283087239718877,
      "learning_rate": 1.0974580244968039e-05,
      "loss": 0.3585,
      "step": 360
    },
    {
      "epoch": 3.907983761840325,
      "grad_norm": 0.10712486173724692,
      "learning_rate": 1.0766561428875036e-05,
      "loss": 0.372,
      "step": 361
    },
    {
      "epoch": 3.918809201623816,
      "grad_norm": 0.10477576856268758,
      "learning_rate": 1.0560225972308925e-05,
      "loss": 0.361,
      "step": 362
    },
    {
      "epoch": 3.9296346414073073,
      "grad_norm": 0.09697172045393422,
      "learning_rate": 1.0355585756757907e-05,
      "loss": 0.3604,
      "step": 363
    },
    {
      "epoch": 3.9404600811907984,
      "grad_norm": 0.10981365350889641,
      "learning_rate": 1.0152652566092582e-05,
      "loss": 0.3609,
      "step": 364
    },
    {
      "epoch": 3.9512855209742894,
      "grad_norm": 0.10519785185770371,
      "learning_rate": 9.951438085887294e-06,
      "loss": 0.3631,
      "step": 365
    },
    {
      "epoch": 3.962110960757781,
      "grad_norm": 0.10740052554928715,
      "learning_rate": 9.751953902747267e-06,
      "loss": 0.368,
      "step": 366
    },
    {
      "epoch": 3.972936400541272,
      "grad_norm": 0.10480987894849778,
      "learning_rate": 9.554211503641416e-06,
      "loss": 0.3565,
      "step": 367
    },
    {
      "epoch": 3.983761840324763,
      "grad_norm": 0.1040412988156162,
      "learning_rate": 9.358222275240884e-06,
      "loss": 0.3592,
      "step": 368
    },
    {
      "epoch": 3.9945872801082545,
      "grad_norm": 0.10495487860148761,
      "learning_rate": 9.163997503263364e-06,
      "loss": 0.3707,
      "step": 369
    },
    {
      "epoch": 4.005412719891746,
      "grad_norm": 0.1875294677786279,
      "learning_rate": 8.971548371823205e-06,
      "loss": 0.5807,
      "step": 370
    },
    {
      "epoch": 4.016238159675237,
      "grad_norm": 0.14041995604948812,
      "learning_rate": 8.780885962787452e-06,
      "loss": 0.3404,
      "step": 371
    },
    {
      "epoch": 4.027063599458728,
      "grad_norm": 0.1253661015397288,
      "learning_rate": 8.592021255137646e-06,
      "loss": 0.3487,
      "step": 372
    },
    {
      "epoch": 4.037889039242219,
      "grad_norm": 0.11634016776170551,
      "learning_rate": 8.404965124337665e-06,
      "loss": 0.338,
      "step": 373
    },
    {
      "epoch": 4.04871447902571,
      "grad_norm": 0.13106693030223066,
      "learning_rate": 8.219728341707425e-06,
      "loss": 0.3418,
      "step": 374
    },
    {
      "epoch": 4.059539918809202,
      "grad_norm": 0.1320753462354552,
      "learning_rate": 8.036321573802706e-06,
      "loss": 0.345,
      "step": 375
    },
    {
      "epoch": 4.070365358592693,
      "grad_norm": 0.1408674098165186,
      "learning_rate": 7.854755381800872e-06,
      "loss": 0.344,
      "step": 376
    },
    {
      "epoch": 4.081190798376184,
      "grad_norm": 0.132540628100584,
      "learning_rate": 7.675040220892782e-06,
      "loss": 0.3441,
      "step": 377
    },
    {
      "epoch": 4.092016238159675,
      "grad_norm": 0.11968734635815752,
      "learning_rate": 7.4971864396806795e-06,
      "loss": 0.3389,
      "step": 378
    },
    {
      "epoch": 4.102841677943166,
      "grad_norm": 0.11698410253549779,
      "learning_rate": 7.321204279582321e-06,
      "loss": 0.3449,
      "step": 379
    },
    {
      "epoch": 4.113667117726657,
      "grad_norm": 0.12255988282586658,
      "learning_rate": 7.147103874241246e-06,
      "loss": 0.3499,
      "step": 380
    },
    {
      "epoch": 4.124492557510149,
      "grad_norm": 0.12100999680393194,
      "learning_rate": 6.974895248943223e-06,
      "loss": 0.3423,
      "step": 381
    },
    {
      "epoch": 4.13531799729364,
      "grad_norm": 0.11527621996191134,
      "learning_rate": 6.804588320038998e-06,
      "loss": 0.3392,
      "step": 382
    },
    {
      "epoch": 4.146143437077131,
      "grad_norm": 0.12139325805430597,
      "learning_rate": 6.636192894373232e-06,
      "loss": 0.3444,
      "step": 383
    },
    {
      "epoch": 4.156968876860622,
      "grad_norm": 0.11122602457716502,
      "learning_rate": 6.4697186687198285e-06,
      "loss": 0.3383,
      "step": 384
    },
    {
      "epoch": 4.167794316644113,
      "grad_norm": 0.11224637325482494,
      "learning_rate": 6.305175229223541e-06,
      "loss": 0.3319,
      "step": 385
    },
    {
      "epoch": 4.178619756427604,
      "grad_norm": 0.10879841592643402,
      "learning_rate": 6.142572050847966e-06,
      "loss": 0.3479,
      "step": 386
    },
    {
      "epoch": 4.189445196211096,
      "grad_norm": 0.12627793027748102,
      "learning_rate": 5.981918496829959e-06,
      "loss": 0.3464,
      "step": 387
    },
    {
      "epoch": 4.200270635994587,
      "grad_norm": 0.12250577402975518,
      "learning_rate": 5.823223818140458e-06,
      "loss": 0.3444,
      "step": 388
    },
    {
      "epoch": 4.211096075778078,
      "grad_norm": 0.11127210268527363,
      "learning_rate": 5.666497152951791e-06,
      "loss": 0.3438,
      "step": 389
    },
    {
      "epoch": 4.2219215155615695,
      "grad_norm": 0.10375456129249835,
      "learning_rate": 5.511747526111459e-06,
      "loss": 0.343,
      "step": 390
    },
    {
      "epoch": 4.2327469553450605,
      "grad_norm": 0.11020190222669736,
      "learning_rate": 5.358983848622452e-06,
      "loss": 0.3443,
      "step": 391
    },
    {
      "epoch": 4.243572395128552,
      "grad_norm": 0.11547044420370078,
      "learning_rate": 5.208214917130132e-06,
      "loss": 0.3456,
      "step": 392
    },
    {
      "epoch": 4.2543978349120435,
      "grad_norm": 0.11195990742206446,
      "learning_rate": 5.059449413415705e-06,
      "loss": 0.3458,
      "step": 393
    },
    {
      "epoch": 4.2652232746955345,
      "grad_norm": 0.1044621266094549,
      "learning_rate": 4.91269590389623e-06,
      "loss": 0.3413,
      "step": 394
    },
    {
      "epoch": 4.276048714479026,
      "grad_norm": 0.09428855278720129,
      "learning_rate": 4.7679628391314615e-06,
      "loss": 0.3387,
      "step": 395
    },
    {
      "epoch": 4.286874154262517,
      "grad_norm": 0.1218060268682413,
      "learning_rate": 4.625258553337108e-06,
      "loss": 0.3373,
      "step": 396
    },
    {
      "epoch": 4.2976995940460085,
      "grad_norm": 0.10192114482593552,
      "learning_rate": 4.484591263904992e-06,
      "loss": 0.3429,
      "step": 397
    },
    {
      "epoch": 4.3085250338295,
      "grad_norm": 0.09490696955645948,
      "learning_rate": 4.345969070929848e-06,
      "loss": 0.3397,
      "step": 398
    },
    {
      "epoch": 4.319350473612991,
      "grad_norm": 0.0943001566489396,
      "learning_rate": 4.20939995674289e-06,
      "loss": 0.3498,
      "step": 399
    },
    {
      "epoch": 4.330175913396482,
      "grad_norm": 0.0910155289601598,
      "learning_rate": 4.07489178545216e-06,
      "loss": 0.3415,
      "step": 400
    },
    {
      "epoch": 4.341001353179973,
      "grad_norm": 0.10041740401971753,
      "learning_rate": 3.942452302489686e-06,
      "loss": 0.3405,
      "step": 401
    },
    {
      "epoch": 4.351826792963464,
      "grad_norm": 0.10532998834210694,
      "learning_rate": 3.8120891341654955e-06,
      "loss": 0.3431,
      "step": 402
    },
    {
      "epoch": 4.362652232746956,
      "grad_norm": 0.09532752858531783,
      "learning_rate": 3.6838097872284382e-06,
      "loss": 0.3422,
      "step": 403
    },
    {
      "epoch": 4.373477672530447,
      "grad_norm": 0.08806314013092435,
      "learning_rate": 3.5576216484339356e-06,
      "loss": 0.3438,
      "step": 404
    },
    {
      "epoch": 4.384303112313938,
      "grad_norm": 0.0902671665456689,
      "learning_rate": 3.4335319841186077e-06,
      "loss": 0.3454,
      "step": 405
    },
    {
      "epoch": 4.395128552097429,
      "grad_norm": 0.09375561638182935,
      "learning_rate": 3.311547939781887e-06,
      "loss": 0.338,
      "step": 406
    },
    {
      "epoch": 4.40595399188092,
      "grad_norm": 0.09234688834689847,
      "learning_rate": 3.1916765396745288e-06,
      "loss": 0.3394,
      "step": 407
    },
    {
      "epoch": 4.416779431664412,
      "grad_norm": 0.08840156590696185,
      "learning_rate": 3.0739246863941717e-06,
      "loss": 0.3472,
      "step": 408
    },
    {
      "epoch": 4.427604871447903,
      "grad_norm": 0.08739334077231752,
      "learning_rate": 2.95829916048779e-06,
      "loss": 0.3409,
      "step": 409
    },
    {
      "epoch": 4.438430311231394,
      "grad_norm": 0.0875172603982304,
      "learning_rate": 2.84480662006132e-06,
      "loss": 0.3477,
      "step": 410
    },
    {
      "epoch": 4.449255751014885,
      "grad_norm": 0.094984179377533,
      "learning_rate": 2.7334536003962166e-06,
      "loss": 0.3393,
      "step": 411
    },
    {
      "epoch": 4.460081190798376,
      "grad_norm": 0.08706812389766655,
      "learning_rate": 2.6242465135731587e-06,
      "loss": 0.3395,
      "step": 412
    },
    {
      "epoch": 4.470906630581867,
      "grad_norm": 0.09622976821594861,
      "learning_rate": 2.5171916481027968e-06,
      "loss": 0.3438,
      "step": 413
    },
    {
      "epoch": 4.481732070365359,
      "grad_norm": 0.09042626515531108,
      "learning_rate": 2.4122951685636674e-06,
      "loss": 0.3397,
      "step": 414
    },
    {
      "epoch": 4.49255751014885,
      "grad_norm": 0.08697445987862922,
      "learning_rate": 2.309563115247184e-06,
      "loss": 0.3442,
      "step": 415
    },
    {
      "epoch": 4.503382949932341,
      "grad_norm": 0.08521377871295831,
      "learning_rate": 2.2090014038098364e-06,
      "loss": 0.343,
      "step": 416
    },
    {
      "epoch": 4.514208389715832,
      "grad_norm": 0.08874899185726369,
      "learning_rate": 2.1106158249325536e-06,
      "loss": 0.336,
      "step": 417
    },
    {
      "epoch": 4.525033829499323,
      "grad_norm": 0.08808087170450281,
      "learning_rate": 2.014412043987233e-06,
      "loss": 0.3428,
      "step": 418
    },
    {
      "epoch": 4.535859269282815,
      "grad_norm": 0.08142283041322351,
      "learning_rate": 1.9203956007105293e-06,
      "loss": 0.3443,
      "step": 419
    },
    {
      "epoch": 4.546684709066306,
      "grad_norm": 0.0851360186887014,
      "learning_rate": 1.8285719088848442e-06,
      "loss": 0.3376,
      "step": 420
    },
    {
      "epoch": 4.557510148849797,
      "grad_norm": 0.08649957637526776,
      "learning_rate": 1.7389462560266102e-06,
      "loss": 0.3467,
      "step": 421
    },
    {
      "epoch": 4.568335588633288,
      "grad_norm": 0.08311650620599292,
      "learning_rate": 1.651523803081778e-06,
      "loss": 0.3451,
      "step": 422
    },
    {
      "epoch": 4.579161028416779,
      "grad_norm": 0.08116356626628798,
      "learning_rate": 1.5663095841286535e-06,
      "loss": 0.3492,
      "step": 423
    },
    {
      "epoch": 4.589986468200271,
      "grad_norm": 0.0824392457885468,
      "learning_rate": 1.4833085060880349e-06,
      "loss": 0.3387,
      "step": 424
    },
    {
      "epoch": 4.600811907983762,
      "grad_norm": 0.08566606903234027,
      "learning_rate": 1.4025253484405954e-06,
      "loss": 0.3459,
      "step": 425
    },
    {
      "epoch": 4.611637347767253,
      "grad_norm": 0.08258042133215138,
      "learning_rate": 1.3239647629517572e-06,
      "loss": 0.3405,
      "step": 426
    },
    {
      "epoch": 4.622462787550744,
      "grad_norm": 0.08262381506050841,
      "learning_rate": 1.2476312734037488e-06,
      "loss": 0.3433,
      "step": 427
    },
    {
      "epoch": 4.633288227334235,
      "grad_norm": 0.08609860936550076,
      "learning_rate": 1.1735292753351346e-06,
      "loss": 0.3375,
      "step": 428
    },
    {
      "epoch": 4.644113667117726,
      "grad_norm": 0.08075622135740361,
      "learning_rate": 1.1016630357877278e-06,
      "loss": 0.343,
      "step": 429
    },
    {
      "epoch": 4.654939106901217,
      "grad_norm": 0.08233697551125709,
      "learning_rate": 1.0320366930608405e-06,
      "loss": 0.3394,
      "step": 430
    },
    {
      "epoch": 4.665764546684709,
      "grad_norm": 0.08316101123672864,
      "learning_rate": 9.64654256473021e-07,
      "loss": 0.3437,
      "step": 431
    },
    {
      "epoch": 4.6765899864682,
      "grad_norm": 0.09300648661066627,
      "learning_rate": 8.995196061311628e-07,
      "loss": 0.3415,
      "step": 432
    },
    {
      "epoch": 4.687415426251691,
      "grad_norm": 0.0875268399310967,
      "learning_rate": 8.366364927070836e-07,
      "loss": 0.3545,
      "step": 433
    },
    {
      "epoch": 4.698240866035182,
      "grad_norm": 0.08398923274446536,
      "learning_rate": 7.760085372215509e-07,
      "loss": 0.3453,
      "step": 434
    },
    {
      "epoch": 4.7090663058186735,
      "grad_norm": 0.08167730199654379,
      "learning_rate": 7.176392308357649e-07,
      "loss": 0.3393,
      "step": 435
    },
    {
      "epoch": 4.719891745602165,
      "grad_norm": 0.07876472278529771,
      "learning_rate": 6.61531934650319e-07,
      "loss": 0.3422,
      "step": 436
    },
    {
      "epoch": 4.730717185385656,
      "grad_norm": 0.07955859230593923,
      "learning_rate": 6.076898795116792e-07,
      "loss": 0.3446,
      "step": 437
    },
    {
      "epoch": 4.7415426251691475,
      "grad_norm": 0.07950730370407931,
      "learning_rate": 5.561161658261238e-07,
      "loss": 0.3497,
      "step": 438
    },
    {
      "epoch": 4.7523680649526385,
      "grad_norm": 0.08188578916605453,
      "learning_rate": 5.068137633812153e-07,
      "loss": 0.3437,
      "step": 439
    },
    {
      "epoch": 4.76319350473613,
      "grad_norm": 0.07643930436958575,
      "learning_rate": 4.5978551117479063e-07,
      "loss": 0.3395,
      "step": 440
    },
    {
      "epoch": 4.7740189445196215,
      "grad_norm": 0.0792066807028224,
      "learning_rate": 4.1503411725147824e-07,
      "loss": 0.345,
      "step": 441
    },
    {
      "epoch": 4.7848443843031125,
      "grad_norm": 0.07917452696979885,
      "learning_rate": 3.725621585467698e-07,
      "loss": 0.3349,
      "step": 442
    },
    {
      "epoch": 4.795669824086604,
      "grad_norm": 0.0814817497437418,
      "learning_rate": 3.3237208073862324e-07,
      "loss": 0.3465,
      "step": 443
    },
    {
      "epoch": 4.806495263870095,
      "grad_norm": 0.07978334424890356,
      "learning_rate": 2.9446619810662434e-07,
      "loss": 0.3444,
      "step": 444
    },
    {
      "epoch": 4.817320703653586,
      "grad_norm": 0.08242514400126746,
      "learning_rate": 2.5884669339875546e-07,
      "loss": 0.3453,
      "step": 445
    },
    {
      "epoch": 4.828146143437078,
      "grad_norm": 0.07808288606410942,
      "learning_rate": 2.2551561770565611e-07,
      "loss": 0.3485,
      "step": 446
    },
    {
      "epoch": 4.838971583220569,
      "grad_norm": 0.07928112358547848,
      "learning_rate": 1.9447489034256196e-07,
      "loss": 0.348,
      "step": 447
    },
    {
      "epoch": 4.84979702300406,
      "grad_norm": 0.0794259887072925,
      "learning_rate": 1.6572629873874869e-07,
      "loss": 0.3414,
      "step": 448
    },
    {
      "epoch": 4.860622462787551,
      "grad_norm": 0.07566795403724712,
      "learning_rate": 1.392714983346366e-07,
      "loss": 0.3435,
      "step": 449
    },
    {
      "epoch": 4.871447902571042,
      "grad_norm": 0.07843464378271181,
      "learning_rate": 1.1511201248643134e-07,
      "loss": 0.3405,
      "step": 450
    },
    {
      "epoch": 4.882273342354533,
      "grad_norm": 0.07703599202280859,
      "learning_rate": 9.324923237842953e-08,
      "loss": 0.3426,
      "step": 451
    },
    {
      "epoch": 4.893098782138025,
      "grad_norm": 0.07979599503404064,
      "learning_rate": 7.368441694290963e-08,
      "loss": 0.3487,
      "step": 452
    },
    {
      "epoch": 4.903924221921516,
      "grad_norm": 0.0781513546026194,
      "learning_rate": 5.641869278761203e-08,
      "loss": 0.3463,
      "step": 453
    },
    {
      "epoch": 4.914749661705007,
      "grad_norm": 0.07970596360712374,
      "learning_rate": 4.145305413089773e-08,
      "loss": 0.3469,
      "step": 454
    },
    {
      "epoch": 4.925575101488498,
      "grad_norm": 0.07818127551791816,
      "learning_rate": 2.8788362744478492e-08,
      "loss": 0.336,
      "step": 455
    },
    {
      "epoch": 4.936400541271989,
      "grad_norm": 0.07869679667959938,
      "learning_rate": 1.8425347903794355e-08,
      "loss": 0.347,
      "step": 456
    },
    {
      "epoch": 4.94722598105548,
      "grad_norm": 0.08181150471394064,
      "learning_rate": 1.0364606346033867e-08,
      "loss": 0.3518,
      "step": 457
    },
    {
      "epoch": 4.958051420838972,
      "grad_norm": 0.07775322978676268,
      "learning_rate": 4.606602235752711e-09,
      "loss": 0.3442,
      "step": 458
    },
    {
      "epoch": 4.968876860622463,
      "grad_norm": 0.07656113821295778,
      "learning_rate": 1.1516671381528454e-09,
      "loss": 0.3409,
      "step": 459
    },
    {
      "epoch": 4.979702300405954,
      "grad_norm": 0.07887018893888488,
      "learning_rate": 0.0,
      "loss": 0.345,
      "step": 460
    },
    {
      "epoch": 4.979702300405954,
      "step": 460,
      "total_flos": 1.1837169167247606e+19,
      "train_loss": 0.4245723386173663,
      "train_runtime": 101661.9148,
      "train_samples_per_second": 2.326,
      "train_steps_per_second": 0.005
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 460,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1837169167247606e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}