{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 335,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 9.992014557607169,
      "learning_rate": 9.090909090909091e-07,
      "loss": 2.7651,
      "step": 1
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.855357081400161,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 2.7106,
      "step": 2
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.452032696456367,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 2.7538,
      "step": 3
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.376280328426684,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 2.5369,
      "step": 4
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.188703375753345,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 2.4723,
      "step": 5
    },
    {
      "epoch": 0.02,
      "grad_norm": 7.173561512899845,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 2.4736,
      "step": 6
    },
    {
      "epoch": 0.02,
      "grad_norm": 7.1450790437206235,
      "learning_rate": 6.363636363636364e-06,
      "loss": 2.4403,
      "step": 7
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.247742053090214,
      "learning_rate": 7.272727272727273e-06,
      "loss": 2.0348,
      "step": 8
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.481406635010285,
      "learning_rate": 8.181818181818183e-06,
      "loss": 1.9127,
      "step": 9
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.521133417095248,
      "learning_rate": 9.090909090909091e-06,
      "loss": 1.891,
      "step": 10
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.731593362969328,
      "learning_rate": 1e-05,
      "loss": 1.8907,
      "step": 11
    },
    {
      "epoch": 0.04,
      "grad_norm": 6.005591131903454,
      "learning_rate": 9.99976495753613e-06,
      "loss": 1.7979,
      "step": 12
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.0731579551147816,
      "learning_rate": 9.999059852242508e-06,
      "loss": 1.6642,
      "step": 13
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.7779072164471383,
      "learning_rate": 9.997884750411004e-06,
      "loss": 1.4797,
      "step": 14
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.3917050377427156,
      "learning_rate": 9.996239762521152e-06,
      "loss": 1.6384,
      "step": 15
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.348652641302203,
      "learning_rate": 9.994125043229753e-06,
      "loss": 1.4289,
      "step": 16
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.7580016074034144,
      "learning_rate": 9.991540791356342e-06,
      "loss": 1.3537,
      "step": 17
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.600060931738886,
      "learning_rate": 9.98848724986449e-06,
      "loss": 1.4414,
      "step": 18
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.402583908854572,
      "learning_rate": 9.98496470583896e-06,
      "loss": 1.4358,
      "step": 19
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.3681534411005365,
      "learning_rate": 9.980973490458728e-06,
      "loss": 1.3664,
      "step": 20
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.417118446828482,
      "learning_rate": 9.976513978965829e-06,
      "loss": 1.2636,
      "step": 21
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.406034881091615,
      "learning_rate": 9.971586590630094e-06,
      "loss": 1.2646,
      "step": 22
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.2282767138519564,
      "learning_rate": 9.966191788709716e-06,
      "loss": 1.2749,
      "step": 23
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.0495586711063853,
      "learning_rate": 9.960330080407712e-06,
      "loss": 1.2495,
      "step": 24
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.232570025850529,
      "learning_rate": 9.954002016824226e-06,
      "loss": 1.255,
      "step": 25
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.9746334176340221,
      "learning_rate": 9.947208192904722e-06,
      "loss": 1.2702,
      "step": 26
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.90630424922607,
      "learning_rate": 9.939949247384046e-06,
      "loss": 1.1942,
      "step": 27
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.0737836420918563,
      "learning_rate": 9.93222586272637e-06,
      "loss": 1.259,
      "step": 28
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.0980819753190514,
      "learning_rate": 9.924038765061042e-06,
      "loss": 1.2254,
      "step": 29
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.07955602263235,
      "learning_rate": 9.915388724114301e-06,
      "loss": 1.1271,
      "step": 30
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.1659071183953023,
      "learning_rate": 9.906276553136924e-06,
      "loss": 1.2434,
      "step": 31
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.1275763181049014,
      "learning_rate": 9.896703108827758e-06,
      "loss": 1.1295,
      "step": 32
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.124992777307241,
      "learning_rate": 9.886669291253178e-06,
      "loss": 1.0909,
      "step": 33
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.1457535781359405,
      "learning_rate": 9.876176043762467e-06,
      "loss": 1.0579,
      "step": 34
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.299797197398565,
      "learning_rate": 9.86522435289912e-06,
      "loss": 1.1403,
      "step": 35
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.9841854125040834,
      "learning_rate": 9.853815248308101e-06,
      "loss": 1.1669,
      "step": 36
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.9449153939112522,
      "learning_rate": 9.841949802639031e-06,
      "loss": 1.177,
      "step": 37
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.823305444585483,
      "learning_rate": 9.829629131445342e-06,
      "loss": 1.0274,
      "step": 38
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.9967356357088837,
      "learning_rate": 9.816854393079402e-06,
      "loss": 1.052,
      "step": 39
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.0742167672185525,
      "learning_rate": 9.803626788583603e-06,
      "loss": 1.0244,
      "step": 40
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.030292578786332,
      "learning_rate": 9.789947561577445e-06,
      "loss": 1.0812,
      "step": 41
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.1285455940700455,
      "learning_rate": 9.775817998140615e-06,
      "loss": 1.0372,
      "step": 42
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.104132559156802,
      "learning_rate": 9.761239426692077e-06,
      "loss": 1.0363,
      "step": 43
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.1801107672452096,
      "learning_rate": 9.74621321786517e-06,
      "loss": 1.1396,
      "step": 44
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.930455012718363,
      "learning_rate": 9.730740784378755e-06,
      "loss": 1.0207,
      "step": 45
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.9114532903731203,
      "learning_rate": 9.71482358090438e-06,
      "loss": 1.0088,
      "step": 46
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.9603910862722937,
      "learning_rate": 9.698463103929542e-06,
      "loss": 1.0367,
      "step": 47
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.0049474141842483,
      "learning_rate": 9.681660891616967e-06,
      "loss": 1.04,
      "step": 48
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.9843717409843054,
      "learning_rate": 9.664418523660004e-06,
      "loss": 1.0541,
      "step": 49
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.8735616571233806,
      "learning_rate": 9.646737621134112e-06,
      "loss": 1.0682,
      "step": 50
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.0255894944970154,
      "learning_rate": 9.628619846344453e-06,
      "loss": 1.072,
      "step": 51
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.8854561919254644,
      "learning_rate": 9.610066902669593e-06,
      "loss": 1.0295,
      "step": 52
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.834056540014944,
      "learning_rate": 9.591080534401371e-06,
      "loss": 1.0185,
      "step": 53
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.9190598160066954,
      "learning_rate": 9.571662526580898e-06,
      "loss": 1.0398,
      "step": 54
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.927342754807542,
      "learning_rate": 9.551814704830734e-06,
      "loss": 1.0799,
      "step": 55
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.9976883134890686,
      "learning_rate": 9.531538935183252e-06,
      "loss": 0.9839,
      "step": 56
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.010561519470896,
      "learning_rate": 9.51083712390519e-06,
      "loss": 1.0452,
      "step": 57
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.8025815822535263,
      "learning_rate": 9.48971121731844e-06,
      "loss": 0.9052,
      "step": 58
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.1672411135729948,
      "learning_rate": 9.468163201617063e-06,
      "loss": 0.9764,
      "step": 59
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.1274344440002197,
      "learning_rate": 9.446195102680531e-06,
      "loss": 1.0161,
      "step": 60
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.9154643005374183,
      "learning_rate": 9.423808985883289e-06,
      "loss": 0.9014,
      "step": 61
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.9253885360090393,
      "learning_rate": 9.401006955900555e-06,
      "loss": 0.9921,
      "step": 62
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.8131597896526979,
      "learning_rate": 9.377791156510456e-06,
      "loss": 0.9996,
      "step": 63
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.047402757936845,
      "learning_rate": 9.35416377039246e-06,
      "loss": 1.0432,
      "step": 64
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.42467076534372,
      "learning_rate": 9.330127018922195e-06,
      "loss": 1.0092,
      "step": 65
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.0406186242464015,
      "learning_rate": 9.305683161962569e-06,
      "loss": 1.0197,
      "step": 66
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.996085329280642,
      "learning_rate": 9.280834497651334e-06,
      "loss": 0.999,
      "step": 67
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.050405093618206,
      "learning_rate": 9.255583362184998e-06,
      "loss": 1.0275,
      "step": 68
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.9391779248159844,
      "learning_rate": 9.229932129599206e-06,
      "loss": 1.0069,
      "step": 69
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8688340050178358,
      "learning_rate": 9.203883211545517e-06,
      "loss": 1.024,
      "step": 70
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.9107718382648906,
      "learning_rate": 9.177439057064684e-06,
      "loss": 0.9796,
      "step": 71
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8884098537907934,
      "learning_rate": 9.150602152356394e-06,
      "loss": 1.0094,
      "step": 72
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.0666939259341337,
      "learning_rate": 9.123375020545534e-06,
      "loss": 1.1245,
      "step": 73
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.9312576824640348,
      "learning_rate": 9.09576022144496e-06,
      "loss": 0.9542,
      "step": 74
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.150486624339303,
      "learning_rate": 9.067760351314838e-06,
      "loss": 0.9796,
      "step": 75
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.1900483954756464,
      "learning_rate": 9.039378042618556e-06,
      "loss": 1.1277,
      "step": 76
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.8759460923333156,
      "learning_rate": 9.01061596377522e-06,
      "loss": 0.9205,
      "step": 77
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.0684517325582923,
      "learning_rate": 8.981476818908778e-06,
      "loss": 1.0655,
      "step": 78
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.7015404285579259,
      "learning_rate": 8.951963347593797e-06,
      "loss": 0.9322,
      "step": 79
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.0128565561264113,
      "learning_rate": 8.92207832459788e-06,
      "loss": 1.0096,
      "step": 80
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.2955683670099836,
      "learning_rate": 8.891824559620801e-06,
      "loss": 0.93,
      "step": 81
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.9819221656020602,
      "learning_rate": 8.861204897030346e-06,
      "loss": 0.8111,
      "step": 82
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.8840968076917093,
      "learning_rate": 8.83022221559489e-06,
      "loss": 0.8334,
      "step": 83
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.8148160639213984,
      "learning_rate": 8.798879428212748e-06,
      "loss": 0.9797,
      "step": 84
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.7219644312125897,
      "learning_rate": 8.767179481638303e-06,
      "loss": 0.9265,
      "step": 85
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.8153657619801007,
      "learning_rate": 8.735125356204982e-06,
      "loss": 0.9308,
      "step": 86
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.1864487165433717,
      "learning_rate": 8.702720065545024e-06,
      "loss": 0.8714,
      "step": 87
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.9319176084767922,
      "learning_rate": 8.669966656306176e-06,
      "loss": 0.8996,
      "step": 88
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.8941836323323051,
      "learning_rate": 8.636868207865244e-06,
      "loss": 0.9559,
      "step": 89
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.0942714525772264,
      "learning_rate": 8.603427832038574e-06,
      "loss": 0.9642,
      "step": 90
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.956607567158922,
      "learning_rate": 8.569648672789496e-06,
      "loss": 0.9556,
      "step": 91
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.0359194563727008,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.9433,
      "step": 92
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.8831570100390131,
      "learning_rate": 8.501086738835843e-06,
      "loss": 0.9653,
      "step": 93
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.970446143582905,
      "learning_rate": 8.466310410117622e-06,
      "loss": 0.9637,
      "step": 94
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.0437287034957903,
      "learning_rate": 8.43120818934367e-06,
      "loss": 0.9639,
      "step": 95
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.9551401347597726,
      "learning_rate": 8.395783376718967e-06,
      "loss": 0.908,
      "step": 96
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.9195968647766901,
      "learning_rate": 8.360039302777614e-06,
      "loss": 0.9048,
      "step": 97
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.853992686163623,
      "learning_rate": 8.323979328069689e-06,
      "loss": 0.9351,
      "step": 98
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.8308983331681774,
      "learning_rate": 8.28760684284532e-06,
      "loss": 0.9443,
      "step": 99
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.1769204299308758,
      "learning_rate": 8.25092526673592e-06,
      "loss": 1.0337,
      "step": 100
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.0101187852614335,
      "learning_rate": 8.213938048432697e-06,
      "loss": 0.9319,
      "step": 101
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.1098348893292616,
      "learning_rate": 8.176648665362426e-06,
      "loss": 0.8138,
      "step": 102
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.018298542101901,
      "learning_rate": 8.139060623360494e-06,
      "loss": 0.9322,
      "step": 103
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.97268436525157,
      "learning_rate": 8.101177456341301e-06,
      "loss": 0.908,
      "step": 104
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.8905564445487661,
      "learning_rate": 8.063002725966014e-06,
      "loss": 0.9059,
      "step": 105
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.8040796552589426,
      "learning_rate": 8.024540021307709e-06,
      "loss": 0.8834,
      "step": 106
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.8099754444058382,
      "learning_rate": 7.985792958513932e-06,
      "loss": 0.949,
      "step": 107
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.859283445011557,
      "learning_rate": 7.946765180466725e-06,
      "loss": 0.8598,
      "step": 108
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.017173689625784,
      "learning_rate": 7.907460356440133e-06,
      "loss": 0.98,
      "step": 109
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.9067977524786353,
      "learning_rate": 7.86788218175523e-06,
      "loss": 0.9113,
      "step": 110
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.8356374617177302,
      "learning_rate": 7.828034377432694e-06,
      "loss": 0.9379,
      "step": 111
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.9671234496145011,
      "learning_rate": 7.787920689842965e-06,
      "loss": 0.971,
      "step": 112
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.9057301922594099,
      "learning_rate": 7.747544890354031e-06,
      "loss": 0.9351,
      "step": 113
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.8176638279870596,
      "learning_rate": 7.706910774976849e-06,
      "loss": 0.8129,
      "step": 114
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.8480732550035386,
      "learning_rate": 7.666022164008458e-06,
      "loss": 0.9127,
      "step": 115
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.8342675263668593,
      "learning_rate": 7.624882901672801e-06,
      "loss": 0.8875,
      "step": 116
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.8479911871602939,
      "learning_rate": 7.5834968557593155e-06,
      "loss": 0.9814,
      "step": 117
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.953930833517739,
      "learning_rate": 7.541867917259278e-06,
      "loss": 1.0459,
      "step": 118
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.816248946402888,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.8606,
      "step": 119
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.829758704776544,
      "learning_rate": 7.457897040276853e-06,
      "loss": 0.9917,
      "step": 120
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.7798304171360442,
      "learning_rate": 7.415562996483193e-06,
      "loss": 0.9226,
      "step": 121
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.949746360544544,
      "learning_rate": 7.373001848738203e-06,
      "loss": 1.0089,
      "step": 122
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.7709880415336614,
      "learning_rate": 7.330217598512696e-06,
      "loss": 0.8108,
      "step": 123
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.8079747568178568,
      "learning_rate": 7.2872142682529045e-06,
      "loss": 0.8246,
      "step": 124
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.8171645027732521,
      "learning_rate": 7.243995901002312e-06,
      "loss": 0.9032,
      "step": 125
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.875345818100136,
      "learning_rate": 7.200566560021525e-06,
      "loss": 0.8692,
      "step": 126
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.9933228492567425,
      "learning_rate": 7.156930328406268e-06,
      "loss": 0.93,
      "step": 127
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.9411283328658857,
      "learning_rate": 7.113091308703498e-06,
      "loss": 0.9071,
      "step": 128
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.090341113079029,
      "learning_rate": 7.069053622525697e-06,
      "loss": 1.043,
      "step": 129
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.91058658362823,
      "learning_rate": 7.0248214101633685e-06,
      "loss": 0.8674,
      "step": 130
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.9696219193781328,
      "learning_rate": 6.980398830195785e-06,
      "loss": 1.0577,
      "step": 131
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.9555717096721157,
      "learning_rate": 6.9357900591000034e-06,
      "loss": 0.9531,
      "step": 132
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.8951036355458846,
      "learning_rate": 6.890999290858213e-06,
      "loss": 1.006,
      "step": 133
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.8568126429433485,
      "learning_rate": 6.8460307365634225e-06,
      "loss": 0.8623,
      "step": 134
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.8704210955739529,
      "learning_rate": 6.800888624023552e-06,
      "loss": 0.9493,
      "step": 135
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.832260849480108,
      "learning_rate": 6.755577197363945e-06,
      "loss": 0.8205,
      "step": 136
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.031520077283429,
      "learning_rate": 6.710100716628345e-06,
      "loss": 0.8792,
      "step": 137
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.7410235600999984,
      "learning_rate": 6.6644634573783825e-06,
      "loss": 0.8951,
      "step": 138
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.843737618356284,
      "learning_rate": 6.618669710291607e-06,
      "loss": 0.8085,
      "step": 139
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.765762644440963,
      "learning_rate": 6.572723780758069e-06,
      "loss": 0.8957,
      "step": 140
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.9511281601416317,
      "learning_rate": 6.526629988475567e-06,
      "loss": 0.9,
      "step": 141
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.7773818546444415,
      "learning_rate": 6.4803926670435e-06,
      "loss": 0.9234,
      "step": 142
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.8692189103556698,
      "learning_rate": 6.434016163555452e-06,
      "loss": 0.8899,
      "step": 143
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.8255302240308842,
      "learning_rate": 6.387504838190479e-06,
      "loss": 0.8086,
      "step": 144
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.8578841539686264,
      "learning_rate": 6.340863063803187e-06,
      "loss": 0.8248,
      "step": 145
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.7994136623563066,
      "learning_rate": 6.294095225512604e-06,
      "loss": 0.933,
      "step": 146
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.0121355236133978,
      "learning_rate": 6.247205720289907e-06,
      "loss": 1.0521,
      "step": 147
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.869460952399974,
      "learning_rate": 6.2001989565450305e-06,
      "loss": 0.8515,
      "step": 148
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.8246450967462686,
      "learning_rate": 6.153079353712201e-06,
      "loss": 0.852,
      "step": 149
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.9648734937960934,
      "learning_rate": 6.105851341834439e-06,
      "loss": 0.8434,
      "step": 150
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.7658190156459108,
      "learning_rate": 6.058519361147055e-06,
      "loss": 0.8406,
      "step": 151
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.8896576006369,
      "learning_rate": 6.011087861660191e-06,
      "loss": 0.8235,
      "step": 152
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.701511265987839,
      "learning_rate": 5.9635613027404495e-06,
      "loss": 0.83,
      "step": 153
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.8458019088871649,
      "learning_rate": 5.915944152691634e-06,
      "loss": 0.8061,
      "step": 154
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.8758600805480732,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.8738,
      "step": 155
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.7770461619321762,
      "learning_rate": 5.820455994586621e-06,
      "loss": 0.8144,
      "step": 156
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.06521139231319,
      "learning_rate": 5.772593964039203e-06,
      "loss": 1.0683,
      "step": 157
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.8949940538942156,
      "learning_rate": 5.724659296536234e-06,
      "loss": 0.8378,
      "step": 158
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.8329328421055082,
      "learning_rate": 5.6766564987506564e-06,
      "loss": 0.8059,
      "step": 159
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.9356837835306697,
      "learning_rate": 5.628590083760815e-06,
      "loss": 0.8759,
      "step": 160
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.914824151289401,
      "learning_rate": 5.5804645706261515e-06,
      "loss": 0.9575,
      "step": 161
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.9610745203544777,
      "learning_rate": 5.532284483962341e-06,
      "loss": 0.8548,
      "step": 162
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.9847850375923817,
      "learning_rate": 5.484054353515896e-06,
      "loss": 0.9005,
      "step": 163
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.900418899811939,
      "learning_rate": 5.435778713738292e-06,
      "loss": 0.9209,
      "step": 164
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.7892930994134306,
      "learning_rate": 5.387462103359655e-06,
      "loss": 0.8555,
      "step": 165
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.774160649412005,
      "learning_rate": 5.339109064962047e-06,
      "loss": 0.7846,
      "step": 166
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.9563920124901706,
      "learning_rate": 5.290724144552379e-06,
      "loss": 0.7538,
      "step": 167
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.9120830031418432,
      "learning_rate": 5.242311891135016e-06,
      "loss": 0.9287,
      "step": 168
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.7820515084191753,
      "learning_rate": 5.193876856284085e-06,
      "loss": 0.7843,
      "step": 169
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.8838516311972966,
      "learning_rate": 5.145423593715558e-06,
      "loss": 0.8848,
      "step": 170
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.8000718579254187,
      "learning_rate": 5.096956658859122e-06,
      "loss": 0.8422,
      "step": 171
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.9252361586236277,
      "learning_rate": 5.048480608429893e-06,
      "loss": 0.788,
      "step": 172
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.7838973065656873,
      "learning_rate": 5e-06,
      "loss": 0.8419,
      "step": 173
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.7406962046278718,
      "learning_rate": 4.951519391570108e-06,
      "loss": 0.8617,
      "step": 174
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.833052653443519,
      "learning_rate": 4.903043341140879e-06,
      "loss": 0.7839,
      "step": 175
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.9635309949293125,
      "learning_rate": 4.854576406284443e-06,
      "loss": 0.745,
      "step": 176
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.8878508304922175,
      "learning_rate": 4.806123143715916e-06,
      "loss": 0.898,
      "step": 177
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.8490253561704422,
      "learning_rate": 4.7576881088649865e-06,
      "loss": 0.8764,
      "step": 178
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.8492342801201405,
      "learning_rate": 4.7092758554476215e-06,
      "loss": 0.8734,
      "step": 179
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.678333641735353,
      "learning_rate": 4.660890935037954e-06,
      "loss": 0.8476,
      "step": 180
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.9197951126550667,
      "learning_rate": 4.6125378966403465e-06,
      "loss": 0.8502,
      "step": 181
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.8065516269516981,
      "learning_rate": 4.564221286261709e-06,
      "loss": 0.8434,
      "step": 182
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.090063964079408,
      "learning_rate": 4.515945646484105e-06,
      "loss": 0.9075,
      "step": 183
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.714803378036167,
      "learning_rate": 4.467715516037659e-06,
      "loss": 0.8979,
      "step": 184
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.8920297684515237,
      "learning_rate": 4.4195354293738484e-06,
      "loss": 0.917,
      "step": 185
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.9646366988613642,
      "learning_rate": 4.371409916239188e-06,
      "loss": 0.8406,
      "step": 186
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.0095779435113643,
      "learning_rate": 4.323343501249346e-06,
      "loss": 0.8105,
      "step": 187
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.84960716754979,
      "learning_rate": 4.275340703463767e-06,
      "loss": 0.709,
      "step": 188
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.8784422747688236,
      "learning_rate": 4.227406035960798e-06,
      "loss": 0.8721,
      "step": 189
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.7892389168859104,
      "learning_rate": 4.17954400541338e-06,
      "loss": 0.8237,
      "step": 190
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.834621493465168,
      "learning_rate": 4.131759111665349e-06,
      "loss": 0.858,
      "step": 191
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.0244248532984312,
      "learning_rate": 4.084055847308367e-06,
      "loss": 0.84,
      "step": 192
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.7956990250048366,
      "learning_rate": 4.036438697259551e-06,
      "loss": 0.8343,
      "step": 193
    },
    {
      "epoch": 0.58,
      "grad_norm": 2.008026972496449,
      "learning_rate": 3.988912138339812e-06,
      "loss": 0.858,
      "step": 194
    },
    {
      "epoch": 0.58,
      "grad_norm": 2.02397192889752,
      "learning_rate": 3.941480638852948e-06,
      "loss": 0.8685,
      "step": 195
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.8202366956390423,
      "learning_rate": 3.894148658165562e-06,
      "loss": 0.8428,
      "step": 196
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.8728629012942652,
      "learning_rate": 3.8469206462878e-06,
      "loss": 0.833,
      "step": 197
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.920110342063804,
      "learning_rate": 3.7998010434549716e-06,
      "loss": 0.923,
      "step": 198
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.8750275450908436,
      "learning_rate": 3.752794279710094e-06,
      "loss": 0.8113,
      "step": 199
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.7555094657763348,
      "learning_rate": 3.705904774487396e-06,
      "loss": 0.8442,
      "step": 200
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.9946734158465282,
      "learning_rate": 3.6591369361968127e-06,
      "loss": 0.8566,
      "step": 201
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.792346256212117,
      "learning_rate": 3.6124951618095224e-06,
      "loss": 0.8306,
      "step": 202
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.7737697004497117,
      "learning_rate": 3.5659838364445505e-06,
      "loss": 0.827,
      "step": 203
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.9896055506585861,
      "learning_rate": 3.519607332956502e-06,
      "loss": 0.8967,
      "step": 204
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.0112406931128954,
      "learning_rate": 3.473370011524435e-06,
      "loss": 0.9582,
      "step": 205
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6933223451923685,
      "learning_rate": 3.427276219241933e-06,
      "loss": 0.7972,
      "step": 206
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.7960032297149964,
      "learning_rate": 3.3813302897083955e-06,
      "loss": 0.91,
      "step": 207
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.8409756298969775,
      "learning_rate": 3.335536542621617e-06,
      "loss": 0.8442,
      "step": 208
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.8943440767619342,
      "learning_rate": 3.289899283371657e-06,
      "loss": 0.8782,
      "step": 209
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.983602448512353,
      "learning_rate": 3.244422802636057e-06,
      "loss": 0.9121,
      "step": 210
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.919904148023849,
      "learning_rate": 3.1991113759764493e-06,
      "loss": 0.8452,
      "step": 211
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7436938594681772,
      "learning_rate": 3.1539692634365788e-06,
      "loss": 0.8367,
      "step": 212
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.9023867686458018,
      "learning_rate": 3.1090007091417884e-06,
      "loss": 0.7576,
      "step": 213
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.8340656721562485,
      "learning_rate": 3.0642099408999982e-06,
      "loss": 0.9066,
      "step": 214
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.123899315024522,
      "learning_rate": 3.019601169804216e-06,
      "loss": 0.8908,
      "step": 215
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.970762510888818,
      "learning_rate": 2.975178589836632e-06,
      "loss": 0.8324,
      "step": 216
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.7131201352172498,
      "learning_rate": 2.9309463774743047e-06,
      "loss": 0.9371,
      "step": 217
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.692464765973636,
      "learning_rate": 2.886908691296504e-06,
      "loss": 0.8384,
      "step": 218
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.754853534401181,
      "learning_rate": 2.843069671593734e-06,
      "loss": 0.8639,
      "step": 219
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7746221308047176,
      "learning_rate": 2.7994334399784773e-06,
      "loss": 0.8123,
      "step": 220
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7797496232199694,
      "learning_rate": 2.7560040989976894e-06,
      "loss": 0.8011,
      "step": 221
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7965856236183813,
      "learning_rate": 2.7127857317470967e-06,
      "loss": 0.8614,
      "step": 222
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.895456729064966,
      "learning_rate": 2.6697824014873076e-06,
      "loss": 0.808,
      "step": 223
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6731115830560004,
      "learning_rate": 2.626998151261798e-06,
      "loss": 0.7644,
      "step": 224
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.722068279724202,
      "learning_rate": 2.5844370035168077e-06,
      "loss": 0.7876,
      "step": 225
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.8560082384134242,
      "learning_rate": 2.5421029597231476e-06,
      "loss": 0.8448,
      "step": 226
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.8503877478285864,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.8241,
      "step": 227
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.9115184771008713,
      "learning_rate": 2.458132082740724e-06,
      "loss": 0.8465,
      "step": 228
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.9363363832594316,
      "learning_rate": 2.4165031442406857e-06,
      "loss": 0.8231,
      "step": 229
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.9209239281193067,
      "learning_rate": 2.3751170983272e-06,
      "loss": 0.9008,
      "step": 230
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.8848510564703413,
      "learning_rate": 2.333977835991545e-06,
      "loss": 0.789,
      "step": 231
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.885613538323485,
      "learning_rate": 2.293089225023152e-06,
      "loss": 0.9544,
      "step": 232
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.9378339418142523,
      "learning_rate": 2.2524551096459703e-06,
      "loss": 0.8512,
      "step": 233
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.8499392383494713,
      "learning_rate": 2.2120793101570366e-06,
      "loss": 0.7881,
      "step": 234
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.8123212923542507,
      "learning_rate": 2.171965622567308e-06,
      "loss": 0.8658,
      "step": 235
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.8978780385206444,
      "learning_rate": 2.132117818244771e-06,
      "loss": 0.8045,
      "step": 236
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.7796628138142092,
      "learning_rate": 2.0925396435598665e-06,
      "loss": 0.813,
      "step": 237
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.8442874060590642,
      "learning_rate": 2.053234819533276e-06,
      "loss": 0.7706,
      "step": 238
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.8289341236387466,
      "learning_rate": 2.0142070414860704e-06,
      "loss": 0.8927,
      "step": 239
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.9766538998425796,
      "learning_rate": 1.9754599786922913e-06,
      "loss": 0.848,
      "step": 240
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.7214028779109611,
      "learning_rate": 1.936997274033986e-06,
      "loss": 0.7225,
      "step": 241
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5814120081508478,
      "learning_rate": 1.8988225436587005e-06,
      "loss": 0.7288,
      "step": 242
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7940729269240667,
      "learning_rate": 1.8609393766395083e-06,
      "loss": 0.9349,
      "step": 243
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.878940225083856,
      "learning_rate": 1.823351334637576e-06,
      "loss": 0.8962,
      "step": 244
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.8155856668279255,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 0.88,
      "step": 245
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.9394793705542583,
      "learning_rate": 1.7490747332640833e-06,
      "loss": 0.897,
      "step": 246
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.8855178833940114,
      "learning_rate": 1.7123931571546826e-06,
      "loss": 0.8955,
      "step": 247
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.680634839138001,
      "learning_rate": 1.6760206719303107e-06,
      "loss": 0.781,
      "step": 248
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.8281579952033755,
      "learning_rate": 1.639960697222388e-06,
      "loss": 0.9578,
      "step": 249
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.8390758231951718,
      "learning_rate": 1.6042166232810346e-06,
      "loss": 0.8377,
      "step": 250
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.637452500506161,
      "learning_rate": 1.5687918106563326e-06,
      "loss": 0.7837,
      "step": 251
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.8615373812049087,
      "learning_rate": 1.5336895898823801e-06,
      "loss": 0.8947,
      "step": 252
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.8795666079438391,
      "learning_rate": 1.4989132611641576e-06,
      "loss": 0.8214,
      "step": 253
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.1220008217273385,
      "learning_rate": 1.4644660940672628e-06,
      "loss": 0.8473,
      "step": 254
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.016744480195016,
      "learning_rate": 1.4303513272105057e-06,
      "loss": 0.7648,
      "step": 255
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.8273845868646243,
      "learning_rate": 1.396572167961427e-06,
      "loss": 0.9056,
      "step": 256
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.9242593040018354,
      "learning_rate": 1.3631317921347564e-06,
      "loss": 0.8546,
      "step": 257
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7602720727922225,
      "learning_rate": 1.330033343693824e-06,
      "loss": 0.8168,
      "step": 258
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7981469172314437,
      "learning_rate": 1.297279934454978e-06,
      "loss": 0.8196,
      "step": 259
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.934034355132605,
      "learning_rate": 1.264874643795021e-06,
      "loss": 0.8641,
      "step": 260
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.731914923464836,
      "learning_rate": 1.2328205183616964e-06,
      "loss": 0.8316,
      "step": 261
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.8671301669332583,
      "learning_rate": 1.2011205717872538e-06,
      "loss": 0.8973,
      "step": 262
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.8258264225083662,
      "learning_rate": 1.1697777844051105e-06,
      "loss": 0.827,
      "step": 263
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.76891519967969,
      "learning_rate": 1.1387951029696543e-06,
      "loss": 0.8073,
      "step": 264
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.8056333135913247,
      "learning_rate": 1.1081754403792e-06,
      "loss": 0.8111,
      "step": 265
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.6693657914010887,
      "learning_rate": 1.0779216754021215e-06,
      "loss": 0.8481,
      "step": 266
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.8914182275757614,
      "learning_rate": 1.0480366524062041e-06,
      "loss": 0.7854,
      "step": 267
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.7952669162789925,
      "learning_rate": 1.0185231810912223e-06,
      "loss": 0.8331,
      "step": 268
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.8041970383317338,
      "learning_rate": 9.893840362247809e-07,
      "loss": 0.808,
      "step": 269
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7349503739814947,
      "learning_rate": 9.606219573814447e-07,
      "loss": 0.782,
      "step": 270
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.971273696916632,
      "learning_rate": 9.322396486851626e-07,
      "loss": 0.8777,
      "step": 271
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6916809211406005,
      "learning_rate": 9.042397785550405e-07,
      "loss": 0.816,
      "step": 272
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.859123501478176,
      "learning_rate": 8.766249794544662e-07,
      "loss": 0.8536,
      "step": 273
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.9296428042034799,
      "learning_rate": 8.49397847643606e-07,
      "loss": 0.9167,
      "step": 274
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.9014753831076923,
      "learning_rate": 8.225609429353187e-07,
      "loss": 0.8984,
      "step": 275
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.9966126008374474,
      "learning_rate": 7.961167884544852e-07,
      "loss": 0.8615,
      "step": 276
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.7845802462997415,
      "learning_rate": 7.700678704007947e-07,
      "loss": 0.8524,
      "step": 277
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.786614071106851,
      "learning_rate": 7.444166378150014e-07,
      "loss": 0.8085,
      "step": 278
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.6646768752142675,
      "learning_rate": 7.191655023486682e-07,
      "loss": 0.867,
      "step": 279
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.934324323024249,
      "learning_rate": 6.94316838037431e-07,
      "loss": 0.724,
      "step": 280
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7670716798370054,
      "learning_rate": 6.698729810778065e-07,
      "loss": 0.7882,
      "step": 281
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.8041863674588816,
      "learning_rate": 6.458362296075399e-07,
      "loss": 0.7681,
      "step": 282
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7612589809223944,
      "learning_rate": 6.222088434895462e-07,
      "loss": 0.8392,
      "step": 283
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.749669895373114,
      "learning_rate": 5.989930440994451e-07,
      "loss": 0.7651,
      "step": 284
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.856400426555766,
      "learning_rate": 5.76191014116711e-07,
      "loss": 0.854,
      "step": 285
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.786307733562441,
      "learning_rate": 5.538048973194699e-07,
      "loss": 0.8124,
      "step": 286
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7598458461279902,
      "learning_rate": 5.318367983829393e-07,
      "loss": 0.822,
      "step": 287
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.6759526002862657,
      "learning_rate": 5.102887826815589e-07,
      "loss": 0.7365,
      "step": 288
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7936397119836314,
      "learning_rate": 4.891628760948114e-07,
      "loss": 0.7666,
      "step": 289
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.8498525973183433,
      "learning_rate": 4.6846106481675035e-07,
      "loss": 0.8682,
      "step": 290
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.7420235116184377,
      "learning_rate": 4.481852951692672e-07,
      "loss": 0.7485,
      "step": 291
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.877611598521712,
      "learning_rate": 4.283374734191037e-07,
      "loss": 0.9166,
      "step": 292
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.9492518664893324,
      "learning_rate": 4.089194655986306e-07,
      "loss": 0.7579,
      "step": 293
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.9457494252701653,
      "learning_rate": 3.899330973304083e-07,
      "loss": 0.9035,
      "step": 294
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.9791967456606612,
      "learning_rate": 3.7138015365554834e-07,
      "loss": 0.7953,
      "step": 295
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.794711427524111,
      "learning_rate": 3.5326237886588734e-07,
      "loss": 0.847,
      "step": 296
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.8794850272880976,
      "learning_rate": 3.355814763399973e-07,
      "loss": 0.8514,
      "step": 297
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.790194856863824,
      "learning_rate": 3.183391083830345e-07,
      "loss": 0.8278,
      "step": 298
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.8717181093829611,
      "learning_rate": 3.015368960704584e-07,
      "loss": 0.8429,
      "step": 299
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.7463890355109786,
      "learning_rate": 2.8517641909562075e-07,
      "loss": 0.8278,
      "step": 300
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.855252737220356,
      "learning_rate": 2.6925921562124867e-07,
      "loss": 0.6983,
      "step": 301
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.8352359161688767,
      "learning_rate": 2.5378678213483057e-07,
      "loss": 0.8415,
      "step": 302
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.8346745144152707,
      "learning_rate": 2.3876057330792344e-07,
      "loss": 0.8257,
      "step": 303
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.8337279964237236,
      "learning_rate": 2.2418200185938488e-07,
      "loss": 0.8155,
      "step": 304
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.9002602072773978,
      "learning_rate": 2.1005243842255552e-07,
      "loss": 0.9803,
      "step": 305
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.7966402315107077,
      "learning_rate": 1.9637321141639743e-07,
      "loss": 0.7586,
      "step": 306
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.8900826834954527,
      "learning_rate": 1.8314560692059836e-07,
      "loss": 0.8701,
      "step": 307
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.986051932218701,
      "learning_rate": 1.7037086855465902e-07,
      "loss": 0.854,
      "step": 308
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.819134244934257,
      "learning_rate": 1.5805019736097105e-07,
      "loss": 0.8146,
      "step": 309
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.7730810336503418,
      "learning_rate": 1.4618475169190017e-07,
      "loss": 0.8358,
      "step": 310
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.8749266769059667,
      "learning_rate": 1.3477564710088097e-07,
      "loss": 0.8806,
      "step": 311
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.7637217906423568,
      "learning_rate": 1.2382395623753484e-07,
      "loss": 0.8348,
      "step": 312
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.7705930846058242,
      "learning_rate": 1.1333070874682217e-07,
      "loss": 0.8438,
      "step": 313
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.9290513445644262,
      "learning_rate": 1.0329689117224262e-07,
      "loss": 0.8837,
      "step": 314
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.7184442508285342,
      "learning_rate": 9.372344686307655e-08,
      "loss": 0.7542,
      "step": 315
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.7508169719519233,
      "learning_rate": 8.461127588570039e-08,
      "loss": 0.8068,
      "step": 316
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.072932043609583,
      "learning_rate": 7.59612349389599e-08,
      "loss": 0.8847,
      "step": 317
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.8418737822218936,
      "learning_rate": 6.777413727363069e-08,
      "loss": 0.7607,
      "step": 318
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.8728345287191652,
      "learning_rate": 6.005075261595495e-08,
      "loss": 0.8279,
      "step": 319
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.9751421080539375,
      "learning_rate": 5.279180709527765e-08,
      "loss": 0.7666,
      "step": 320
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.742439764943798,
      "learning_rate": 4.599798317577342e-08,
      "loss": 0.8479,
      "step": 321
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.8692495061239316,
      "learning_rate": 3.9669919592288385e-08,
      "loss": 0.8109,
      "step": 322
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.670765464035315,
      "learning_rate": 3.3808211290284886e-08,
      "loss": 0.7585,
      "step": 323
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.062997844606162,
      "learning_rate": 2.8413409369907887e-08,
      "loss": 0.8709,
      "step": 324
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.9091961716040289,
      "learning_rate": 2.3486021034170857e-08,
      "loss": 0.818,
      "step": 325
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.8898711790354914,
      "learning_rate": 1.9026509541272276e-08,
      "loss": 0.8346,
      "step": 326
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.6576693409221155,
      "learning_rate": 1.5035294161039882e-08,
      "loss": 0.7996,
      "step": 327
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.7571989556959642,
      "learning_rate": 1.1512750135511674e-08,
      "loss": 0.7742,
      "step": 328
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.8431201117480818,
      "learning_rate": 8.459208643659122e-09,
      "loss": 0.8858,
      "step": 329
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.8220210754565818,
      "learning_rate": 5.874956770248186e-09,
      "loss": 0.8481,
      "step": 330
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.8136218315284558,
      "learning_rate": 3.760237478849793e-09,
      "loss": 0.6899,
      "step": 331
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.9765791609593397,
      "learning_rate": 2.1152495889970035e-09,
      "loss": 0.8208,
      "step": 332
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.8275429866969195,
      "learning_rate": 9.401477574932927e-10,
      "loss": 0.8589,
      "step": 333
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.8063515591696528,
      "learning_rate": 2.3504246386918394e-10,
      "loss": 0.781,
      "step": 334
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.9072496888398838,
      "learning_rate": 0.0,
      "loss": 0.8156,
      "step": 335
    },
    {
      "epoch": 1.0,
      "step": 335,
      "total_flos": 79804470312960.0,
      "train_loss": 0.9642956034461064,
      "train_runtime": 7158.0735,
      "train_samples_per_second": 0.374,
      "train_steps_per_second": 0.047
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 335,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2056,
  "total_flos": 79804470312960.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}