{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.992911392405063,
  "eval_steps": 500,
  "global_step": 1230,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004050632911392405,
      "grad_norm": 6.267872724399711,
      "learning_rate": 3.2520325203252037e-07,
      "loss": 1.0318,
      "step": 1
    },
    {
      "epoch": 0.00810126582278481,
      "grad_norm": 6.326748867654013,
      "learning_rate": 6.504065040650407e-07,
      "loss": 1.0362,
      "step": 2
    },
    {
      "epoch": 0.012151898734177215,
      "grad_norm": 6.4512224906290045,
      "learning_rate": 9.75609756097561e-07,
      "loss": 1.0451,
      "step": 3
    },
    {
      "epoch": 0.01620253164556962,
      "grad_norm": 6.331043240164785,
      "learning_rate": 1.3008130081300815e-06,
      "loss": 1.0315,
      "step": 4
    },
    {
      "epoch": 0.020253164556962026,
      "grad_norm": 6.008751321558461,
      "learning_rate": 1.6260162601626018e-06,
      "loss": 1.0318,
      "step": 5
    },
    {
      "epoch": 0.02430379746835443,
      "grad_norm": 5.744755336134939,
      "learning_rate": 1.951219512195122e-06,
      "loss": 1.0341,
      "step": 6
    },
    {
      "epoch": 0.028354430379746835,
      "grad_norm": 4.501500311867758,
      "learning_rate": 2.2764227642276426e-06,
      "loss": 0.988,
      "step": 7
    },
    {
      "epoch": 0.03240506329113924,
      "grad_norm": 4.194837495264538,
      "learning_rate": 2.601626016260163e-06,
      "loss": 0.9897,
      "step": 8
    },
    {
      "epoch": 0.03645569620253165,
      "grad_norm": 2.7261328178309,
      "learning_rate": 2.926829268292683e-06,
      "loss": 0.9717,
      "step": 9
    },
    {
      "epoch": 0.04050632911392405,
      "grad_norm": 2.5599813418284407,
      "learning_rate": 3.2520325203252037e-06,
      "loss": 0.9558,
      "step": 10
    },
    {
      "epoch": 0.044556962025316456,
      "grad_norm": 2.5049655673890014,
      "learning_rate": 3.577235772357724e-06,
      "loss": 0.9581,
      "step": 11
    },
    {
      "epoch": 0.04860759493670886,
      "grad_norm": 3.7642705402513488,
      "learning_rate": 3.902439024390244e-06,
      "loss": 0.9367,
      "step": 12
    },
    {
      "epoch": 0.052658227848101265,
      "grad_norm": 4.122407627161563,
      "learning_rate": 4.227642276422765e-06,
      "loss": 0.9219,
      "step": 13
    },
    {
      "epoch": 0.05670886075949367,
      "grad_norm": 4.044000827928146,
      "learning_rate": 4.552845528455285e-06,
      "loss": 0.9427,
      "step": 14
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 3.71847356841669,
      "learning_rate": 4.8780487804878055e-06,
      "loss": 0.9265,
      "step": 15
    },
    {
      "epoch": 0.06481012658227848,
      "grad_norm": 2.783007919958198,
      "learning_rate": 5.203252032520326e-06,
      "loss": 0.8881,
      "step": 16
    },
    {
      "epoch": 0.06886075949367089,
      "grad_norm": 2.5352205848958316,
      "learning_rate": 5.528455284552846e-06,
      "loss": 0.889,
      "step": 17
    },
    {
      "epoch": 0.0729113924050633,
      "grad_norm": 2.0686996174325447,
      "learning_rate": 5.853658536585366e-06,
      "loss": 0.8589,
      "step": 18
    },
    {
      "epoch": 0.0769620253164557,
      "grad_norm": 1.5276515527214722,
      "learning_rate": 6.178861788617887e-06,
      "loss": 0.8528,
      "step": 19
    },
    {
      "epoch": 0.0810126582278481,
      "grad_norm": 1.1326350883408838,
      "learning_rate": 6.504065040650407e-06,
      "loss": 0.8342,
      "step": 20
    },
    {
      "epoch": 0.08506329113924051,
      "grad_norm": 1.0783825344758553,
      "learning_rate": 6.829268292682928e-06,
      "loss": 0.83,
      "step": 21
    },
    {
      "epoch": 0.08911392405063291,
      "grad_norm": 1.2060700957675083,
      "learning_rate": 7.154471544715448e-06,
      "loss": 0.8267,
      "step": 22
    },
    {
      "epoch": 0.09316455696202532,
      "grad_norm": 1.129321046006167,
      "learning_rate": 7.4796747967479676e-06,
      "loss": 0.8175,
      "step": 23
    },
    {
      "epoch": 0.09721518987341772,
      "grad_norm": 0.9070211157102742,
      "learning_rate": 7.804878048780489e-06,
      "loss": 0.8025,
      "step": 24
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.9150611038461398,
      "learning_rate": 8.130081300813009e-06,
      "loss": 0.7856,
      "step": 25
    },
    {
      "epoch": 0.10531645569620253,
      "grad_norm": 0.883057370283813,
      "learning_rate": 8.45528455284553e-06,
      "loss": 0.789,
      "step": 26
    },
    {
      "epoch": 0.10936708860759493,
      "grad_norm": 0.8174541484063313,
      "learning_rate": 8.78048780487805e-06,
      "loss": 0.7627,
      "step": 27
    },
    {
      "epoch": 0.11341772151898734,
      "grad_norm": 0.8301141133287909,
      "learning_rate": 9.10569105691057e-06,
      "loss": 0.7842,
      "step": 28
    },
    {
      "epoch": 0.11746835443037974,
      "grad_norm": 0.7422043145548739,
      "learning_rate": 9.43089430894309e-06,
      "loss": 0.7473,
      "step": 29
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 0.7711284126817353,
      "learning_rate": 9.756097560975611e-06,
      "loss": 0.7696,
      "step": 30
    },
    {
      "epoch": 0.12556962025316457,
      "grad_norm": 0.7074324067983795,
      "learning_rate": 1.008130081300813e-05,
      "loss": 0.7524,
      "step": 31
    },
    {
      "epoch": 0.12962025316455697,
      "grad_norm": 0.6532935382997638,
      "learning_rate": 1.0406504065040652e-05,
      "loss": 0.7518,
      "step": 32
    },
    {
      "epoch": 0.13367088607594937,
      "grad_norm": 0.7024393288761571,
      "learning_rate": 1.0731707317073172e-05,
      "loss": 0.7602,
      "step": 33
    },
    {
      "epoch": 0.13772151898734178,
      "grad_norm": 0.6644580967730146,
      "learning_rate": 1.1056910569105692e-05,
      "loss": 0.7488,
      "step": 34
    },
    {
      "epoch": 0.14177215189873418,
      "grad_norm": 0.7026479346217505,
      "learning_rate": 1.1382113821138213e-05,
      "loss": 0.7293,
      "step": 35
    },
    {
      "epoch": 0.1458227848101266,
      "grad_norm": 0.5355246445144775,
      "learning_rate": 1.1707317073170731e-05,
      "loss": 0.728,
      "step": 36
    },
    {
      "epoch": 0.149873417721519,
      "grad_norm": 0.7526778631404095,
      "learning_rate": 1.2032520325203254e-05,
      "loss": 0.7198,
      "step": 37
    },
    {
      "epoch": 0.1539240506329114,
      "grad_norm": 0.6115728706793658,
      "learning_rate": 1.2357723577235774e-05,
      "loss": 0.7401,
      "step": 38
    },
    {
      "epoch": 0.1579746835443038,
      "grad_norm": 0.5083235141951238,
      "learning_rate": 1.2682926829268294e-05,
      "loss": 0.719,
      "step": 39
    },
    {
      "epoch": 0.1620253164556962,
      "grad_norm": 0.580015093532721,
      "learning_rate": 1.3008130081300815e-05,
      "loss": 0.7124,
      "step": 40
    },
    {
      "epoch": 0.1660759493670886,
      "grad_norm": 0.5241642723943151,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.7131,
      "step": 41
    },
    {
      "epoch": 0.17012658227848101,
      "grad_norm": 0.5598189805475433,
      "learning_rate": 1.3658536585365855e-05,
      "loss": 0.7295,
      "step": 42
    },
    {
      "epoch": 0.17417721518987342,
      "grad_norm": 0.5694695362906147,
      "learning_rate": 1.3983739837398376e-05,
      "loss": 0.7206,
      "step": 43
    },
    {
      "epoch": 0.17822784810126582,
      "grad_norm": 0.5242894493848274,
      "learning_rate": 1.4308943089430896e-05,
      "loss": 0.7217,
      "step": 44
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 0.4972633187259291,
      "learning_rate": 1.4634146341463415e-05,
      "loss": 0.7039,
      "step": 45
    },
    {
      "epoch": 0.18632911392405063,
      "grad_norm": 0.5413644008948799,
      "learning_rate": 1.4959349593495935e-05,
      "loss": 0.7215,
      "step": 46
    },
    {
      "epoch": 0.19037974683544304,
      "grad_norm": 0.5647497598983527,
      "learning_rate": 1.528455284552846e-05,
      "loss": 0.7094,
      "step": 47
    },
    {
      "epoch": 0.19443037974683544,
      "grad_norm": 0.45166760193682776,
      "learning_rate": 1.5609756097560978e-05,
      "loss": 0.7038,
      "step": 48
    },
    {
      "epoch": 0.19848101265822785,
      "grad_norm": 0.5038741798470056,
      "learning_rate": 1.5934959349593496e-05,
      "loss": 0.701,
      "step": 49
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.5225355106005427,
      "learning_rate": 1.6260162601626018e-05,
      "loss": 0.7111,
      "step": 50
    },
    {
      "epoch": 0.20658227848101265,
      "grad_norm": 0.42904695949111665,
      "learning_rate": 1.6585365853658537e-05,
      "loss": 0.6968,
      "step": 51
    },
    {
      "epoch": 0.21063291139240506,
      "grad_norm": 0.45293533492327254,
      "learning_rate": 1.691056910569106e-05,
      "loss": 0.7017,
      "step": 52
    },
    {
      "epoch": 0.21468354430379746,
      "grad_norm": 0.6080862526027445,
      "learning_rate": 1.7235772357723578e-05,
      "loss": 0.6991,
      "step": 53
    },
    {
      "epoch": 0.21873417721518987,
      "grad_norm": 0.57896770760262,
      "learning_rate": 1.75609756097561e-05,
      "loss": 0.6904,
      "step": 54
    },
    {
      "epoch": 0.22278481012658227,
      "grad_norm": 0.5464023506962115,
      "learning_rate": 1.788617886178862e-05,
      "loss": 0.6956,
      "step": 55
    },
    {
      "epoch": 0.22683544303797468,
      "grad_norm": 0.693358648391038,
      "learning_rate": 1.821138211382114e-05,
      "loss": 0.7036,
      "step": 56
    },
    {
      "epoch": 0.23088607594936708,
      "grad_norm": 0.6079077786196293,
      "learning_rate": 1.8536585365853663e-05,
      "loss": 0.6822,
      "step": 57
    },
    {
      "epoch": 0.23493670886075949,
      "grad_norm": 0.5521772770370854,
      "learning_rate": 1.886178861788618e-05,
      "loss": 0.6963,
      "step": 58
    },
    {
      "epoch": 0.2389873417721519,
      "grad_norm": 0.5692921611903945,
      "learning_rate": 1.91869918699187e-05,
      "loss": 0.6841,
      "step": 59
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 0.4277583163069273,
      "learning_rate": 1.9512195121951222e-05,
      "loss": 0.6843,
      "step": 60
    },
    {
      "epoch": 0.2470886075949367,
      "grad_norm": 0.5946444658791429,
      "learning_rate": 1.983739837398374e-05,
      "loss": 0.688,
      "step": 61
    },
    {
      "epoch": 0.25113924050632913,
      "grad_norm": 0.5193866518361615,
      "learning_rate": 2.016260162601626e-05,
      "loss": 0.6958,
      "step": 62
    },
    {
      "epoch": 0.25518987341772154,
      "grad_norm": 0.5506298399665518,
      "learning_rate": 2.048780487804878e-05,
      "loss": 0.6818,
      "step": 63
    },
    {
      "epoch": 0.25924050632911394,
      "grad_norm": 0.5842342548813322,
      "learning_rate": 2.0813008130081303e-05,
      "loss": 0.6866,
      "step": 64
    },
    {
      "epoch": 0.26329113924050634,
      "grad_norm": 0.6423887940602377,
      "learning_rate": 2.1138211382113822e-05,
      "loss": 0.669,
      "step": 65
    },
    {
      "epoch": 0.26734177215189875,
      "grad_norm": 0.646210695645398,
      "learning_rate": 2.1463414634146344e-05,
      "loss": 0.6698,
      "step": 66
    },
    {
      "epoch": 0.27139240506329115,
      "grad_norm": 0.8296824329026052,
      "learning_rate": 2.1788617886178863e-05,
      "loss": 0.6885,
      "step": 67
    },
    {
      "epoch": 0.27544303797468356,
      "grad_norm": 0.6730718572400709,
      "learning_rate": 2.2113821138211385e-05,
      "loss": 0.6641,
      "step": 68
    },
    {
      "epoch": 0.27949367088607596,
      "grad_norm": 0.7207669311028455,
      "learning_rate": 2.2439024390243907e-05,
      "loss": 0.6824,
      "step": 69
    },
    {
      "epoch": 0.28354430379746837,
      "grad_norm": 1.1649580733782816,
      "learning_rate": 2.2764227642276426e-05,
      "loss": 0.6828,
      "step": 70
    },
    {
      "epoch": 0.28759493670886077,
      "grad_norm": 1.040002607617539,
      "learning_rate": 2.3089430894308948e-05,
      "loss": 0.6695,
      "step": 71
    },
    {
      "epoch": 0.2916455696202532,
      "grad_norm": 0.7951691449017257,
      "learning_rate": 2.3414634146341463e-05,
      "loss": 0.663,
      "step": 72
    },
    {
      "epoch": 0.2956962025316456,
      "grad_norm": 0.8207260730491389,
      "learning_rate": 2.3739837398373985e-05,
      "loss": 0.6899,
      "step": 73
    },
    {
      "epoch": 0.299746835443038,
      "grad_norm": 0.8037932980878757,
      "learning_rate": 2.4065040650406507e-05,
      "loss": 0.6671,
      "step": 74
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.902114068595089,
      "learning_rate": 2.4390243902439026e-05,
      "loss": 0.6649,
      "step": 75
    },
    {
      "epoch": 0.3078481012658228,
      "grad_norm": 1.0619080670802745,
      "learning_rate": 2.4715447154471548e-05,
      "loss": 0.6651,
      "step": 76
    },
    {
      "epoch": 0.3118987341772152,
      "grad_norm": 1.0873503274000962,
      "learning_rate": 2.5040650406504066e-05,
      "loss": 0.663,
      "step": 77
    },
    {
      "epoch": 0.3159493670886076,
      "grad_norm": 1.1370688620274763,
      "learning_rate": 2.536585365853659e-05,
      "loss": 0.6709,
      "step": 78
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.8572076298641462,
      "learning_rate": 2.569105691056911e-05,
      "loss": 0.664,
      "step": 79
    },
    {
      "epoch": 0.3240506329113924,
      "grad_norm": 0.8362187882636416,
      "learning_rate": 2.601626016260163e-05,
      "loss": 0.6538,
      "step": 80
    },
    {
      "epoch": 0.3281012658227848,
      "grad_norm": 0.7817386701561712,
      "learning_rate": 2.634146341463415e-05,
      "loss": 0.6632,
      "step": 81
    },
    {
      "epoch": 0.3321518987341772,
      "grad_norm": 0.9028371663839354,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.6652,
      "step": 82
    },
    {
      "epoch": 0.3362025316455696,
      "grad_norm": 1.0367513031458186,
      "learning_rate": 2.699186991869919e-05,
      "loss": 0.6554,
      "step": 83
    },
    {
      "epoch": 0.34025316455696203,
      "grad_norm": 1.0697006804633573,
      "learning_rate": 2.731707317073171e-05,
      "loss": 0.6749,
      "step": 84
    },
    {
      "epoch": 0.34430379746835443,
      "grad_norm": 1.1717745271852935,
      "learning_rate": 2.764227642276423e-05,
      "loss": 0.675,
      "step": 85
    },
    {
      "epoch": 0.34835443037974684,
      "grad_norm": 1.3144098331853418,
      "learning_rate": 2.796747967479675e-05,
      "loss": 0.6579,
      "step": 86
    },
    {
      "epoch": 0.35240506329113924,
      "grad_norm": 1.271474973821305,
      "learning_rate": 2.829268292682927e-05,
      "loss": 0.669,
      "step": 87
    },
    {
      "epoch": 0.35645569620253165,
      "grad_norm": 1.0541040846764878,
      "learning_rate": 2.8617886178861792e-05,
      "loss": 0.6616,
      "step": 88
    },
    {
      "epoch": 0.36050632911392405,
      "grad_norm": 1.1531193074376433,
      "learning_rate": 2.8943089430894314e-05,
      "loss": 0.6582,
      "step": 89
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 1.5716357961110212,
      "learning_rate": 2.926829268292683e-05,
      "loss": 0.6759,
      "step": 90
    },
    {
      "epoch": 0.36860759493670886,
      "grad_norm": 0.7630741231316559,
      "learning_rate": 2.959349593495935e-05,
      "loss": 0.6794,
      "step": 91
    },
    {
      "epoch": 0.37265822784810126,
      "grad_norm": 1.5790253172013782,
      "learning_rate": 2.991869918699187e-05,
      "loss": 0.6676,
      "step": 92
    },
    {
      "epoch": 0.37670886075949367,
      "grad_norm": 0.7809102441019725,
      "learning_rate": 3.0243902439024392e-05,
      "loss": 0.677,
      "step": 93
    },
    {
      "epoch": 0.3807594936708861,
      "grad_norm": 1.061648856520805,
      "learning_rate": 3.056910569105692e-05,
      "loss": 0.6541,
      "step": 94
    },
    {
      "epoch": 0.3848101265822785,
      "grad_norm": 1.1537698047167397,
      "learning_rate": 3.089430894308943e-05,
      "loss": 0.6588,
      "step": 95
    },
    {
      "epoch": 0.3888607594936709,
      "grad_norm": 1.032394917899006,
      "learning_rate": 3.1219512195121955e-05,
      "loss": 0.6654,
      "step": 96
    },
    {
      "epoch": 0.3929113924050633,
      "grad_norm": 1.1859144492530167,
      "learning_rate": 3.154471544715447e-05,
      "loss": 0.6604,
      "step": 97
    },
    {
      "epoch": 0.3969620253164557,
      "grad_norm": 1.1252447900860991,
      "learning_rate": 3.186991869918699e-05,
      "loss": 0.6504,
      "step": 98
    },
    {
      "epoch": 0.4010126582278481,
      "grad_norm": 1.0317149402148316,
      "learning_rate": 3.2195121951219514e-05,
      "loss": 0.6607,
      "step": 99
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 1.0974131345676619,
      "learning_rate": 3.2520325203252037e-05,
      "loss": 0.6663,
      "step": 100
    },
    {
      "epoch": 0.4091139240506329,
      "grad_norm": 0.8995186616408428,
      "learning_rate": 3.284552845528456e-05,
      "loss": 0.6496,
      "step": 101
    },
    {
      "epoch": 0.4131645569620253,
      "grad_norm": 1.003593166336955,
      "learning_rate": 3.3170731707317074e-05,
      "loss": 0.6622,
      "step": 102
    },
    {
      "epoch": 0.4172151898734177,
      "grad_norm": 1.6764037487201724,
      "learning_rate": 3.3495934959349596e-05,
      "loss": 0.6744,
      "step": 103
    },
    {
      "epoch": 0.4212658227848101,
      "grad_norm": 0.906417500134716,
      "learning_rate": 3.382113821138212e-05,
      "loss": 0.6638,
      "step": 104
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 1.8129482587737875,
      "learning_rate": 3.414634146341463e-05,
      "loss": 0.662,
      "step": 105
    },
    {
      "epoch": 0.4293670886075949,
      "grad_norm": 1.3080715671875762,
      "learning_rate": 3.4471544715447155e-05,
      "loss": 0.6561,
      "step": 106
    },
    {
      "epoch": 0.43341772151898733,
      "grad_norm": 1.560236294789284,
      "learning_rate": 3.479674796747968e-05,
      "loss": 0.6756,
      "step": 107
    },
    {
      "epoch": 0.43746835443037974,
      "grad_norm": 1.5809286261119126,
      "learning_rate": 3.51219512195122e-05,
      "loss": 0.659,
      "step": 108
    },
    {
      "epoch": 0.44151898734177214,
      "grad_norm": 1.123593624417136,
      "learning_rate": 3.544715447154472e-05,
      "loss": 0.6579,
      "step": 109
    },
    {
      "epoch": 0.44556962025316454,
      "grad_norm": 1.4558926194831685,
      "learning_rate": 3.577235772357724e-05,
      "loss": 0.657,
      "step": 110
    },
    {
      "epoch": 0.44962025316455695,
      "grad_norm": 1.0393978988072028,
      "learning_rate": 3.609756097560976e-05,
      "loss": 0.6488,
      "step": 111
    },
    {
      "epoch": 0.45367088607594935,
      "grad_norm": 1.6976376758116054,
      "learning_rate": 3.642276422764228e-05,
      "loss": 0.6536,
      "step": 112
    },
    {
      "epoch": 0.45772151898734176,
      "grad_norm": 1.3834562832314612,
      "learning_rate": 3.67479674796748e-05,
      "loss": 0.6595,
      "step": 113
    },
    {
      "epoch": 0.46177215189873416,
      "grad_norm": 1.4295728740809213,
      "learning_rate": 3.7073170731707325e-05,
      "loss": 0.6503,
      "step": 114
    },
    {
      "epoch": 0.46582278481012657,
      "grad_norm": 1.1814226080747972,
      "learning_rate": 3.739837398373984e-05,
      "loss": 0.6497,
      "step": 115
    },
    {
      "epoch": 0.46987341772151897,
      "grad_norm": 1.536926787138022,
      "learning_rate": 3.772357723577236e-05,
      "loss": 0.6725,
      "step": 116
    },
    {
      "epoch": 0.4739240506329114,
      "grad_norm": 0.9426335602063272,
      "learning_rate": 3.804878048780488e-05,
      "loss": 0.6516,
      "step": 117
    },
    {
      "epoch": 0.4779746835443038,
      "grad_norm": 1.1985616214745298,
      "learning_rate": 3.83739837398374e-05,
      "loss": 0.6548,
      "step": 118
    },
    {
      "epoch": 0.4820253164556962,
      "grad_norm": 1.079783774562937,
      "learning_rate": 3.869918699186992e-05,
      "loss": 0.6537,
      "step": 119
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 1.028334101766473,
      "learning_rate": 3.9024390243902444e-05,
      "loss": 0.6359,
      "step": 120
    },
    {
      "epoch": 0.490126582278481,
      "grad_norm": 1.3559404729784237,
      "learning_rate": 3.9349593495934966e-05,
      "loss": 0.6542,
      "step": 121
    },
    {
      "epoch": 0.4941772151898734,
      "grad_norm": 0.9823084891166122,
      "learning_rate": 3.967479674796748e-05,
      "loss": 0.6486,
      "step": 122
    },
    {
      "epoch": 0.4982278481012658,
      "grad_norm": 1.5356930470801822,
      "learning_rate": 4e-05,
      "loss": 0.6524,
      "step": 123
    },
    {
      "epoch": 0.5022784810126583,
      "grad_norm": 1.2078727809503662,
      "learning_rate": 3.999991946137476e-05,
      "loss": 0.6456,
      "step": 124
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 1.3281820029781153,
      "learning_rate": 3.999967784614766e-05,
      "loss": 0.6462,
      "step": 125
    },
    {
      "epoch": 0.5103797468354431,
      "grad_norm": 1.0924515901018872,
      "learning_rate": 3.9999275156264656e-05,
      "loss": 0.6365,
      "step": 126
    },
    {
      "epoch": 0.5144303797468355,
      "grad_norm": 1.4511733061268617,
      "learning_rate": 3.999871139496895e-05,
      "loss": 0.6473,
      "step": 127
    },
    {
      "epoch": 0.5184810126582279,
      "grad_norm": 0.983834322086371,
      "learning_rate": 3.9997986566800995e-05,
      "loss": 0.6498,
      "step": 128
    },
    {
      "epoch": 0.5225316455696203,
      "grad_norm": 1.0478739898375304,
      "learning_rate": 3.999710067759846e-05,
      "loss": 0.6443,
      "step": 129
    },
    {
      "epoch": 0.5265822784810127,
      "grad_norm": 0.9148346245105307,
      "learning_rate": 3.999605373449617e-05,
      "loss": 0.6422,
      "step": 130
    },
    {
      "epoch": 0.5306329113924051,
      "grad_norm": 0.884961601906783,
      "learning_rate": 3.9994845745926075e-05,
      "loss": 0.6429,
      "step": 131
    },
    {
      "epoch": 0.5346835443037975,
      "grad_norm": 1.1056992538002086,
      "learning_rate": 3.999347672161713e-05,
      "loss": 0.6473,
      "step": 132
    },
    {
      "epoch": 0.5387341772151899,
      "grad_norm": 1.7492897269601604,
      "learning_rate": 3.999194667259528e-05,
      "loss": 0.6492,
      "step": 133
    },
    {
      "epoch": 0.5427848101265823,
      "grad_norm": 1.0344914348244,
      "learning_rate": 3.999025561118334e-05,
      "loss": 0.6664,
      "step": 134
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 2.099654893858135,
      "learning_rate": 3.998840355100086e-05,
      "loss": 0.6684,
      "step": 135
    },
    {
      "epoch": 0.5508860759493671,
      "grad_norm": 1.4078670098069543,
      "learning_rate": 3.998639050696409e-05,
      "loss": 0.6487,
      "step": 136
    },
    {
      "epoch": 0.5549367088607595,
      "grad_norm": 2.114098984285829,
      "learning_rate": 3.998421649528582e-05,
      "loss": 0.6461,
      "step": 137
    },
    {
      "epoch": 0.5589873417721519,
      "grad_norm": 1.5520858273645852,
      "learning_rate": 3.9981881533475234e-05,
      "loss": 0.6608,
      "step": 138
    },
    {
      "epoch": 0.5630379746835443,
      "grad_norm": 1.9458622266732575,
      "learning_rate": 3.997938564033779e-05,
      "loss": 0.6692,
      "step": 139
    },
    {
      "epoch": 0.5670886075949367,
      "grad_norm": 1.681021751813206,
      "learning_rate": 3.9976728835975064e-05,
      "loss": 0.6437,
      "step": 140
    },
    {
      "epoch": 0.5711392405063291,
      "grad_norm": 1.6056234864966519,
      "learning_rate": 3.9973911141784605e-05,
      "loss": 0.6471,
      "step": 141
    },
    {
      "epoch": 0.5751898734177215,
      "grad_norm": 1.5560674387671527,
      "learning_rate": 3.997093258045973e-05,
      "loss": 0.6574,
      "step": 142
    },
    {
      "epoch": 0.579240506329114,
      "grad_norm": 1.1402626698710554,
      "learning_rate": 3.996779317598936e-05,
      "loss": 0.6344,
      "step": 143
    },
    {
      "epoch": 0.5832911392405064,
      "grad_norm": 1.3295627998677513,
      "learning_rate": 3.996449295365782e-05,
      "loss": 0.6492,
      "step": 144
    },
    {
      "epoch": 0.5873417721518988,
      "grad_norm": 1.0860941043660022,
      "learning_rate": 3.996103194004467e-05,
      "loss": 0.6428,
      "step": 145
    },
    {
      "epoch": 0.5913924050632912,
      "grad_norm": 1.3307516200575613,
      "learning_rate": 3.995741016302441e-05,
      "loss": 0.6478,
      "step": 146
    },
    {
      "epoch": 0.5954430379746836,
      "grad_norm": 1.0881273230169572,
      "learning_rate": 3.9953627651766364e-05,
      "loss": 0.656,
      "step": 147
    },
    {
      "epoch": 0.599493670886076,
      "grad_norm": 0.9713834409449049,
      "learning_rate": 3.9949684436734325e-05,
      "loss": 0.6386,
      "step": 148
    },
    {
      "epoch": 0.6035443037974684,
      "grad_norm": 1.1526825064464896,
      "learning_rate": 3.994558054968643e-05,
      "loss": 0.6374,
      "step": 149
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.870181241816471,
      "learning_rate": 3.994131602367481e-05,
      "loss": 0.6418,
      "step": 150
    },
    {
      "epoch": 0.6116455696202532,
      "grad_norm": 1.1060315453225653,
      "learning_rate": 3.9936890893045376e-05,
      "loss": 0.6184,
      "step": 151
    },
    {
      "epoch": 0.6156962025316456,
      "grad_norm": 0.7952240419957377,
      "learning_rate": 3.993230519343752e-05,
      "loss": 0.6364,
      "step": 152
    },
    {
      "epoch": 0.619746835443038,
      "grad_norm": 0.9540127354014867,
      "learning_rate": 3.992755896178383e-05,
      "loss": 0.6317,
      "step": 153
    },
    {
      "epoch": 0.6237974683544304,
      "grad_norm": 0.7891401506524477,
      "learning_rate": 3.992265223630981e-05,
      "loss": 0.6437,
      "step": 154
    },
    {
      "epoch": 0.6278481012658228,
      "grad_norm": 0.942911040659031,
      "learning_rate": 3.991758505653355e-05,
      "loss": 0.6447,
      "step": 155
    },
    {
      "epoch": 0.6318987341772152,
      "grad_norm": 0.7558434465628108,
      "learning_rate": 3.991235746326543e-05,
      "loss": 0.6313,
      "step": 156
    },
    {
      "epoch": 0.6359493670886076,
      "grad_norm": 0.7795684086309097,
      "learning_rate": 3.9906969498607745e-05,
      "loss": 0.6469,
      "step": 157
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.7448666075821628,
      "learning_rate": 3.990142120595444e-05,
      "loss": 0.6334,
      "step": 158
    },
    {
      "epoch": 0.6440506329113924,
      "grad_norm": 0.7239485624173987,
      "learning_rate": 3.98957126299907e-05,
      "loss": 0.6248,
      "step": 159
    },
    {
      "epoch": 0.6481012658227848,
      "grad_norm": 0.591155817523153,
      "learning_rate": 3.9889843816692596e-05,
      "loss": 0.6406,
      "step": 160
    },
    {
      "epoch": 0.6521518987341772,
      "grad_norm": 0.607708955695484,
      "learning_rate": 3.9883814813326766e-05,
      "loss": 0.6202,
      "step": 161
    },
    {
      "epoch": 0.6562025316455696,
      "grad_norm": 0.6156078218694333,
      "learning_rate": 3.9877625668449956e-05,
      "loss": 0.6357,
      "step": 162
    },
    {
      "epoch": 0.660253164556962,
      "grad_norm": 0.5604971329767237,
      "learning_rate": 3.98712764319087e-05,
      "loss": 0.6463,
      "step": 163
    },
    {
      "epoch": 0.6643037974683544,
      "grad_norm": 0.6533794739116997,
      "learning_rate": 3.9864767154838864e-05,
      "loss": 0.63,
      "step": 164
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 0.5417533767740549,
      "learning_rate": 3.9858097889665277e-05,
      "loss": 0.6418,
      "step": 165
    },
    {
      "epoch": 0.6724050632911392,
      "grad_norm": 0.7329668151245794,
      "learning_rate": 3.985126869010129e-05,
      "loss": 0.62,
      "step": 166
    },
    {
      "epoch": 0.6764556962025317,
      "grad_norm": 0.5324817345898383,
      "learning_rate": 3.984427961114833e-05,
      "loss": 0.6251,
      "step": 167
    },
    {
      "epoch": 0.6805063291139241,
      "grad_norm": 0.7285590655077412,
      "learning_rate": 3.9837130709095475e-05,
      "loss": 0.615,
      "step": 168
    },
    {
      "epoch": 0.6845569620253165,
      "grad_norm": 0.5994718506820276,
      "learning_rate": 3.982982204151901e-05,
      "loss": 0.6234,
      "step": 169
    },
    {
      "epoch": 0.6886075949367089,
      "grad_norm": 0.603072195674644,
      "learning_rate": 3.982235366728193e-05,
      "loss": 0.6295,
      "step": 170
    },
    {
      "epoch": 0.6926582278481013,
      "grad_norm": 0.6589057168853385,
      "learning_rate": 3.9814725646533505e-05,
      "loss": 0.6179,
      "step": 171
    },
    {
      "epoch": 0.6967088607594937,
      "grad_norm": 0.7004049506938446,
      "learning_rate": 3.9806938040708746e-05,
      "loss": 0.6211,
      "step": 172
    },
    {
      "epoch": 0.7007594936708861,
      "grad_norm": 0.6229989531851842,
      "learning_rate": 3.9798990912527976e-05,
      "loss": 0.6237,
      "step": 173
    },
    {
      "epoch": 0.7048101265822785,
      "grad_norm": 0.5080658785134241,
      "learning_rate": 3.979088432599627e-05,
      "loss": 0.6149,
      "step": 174
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.7714215557946642,
      "learning_rate": 3.9782618346402964e-05,
      "loss": 0.6326,
      "step": 175
    },
    {
      "epoch": 0.7129113924050633,
      "grad_norm": 0.8182796417435333,
      "learning_rate": 3.977419304032111e-05,
      "loss": 0.634,
      "step": 176
    },
    {
      "epoch": 0.7169620253164557,
      "grad_norm": 0.7537452717458883,
      "learning_rate": 3.976560847560697e-05,
      "loss": 0.6124,
      "step": 177
    },
    {
      "epoch": 0.7210126582278481,
      "grad_norm": 0.729089731848073,
      "learning_rate": 3.9756864721399456e-05,
      "loss": 0.6343,
      "step": 178
    },
    {
      "epoch": 0.7250632911392405,
      "grad_norm": 0.5027855711233715,
      "learning_rate": 3.974796184811956e-05,
      "loss": 0.6168,
      "step": 179
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 0.478461909753587,
      "learning_rate": 3.973889992746979e-05,
      "loss": 0.6125,
      "step": 180
    },
    {
      "epoch": 0.7331645569620253,
      "grad_norm": 0.6519906445020534,
      "learning_rate": 3.972967903243361e-05,
      "loss": 0.6291,
      "step": 181
    },
    {
      "epoch": 0.7372151898734177,
      "grad_norm": 0.7509662839843748,
      "learning_rate": 3.972029923727486e-05,
      "loss": 0.6441,
      "step": 182
    },
    {
      "epoch": 0.7412658227848101,
      "grad_norm": 0.6894343086025367,
      "learning_rate": 3.971076061753709e-05,
      "loss": 0.6269,
      "step": 183
    },
    {
      "epoch": 0.7453164556962025,
      "grad_norm": 0.6940826007520512,
      "learning_rate": 3.9701063250043066e-05,
      "loss": 0.6297,
      "step": 184
    },
    {
      "epoch": 0.7493670886075949,
      "grad_norm": 0.6483458725833058,
      "learning_rate": 3.969120721289402e-05,
      "loss": 0.6213,
      "step": 185
    },
    {
      "epoch": 0.7534177215189873,
      "grad_norm": 0.8571625244918416,
      "learning_rate": 3.9681192585469146e-05,
      "loss": 0.6169,
      "step": 186
    },
    {
      "epoch": 0.7574683544303797,
      "grad_norm": 0.9218713814556457,
      "learning_rate": 3.9671019448424865e-05,
      "loss": 0.6223,
      "step": 187
    },
    {
      "epoch": 0.7615189873417721,
      "grad_norm": 0.7785728531131607,
      "learning_rate": 3.966068788369422e-05,
      "loss": 0.6207,
      "step": 188
    },
    {
      "epoch": 0.7655696202531646,
      "grad_norm": 0.5684131322274167,
      "learning_rate": 3.965019797448622e-05,
      "loss": 0.6206,
      "step": 189
    },
    {
      "epoch": 0.769620253164557,
      "grad_norm": 0.7018323172643347,
      "learning_rate": 3.963954980528515e-05,
      "loss": 0.6355,
      "step": 190
    },
    {
      "epoch": 0.7736708860759494,
      "grad_norm": 0.8494235094588225,
      "learning_rate": 3.9628743461849905e-05,
      "loss": 0.6229,
      "step": 191
    },
    {
      "epoch": 0.7777215189873418,
      "grad_norm": 0.7050833386965384,
      "learning_rate": 3.961777903121329e-05,
      "loss": 0.6154,
      "step": 192
    },
    {
      "epoch": 0.7817721518987342,
      "grad_norm": 0.5506331638243598,
      "learning_rate": 3.960665660168131e-05,
      "loss": 0.6323,
      "step": 193
    },
    {
      "epoch": 0.7858227848101266,
      "grad_norm": 0.6328044747598569,
      "learning_rate": 3.9595376262832485e-05,
      "loss": 0.6329,
      "step": 194
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 0.7846775205413099,
      "learning_rate": 3.9583938105517127e-05,
      "loss": 0.6164,
      "step": 195
    },
    {
      "epoch": 0.7939240506329114,
      "grad_norm": 0.5873785974553063,
      "learning_rate": 3.957234222185657e-05,
      "loss": 0.6128,
      "step": 196
    },
    {
      "epoch": 0.7979746835443038,
      "grad_norm": 0.4306322267214671,
      "learning_rate": 3.9560588705242474e-05,
      "loss": 0.6228,
      "step": 197
    },
    {
      "epoch": 0.8020253164556962,
      "grad_norm": 0.621403740157461,
      "learning_rate": 3.954867765033605e-05,
      "loss": 0.6173,
      "step": 198
    },
    {
      "epoch": 0.8060759493670886,
      "grad_norm": 0.6392451258526016,
      "learning_rate": 3.953660915306728e-05,
      "loss": 0.6148,
      "step": 199
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.5258670200002085,
      "learning_rate": 3.952438331063419e-05,
      "loss": 0.6249,
      "step": 200
    },
    {
      "epoch": 0.8141772151898734,
      "grad_norm": 0.4859036412572938,
      "learning_rate": 3.951200022150205e-05,
      "loss": 0.6142,
      "step": 201
    },
    {
      "epoch": 0.8182278481012658,
      "grad_norm": 0.5731696479410778,
      "learning_rate": 3.949945998540253e-05,
      "loss": 0.623,
      "step": 202
    },
    {
      "epoch": 0.8222784810126582,
      "grad_norm": 0.5563395478338331,
      "learning_rate": 3.9486762703332993e-05,
      "loss": 0.6197,
      "step": 203
    },
    {
      "epoch": 0.8263291139240506,
      "grad_norm": 0.6398044235769332,
      "learning_rate": 3.947390847755559e-05,
      "loss": 0.6239,
      "step": 204
    },
    {
      "epoch": 0.830379746835443,
      "grad_norm": 0.5760457880690834,
      "learning_rate": 3.946089741159648e-05,
      "loss": 0.6179,
      "step": 205
    },
    {
      "epoch": 0.8344303797468354,
      "grad_norm": 0.47228702613295065,
      "learning_rate": 3.944772961024501e-05,
      "loss": 0.6284,
      "step": 206
    },
    {
      "epoch": 0.8384810126582278,
      "grad_norm": 0.4819098603372228,
      "learning_rate": 3.943440517955285e-05,
      "loss": 0.6153,
      "step": 207
    },
    {
      "epoch": 0.8425316455696202,
      "grad_norm": 0.43345086320940657,
      "learning_rate": 3.9420924226833126e-05,
      "loss": 0.6091,
      "step": 208
    },
    {
      "epoch": 0.8465822784810126,
      "grad_norm": 0.5440822105813555,
      "learning_rate": 3.9407286860659566e-05,
      "loss": 0.6139,
      "step": 209
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 0.4780965303118637,
      "learning_rate": 3.9393493190865657e-05,
      "loss": 0.6148,
      "step": 210
    },
    {
      "epoch": 0.8546835443037974,
      "grad_norm": 0.4118132628354046,
      "learning_rate": 3.937954332854371e-05,
      "loss": 0.6011,
      "step": 211
    },
    {
      "epoch": 0.8587341772151899,
      "grad_norm": 0.47941929279112827,
      "learning_rate": 3.9365437386044016e-05,
      "loss": 0.6168,
      "step": 212
    },
    {
      "epoch": 0.8627848101265823,
      "grad_norm": 0.36983057672873043,
      "learning_rate": 3.935117547697387e-05,
      "loss": 0.6052,
      "step": 213
    },
    {
      "epoch": 0.8668354430379747,
      "grad_norm": 0.6161175215621785,
      "learning_rate": 3.933675771619675e-05,
      "loss": 0.6125,
      "step": 214
    },
    {
      "epoch": 0.8708860759493671,
      "grad_norm": 0.4567894305935037,
      "learning_rate": 3.932218421983131e-05,
      "loss": 0.6162,
      "step": 215
    },
    {
      "epoch": 0.8749367088607595,
      "grad_norm": 0.5135328049547743,
      "learning_rate": 3.9307455105250484e-05,
      "loss": 0.6194,
      "step": 216
    },
    {
      "epoch": 0.8789873417721519,
      "grad_norm": 0.5036802589693263,
      "learning_rate": 3.929257049108054e-05,
      "loss": 0.6304,
      "step": 217
    },
    {
      "epoch": 0.8830379746835443,
      "grad_norm": 0.5137560791338927,
      "learning_rate": 3.927753049720011e-05,
      "loss": 0.61,
      "step": 218
    },
    {
      "epoch": 0.8870886075949367,
      "grad_norm": 0.609964000538998,
      "learning_rate": 3.9262335244739234e-05,
      "loss": 0.6126,
      "step": 219
    },
    {
      "epoch": 0.8911392405063291,
      "grad_norm": 0.4083823458324012,
      "learning_rate": 3.92469848560784e-05,
      "loss": 0.613,
      "step": 220
    },
    {
      "epoch": 0.8951898734177215,
      "grad_norm": 0.4584784997524374,
      "learning_rate": 3.923147945484751e-05,
      "loss": 0.594,
      "step": 221
    },
    {
      "epoch": 0.8992405063291139,
      "grad_norm": 0.430173210479083,
      "learning_rate": 3.9215819165924956e-05,
      "loss": 0.6251,
      "step": 222
    },
    {
      "epoch": 0.9032911392405063,
      "grad_norm": 0.4294785919890423,
      "learning_rate": 3.920000411543654e-05,
      "loss": 0.6197,
      "step": 223
    },
    {
      "epoch": 0.9073417721518987,
      "grad_norm": 0.5263246021566597,
      "learning_rate": 3.9184034430754495e-05,
      "loss": 0.6243,
      "step": 224
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.43691586737376925,
      "learning_rate": 3.916791024049648e-05,
      "loss": 0.6212,
      "step": 225
    },
    {
      "epoch": 0.9154430379746835,
      "grad_norm": 0.43117615132383846,
      "learning_rate": 3.91516316745245e-05,
      "loss": 0.6215,
      "step": 226
    },
    {
      "epoch": 0.9194936708860759,
      "grad_norm": 0.5020281978660368,
      "learning_rate": 3.913519886394389e-05,
      "loss": 0.6166,
      "step": 227
    },
    {
      "epoch": 0.9235443037974683,
      "grad_norm": 0.44463326722302626,
      "learning_rate": 3.911861194110225e-05,
      "loss": 0.6008,
      "step": 228
    },
    {
      "epoch": 0.9275949367088607,
      "grad_norm": 0.541206763409399,
      "learning_rate": 3.910187103958837e-05,
      "loss": 0.6087,
      "step": 229
    },
    {
      "epoch": 0.9316455696202531,
      "grad_norm": 0.48937543842451026,
      "learning_rate": 3.908497629423117e-05,
      "loss": 0.6072,
      "step": 230
    },
    {
      "epoch": 0.9356962025316455,
      "grad_norm": 0.3984103174409041,
      "learning_rate": 3.9067927841098614e-05,
      "loss": 0.6244,
      "step": 231
    },
    {
      "epoch": 0.9397468354430379,
      "grad_norm": 0.6246548628267016,
      "learning_rate": 3.9050725817496594e-05,
      "loss": 0.6059,
      "step": 232
    },
    {
      "epoch": 0.9437974683544303,
      "grad_norm": 0.4951756160965602,
      "learning_rate": 3.9033370361967844e-05,
      "loss": 0.6161,
      "step": 233
    },
    {
      "epoch": 0.9478481012658228,
      "grad_norm": 0.4798622926501674,
      "learning_rate": 3.901586161429081e-05,
      "loss": 0.6108,
      "step": 234
    },
    {
      "epoch": 0.9518987341772152,
      "grad_norm": 0.5282429391645367,
      "learning_rate": 3.8998199715478545e-05,
      "loss": 0.613,
      "step": 235
    },
    {
      "epoch": 0.9559493670886076,
      "grad_norm": 0.4885947430565363,
      "learning_rate": 3.8980384807777564e-05,
      "loss": 0.6269,
      "step": 236
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.45238536597414436,
      "learning_rate": 3.896241703466667e-05,
      "loss": 0.616,
      "step": 237
    },
    {
      "epoch": 0.9640506329113924,
      "grad_norm": 0.39313656198512664,
      "learning_rate": 3.894429654085585e-05,
      "loss": 0.6092,
      "step": 238
    },
    {
      "epoch": 0.9681012658227848,
      "grad_norm": 0.40004888515504505,
      "learning_rate": 3.892602347228505e-05,
      "loss": 0.6185,
      "step": 239
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 0.5189156799720501,
      "learning_rate": 3.890759797612307e-05,
      "loss": 0.6128,
      "step": 240
    },
    {
      "epoch": 0.9762025316455696,
      "grad_norm": 0.3756698390039139,
      "learning_rate": 3.888902020076632e-05,
      "loss": 0.6308,
      "step": 241
    },
    {
      "epoch": 0.980253164556962,
      "grad_norm": 0.5956276927917231,
      "learning_rate": 3.887029029583764e-05,
      "loss": 0.6002,
      "step": 242
    },
    {
      "epoch": 0.9843037974683544,
      "grad_norm": 0.4103752806299911,
      "learning_rate": 3.8851408412185125e-05,
      "loss": 0.6067,
      "step": 243
    },
    {
      "epoch": 0.9883544303797468,
      "grad_norm": 0.650797224572476,
      "learning_rate": 3.8832374701880855e-05,
      "loss": 0.613,
      "step": 244
    },
    {
      "epoch": 0.9924050632911392,
      "grad_norm": 0.769375813051049,
      "learning_rate": 3.881318931821972e-05,
      "loss": 0.6199,
      "step": 245
    },
    {
      "epoch": 0.9964556962025316,
      "grad_norm": 0.4865980388014915,
      "learning_rate": 3.879385241571817e-05,
      "loss": 0.6045,
      "step": 246
    },
    {
      "epoch": 1.000506329113924,
      "grad_norm": 0.6016551209494891,
      "learning_rate": 3.8774364150112955e-05,
      "loss": 0.5991,
      "step": 247
    },
    {
      "epoch": 1.0045569620253165,
      "grad_norm": 0.6762739783915436,
      "learning_rate": 3.8754724678359884e-05,
      "loss": 0.5742,
      "step": 248
    },
    {
      "epoch": 1.0086075949367088,
      "grad_norm": 0.5863736625895714,
      "learning_rate": 3.873493415863256e-05,
      "loss": 0.5716,
      "step": 249
    },
    {
      "epoch": 1.0126582278481013,
      "grad_norm": 0.6148252052279031,
      "learning_rate": 3.871499275032111e-05,
      "loss": 0.5586,
      "step": 250
    },
    {
      "epoch": 1.0167088607594936,
      "grad_norm": 0.47516675865753616,
      "learning_rate": 3.869490061403091e-05,
      "loss": 0.5564,
      "step": 251
    },
    {
      "epoch": 1.0207594936708861,
      "grad_norm": 0.6290497022486202,
      "learning_rate": 3.867465791158124e-05,
      "loss": 0.5582,
      "step": 252
    },
    {
      "epoch": 1.0248101265822784,
      "grad_norm": 0.46998886795568384,
      "learning_rate": 3.865426480600407e-05,
      "loss": 0.5626,
      "step": 253
    },
    {
      "epoch": 1.028860759493671,
      "grad_norm": 0.4214985275717433,
      "learning_rate": 3.863372146154264e-05,
      "loss": 0.5831,
      "step": 254
    },
    {
      "epoch": 1.0329113924050632,
      "grad_norm": 0.6229667172269647,
      "learning_rate": 3.861302804365024e-05,
      "loss": 0.5613,
      "step": 255
    },
    {
      "epoch": 1.0369620253164558,
      "grad_norm": 0.7145326946685512,
      "learning_rate": 3.85921847189888e-05,
      "loss": 0.5781,
      "step": 256
    },
    {
      "epoch": 1.041012658227848,
      "grad_norm": 0.49877846974194767,
      "learning_rate": 3.85711916554276e-05,
      "loss": 0.5627,
      "step": 257
    },
    {
      "epoch": 1.0450632911392406,
      "grad_norm": 0.5586122780653295,
      "learning_rate": 3.85500490220419e-05,
      "loss": 0.5554,
      "step": 258
    },
    {
      "epoch": 1.0491139240506329,
      "grad_norm": 0.5683943531475496,
      "learning_rate": 3.852875698911154e-05,
      "loss": 0.569,
      "step": 259
    },
    {
      "epoch": 1.0531645569620254,
      "grad_norm": 0.386096752292213,
      "learning_rate": 3.850731572811963e-05,
      "loss": 0.5576,
      "step": 260
    },
    {
      "epoch": 1.0572151898734177,
      "grad_norm": 0.3937830104443795,
      "learning_rate": 3.848572541175116e-05,
      "loss": 0.5617,
      "step": 261
    },
    {
      "epoch": 1.0612658227848102,
      "grad_norm": 0.5055630358055839,
      "learning_rate": 3.846398621389154e-05,
      "loss": 0.5657,
      "step": 262
    },
    {
      "epoch": 1.0653164556962025,
      "grad_norm": 0.5038621034888336,
      "learning_rate": 3.84420983096253e-05,
      "loss": 0.5577,
      "step": 263
    },
    {
      "epoch": 1.069367088607595,
      "grad_norm": 0.42363741759197754,
      "learning_rate": 3.8420061875234606e-05,
      "loss": 0.5629,
      "step": 264
    },
    {
      "epoch": 1.0734177215189873,
      "grad_norm": 0.5772346836554705,
      "learning_rate": 3.839787708819787e-05,
      "loss": 0.5682,
      "step": 265
    },
    {
      "epoch": 1.0774683544303798,
      "grad_norm": 0.6755048896943167,
      "learning_rate": 3.8375544127188325e-05,
      "loss": 0.5649,
      "step": 266
    },
    {
      "epoch": 1.081518987341772,
      "grad_norm": 0.5733935239916316,
      "learning_rate": 3.8353063172072564e-05,
      "loss": 0.5684,
      "step": 267
    },
    {
      "epoch": 1.0855696202531646,
      "grad_norm": 0.4314340838032296,
      "learning_rate": 3.8330434403909105e-05,
      "loss": 0.5481,
      "step": 268
    },
    {
      "epoch": 1.089620253164557,
      "grad_norm": 0.4961703285912229,
      "learning_rate": 3.8307658004946934e-05,
      "loss": 0.5695,
      "step": 269
    },
    {
      "epoch": 1.0936708860759494,
      "grad_norm": 0.5401640800850352,
      "learning_rate": 3.8284734158624046e-05,
      "loss": 0.5668,
      "step": 270
    },
    {
      "epoch": 1.0977215189873417,
      "grad_norm": 0.49988381319092134,
      "learning_rate": 3.826166304956594e-05,
      "loss": 0.5533,
      "step": 271
    },
    {
      "epoch": 1.1017721518987342,
      "grad_norm": 0.5357146211019791,
      "learning_rate": 3.8238444863584164e-05,
      "loss": 0.5539,
      "step": 272
    },
    {
      "epoch": 1.1058227848101265,
      "grad_norm": 0.4957272632921066,
      "learning_rate": 3.821507978767479e-05,
      "loss": 0.5638,
      "step": 273
    },
    {
      "epoch": 1.109873417721519,
      "grad_norm": 0.5209552525390151,
      "learning_rate": 3.819156801001693e-05,
      "loss": 0.567,
      "step": 274
    },
    {
      "epoch": 1.1139240506329113,
      "grad_norm": 0.43524693112384316,
      "learning_rate": 3.816790971997121e-05,
      "loss": 0.5607,
      "step": 275
    },
    {
      "epoch": 1.1179746835443038,
      "grad_norm": 0.6777140536339052,
      "learning_rate": 3.8144105108078246e-05,
      "loss": 0.5638,
      "step": 276
    },
    {
      "epoch": 1.1220253164556961,
      "grad_norm": 0.5783928516930995,
      "learning_rate": 3.81201543660571e-05,
      "loss": 0.554,
      "step": 277
    },
    {
      "epoch": 1.1260759493670887,
      "grad_norm": 0.603612979218957,
      "learning_rate": 3.809605768680377e-05,
      "loss": 0.5673,
      "step": 278
    },
    {
      "epoch": 1.130126582278481,
      "grad_norm": 0.5740200811030863,
      "learning_rate": 3.807181526438958e-05,
      "loss": 0.5652,
      "step": 279
    },
    {
      "epoch": 1.1341772151898735,
      "grad_norm": 0.5374091625854728,
      "learning_rate": 3.8047427294059697e-05,
      "loss": 0.5654,
      "step": 280
    },
    {
      "epoch": 1.1382278481012658,
      "grad_norm": 0.5827311647458365,
      "learning_rate": 3.802289397223145e-05,
      "loss": 0.5711,
      "step": 281
    },
    {
      "epoch": 1.1422784810126583,
      "grad_norm": 0.5259696586839012,
      "learning_rate": 3.7998215496492854e-05,
      "loss": 0.5705,
      "step": 282
    },
    {
      "epoch": 1.1463291139240506,
      "grad_norm": 0.4907255685168625,
      "learning_rate": 3.797339206560096e-05,
      "loss": 0.5623,
      "step": 283
    },
    {
      "epoch": 1.150379746835443,
      "grad_norm": 0.4673939863398156,
      "learning_rate": 3.794842387948027e-05,
      "loss": 0.5528,
      "step": 284
    },
    {
      "epoch": 1.1544303797468354,
      "grad_norm": 0.4869927177226448,
      "learning_rate": 3.7923311139221114e-05,
      "loss": 0.568,
      "step": 285
    },
    {
      "epoch": 1.158481012658228,
      "grad_norm": 0.4772030412442051,
      "learning_rate": 3.7898054047078054e-05,
      "loss": 0.5529,
      "step": 286
    },
    {
      "epoch": 1.1625316455696202,
      "grad_norm": 0.41354705528597935,
      "learning_rate": 3.787265280646825e-05,
      "loss": 0.5655,
      "step": 287
    },
    {
      "epoch": 1.1665822784810127,
      "grad_norm": 0.4231512718054825,
      "learning_rate": 3.7847107621969786e-05,
      "loss": 0.5571,
      "step": 288
    },
    {
      "epoch": 1.170632911392405,
      "grad_norm": 0.44463419802142096,
      "learning_rate": 3.7821418699320064e-05,
      "loss": 0.559,
      "step": 289
    },
    {
      "epoch": 1.1746835443037975,
      "grad_norm": 0.5314268811760819,
      "learning_rate": 3.7795586245414145e-05,
      "loss": 0.5558,
      "step": 290
    },
    {
      "epoch": 1.1787341772151898,
      "grad_norm": 0.49566308382488355,
      "learning_rate": 3.776961046830306e-05,
      "loss": 0.5612,
      "step": 291
    },
    {
      "epoch": 1.1827848101265823,
      "grad_norm": 0.39238583204386734,
      "learning_rate": 3.774349157719215e-05,
      "loss": 0.5593,
      "step": 292
    },
    {
      "epoch": 1.1868354430379746,
      "grad_norm": 0.42620692082882106,
      "learning_rate": 3.7717229782439365e-05,
      "loss": 0.5688,
      "step": 293
    },
    {
      "epoch": 1.1908860759493671,
      "grad_norm": 0.6940594090835736,
      "learning_rate": 3.769082529555359e-05,
      "loss": 0.5653,
      "step": 294
    },
    {
      "epoch": 1.1949367088607594,
      "grad_norm": 0.6885079643144056,
      "learning_rate": 3.766427832919294e-05,
      "loss": 0.553,
      "step": 295
    },
    {
      "epoch": 1.198987341772152,
      "grad_norm": 0.679794751631822,
      "learning_rate": 3.7637589097163024e-05,
      "loss": 0.5771,
      "step": 296
    },
    {
      "epoch": 1.2030379746835442,
      "grad_norm": 0.7076196679797153,
      "learning_rate": 3.761075781441526e-05,
      "loss": 0.5672,
      "step": 297
    },
    {
      "epoch": 1.2070886075949367,
      "grad_norm": 0.4798044569766114,
      "learning_rate": 3.75837846970451e-05,
      "loss": 0.5534,
      "step": 298
    },
    {
      "epoch": 1.211139240506329,
      "grad_norm": 0.5119432958734289,
      "learning_rate": 3.755666996229032e-05,
      "loss": 0.5696,
      "step": 299
    },
    {
      "epoch": 1.2151898734177216,
      "grad_norm": 0.6019454537380619,
      "learning_rate": 3.752941382852927e-05,
      "loss": 0.5645,
      "step": 300
    },
    {
      "epoch": 1.2192405063291138,
      "grad_norm": 0.6684724744803842,
      "learning_rate": 3.7502016515279115e-05,
      "loss": 0.5605,
      "step": 301
    },
    {
      "epoch": 1.2232911392405064,
      "grad_norm": 0.6860990846108473,
      "learning_rate": 3.7474478243194043e-05,
      "loss": 0.5467,
      "step": 302
    },
    {
      "epoch": 1.2273417721518987,
      "grad_norm": 0.5929264968811373,
      "learning_rate": 3.744679923406351e-05,
      "loss": 0.5629,
      "step": 303
    },
    {
      "epoch": 1.2313924050632912,
      "grad_norm": 0.42346267369721885,
      "learning_rate": 3.741897971081043e-05,
      "loss": 0.5661,
      "step": 304
    },
    {
      "epoch": 1.2354430379746835,
      "grad_norm": 0.5106134506040135,
      "learning_rate": 3.739101989748946e-05,
      "loss": 0.5754,
      "step": 305
    },
    {
      "epoch": 1.239493670886076,
      "grad_norm": 0.5580550679776318,
      "learning_rate": 3.7362920019285066e-05,
      "loss": 0.5639,
      "step": 306
    },
    {
      "epoch": 1.2435443037974683,
      "grad_norm": 0.5002032256052342,
      "learning_rate": 3.73346803025098e-05,
      "loss": 0.5574,
      "step": 307
    },
    {
      "epoch": 1.2475949367088608,
      "grad_norm": 0.4528146573608629,
      "learning_rate": 3.730630097460247e-05,
      "loss": 0.5541,
      "step": 308
    },
    {
      "epoch": 1.251645569620253,
      "grad_norm": 0.597622326396268,
      "learning_rate": 3.727778226412628e-05,
      "loss": 0.5612,
      "step": 309
    },
    {
      "epoch": 1.2556962025316456,
      "grad_norm": 0.4419183480883549,
      "learning_rate": 3.7249124400767006e-05,
      "loss": 0.5587,
      "step": 310
    },
    {
      "epoch": 1.259746835443038,
      "grad_norm": 0.4510247001313569,
      "learning_rate": 3.722032761533114e-05,
      "loss": 0.5566,
      "step": 311
    },
    {
      "epoch": 1.2637974683544304,
      "grad_norm": 0.45438823479056306,
      "learning_rate": 3.719139213974403e-05,
      "loss": 0.5549,
      "step": 312
    },
    {
      "epoch": 1.2678481012658227,
      "grad_norm": 0.43451390910393306,
      "learning_rate": 3.7162318207048006e-05,
      "loss": 0.5649,
      "step": 313
    },
    {
      "epoch": 1.2718987341772152,
      "grad_norm": 0.45383309551800477,
      "learning_rate": 3.713310605140055e-05,
      "loss": 0.5705,
      "step": 314
    },
    {
      "epoch": 1.2759493670886077,
      "grad_norm": 0.3573474422880059,
      "learning_rate": 3.710375590807233e-05,
      "loss": 0.5658,
      "step": 315
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.47101510476430053,
      "learning_rate": 3.7074268013445365e-05,
      "loss": 0.5576,
      "step": 316
    },
    {
      "epoch": 1.2840506329113923,
      "grad_norm": 0.3989854299594064,
      "learning_rate": 3.7044642605011114e-05,
      "loss": 0.5543,
      "step": 317
    },
    {
      "epoch": 1.2881012658227848,
      "grad_norm": 0.4770750948019606,
      "learning_rate": 3.701487992136854e-05,
      "loss": 0.5675,
      "step": 318
    },
    {
      "epoch": 1.2921518987341774,
      "grad_norm": 0.5404013516265972,
      "learning_rate": 3.69849802022222e-05,
      "loss": 0.571,
      "step": 319
    },
    {
      "epoch": 1.2962025316455696,
      "grad_norm": 0.4572720913218892,
      "learning_rate": 3.6954943688380334e-05,
      "loss": 0.5461,
      "step": 320
    },
    {
      "epoch": 1.300253164556962,
      "grad_norm": 0.45000001265573575,
      "learning_rate": 3.692477062175289e-05,
      "loss": 0.5527,
      "step": 321
    },
    {
      "epoch": 1.3043037974683545,
      "grad_norm": 0.36452787051932667,
      "learning_rate": 3.689446124534958e-05,
      "loss": 0.5554,
      "step": 322
    },
    {
      "epoch": 1.308354430379747,
      "grad_norm": 0.4133301099569378,
      "learning_rate": 3.686401580327799e-05,
      "loss": 0.5643,
      "step": 323
    },
    {
      "epoch": 1.3124050632911393,
      "grad_norm": 0.45361344710792584,
      "learning_rate": 3.683343454074149e-05,
      "loss": 0.5647,
      "step": 324
    },
    {
      "epoch": 1.3164556962025316,
      "grad_norm": 0.3889810619626077,
      "learning_rate": 3.6802717704037386e-05,
      "loss": 0.5592,
      "step": 325
    },
    {
      "epoch": 1.320506329113924,
      "grad_norm": 0.4580655141182608,
      "learning_rate": 3.6771865540554855e-05,
      "loss": 0.5515,
      "step": 326
    },
    {
      "epoch": 1.3245569620253166,
      "grad_norm": 0.5034051518623537,
      "learning_rate": 3.674087829877297e-05,
      "loss": 0.5648,
      "step": 327
    },
    {
      "epoch": 1.3286075949367089,
      "grad_norm": 0.5163634775205252,
      "learning_rate": 3.6709756228258735e-05,
      "loss": 0.5493,
      "step": 328
    },
    {
      "epoch": 1.3326582278481012,
      "grad_norm": 0.3763193277537522,
      "learning_rate": 3.667849957966501e-05,
      "loss": 0.5508,
      "step": 329
    },
    {
      "epoch": 1.3367088607594937,
      "grad_norm": 0.45209201898444756,
      "learning_rate": 3.6647108604728546e-05,
      "loss": 0.5785,
      "step": 330
    },
    {
      "epoch": 1.3407594936708862,
      "grad_norm": 0.47971066944542157,
      "learning_rate": 3.661558355626795e-05,
      "loss": 0.5668,
      "step": 331
    },
    {
      "epoch": 1.3448101265822785,
      "grad_norm": 0.4875552419687559,
      "learning_rate": 3.658392468818163e-05,
      "loss": 0.5714,
      "step": 332
    },
    {
      "epoch": 1.3488607594936708,
      "grad_norm": 0.4374907884566699,
      "learning_rate": 3.655213225544574e-05,
      "loss": 0.5618,
      "step": 333
    },
    {
      "epoch": 1.3529113924050633,
      "grad_norm": 0.543043207147836,
      "learning_rate": 3.652020651411218e-05,
      "loss": 0.5564,
      "step": 334
    },
    {
      "epoch": 1.3569620253164558,
      "grad_norm": 0.45358785612240554,
      "learning_rate": 3.6488147721306474e-05,
      "loss": 0.5641,
      "step": 335
    },
    {
      "epoch": 1.3610126582278481,
      "grad_norm": 0.5163627473384448,
      "learning_rate": 3.645595613522574e-05,
      "loss": 0.5456,
      "step": 336
    },
    {
      "epoch": 1.3650632911392404,
      "grad_norm": 0.538883867876519,
      "learning_rate": 3.642363201513657e-05,
      "loss": 0.5592,
      "step": 337
    },
    {
      "epoch": 1.369113924050633,
      "grad_norm": 0.38966291493082816,
      "learning_rate": 3.6391175621373006e-05,
      "loss": 0.5569,
      "step": 338
    },
    {
      "epoch": 1.3731645569620254,
      "grad_norm": 0.4424561195938351,
      "learning_rate": 3.6358587215334355e-05,
      "loss": 0.5554,
      "step": 339
    },
    {
      "epoch": 1.3772151898734177,
      "grad_norm": 0.44292641029124924,
      "learning_rate": 3.632586705948318e-05,
      "loss": 0.5566,
      "step": 340
    },
    {
      "epoch": 1.38126582278481,
      "grad_norm": 0.42097728327879613,
      "learning_rate": 3.629301541734311e-05,
      "loss": 0.5452,
      "step": 341
    },
    {
      "epoch": 1.3853164556962025,
      "grad_norm": 0.36325809070408344,
      "learning_rate": 3.626003255349676e-05,
      "loss": 0.5607,
      "step": 342
    },
    {
      "epoch": 1.389367088607595,
      "grad_norm": 0.5230295688806101,
      "learning_rate": 3.622691873358357e-05,
      "loss": 0.5587,
      "step": 343
    },
    {
      "epoch": 1.3934177215189874,
      "grad_norm": 0.36704606903220616,
      "learning_rate": 3.61936742242977e-05,
      "loss": 0.5639,
      "step": 344
    },
    {
      "epoch": 1.3974683544303796,
      "grad_norm": 0.3881781310067377,
      "learning_rate": 3.6160299293385864e-05,
      "loss": 0.5599,
      "step": 345
    },
    {
      "epoch": 1.4015189873417722,
      "grad_norm": 0.3976681360880836,
      "learning_rate": 3.612679420964516e-05,
      "loss": 0.5683,
      "step": 346
    },
    {
      "epoch": 1.4055696202531647,
      "grad_norm": 0.4107743714390722,
      "learning_rate": 3.609315924292092e-05,
      "loss": 0.5534,
      "step": 347
    },
    {
      "epoch": 1.409620253164557,
      "grad_norm": 0.4400201890346314,
      "learning_rate": 3.6059394664104554e-05,
      "loss": 0.5541,
      "step": 348
    },
    {
      "epoch": 1.4136708860759493,
      "grad_norm": 0.4076815237052884,
      "learning_rate": 3.602550074513133e-05,
      "loss": 0.5639,
      "step": 349
    },
    {
      "epoch": 1.4177215189873418,
      "grad_norm": 0.4148904790786526,
      "learning_rate": 3.599147775897822e-05,
      "loss": 0.5541,
      "step": 350
    },
    {
      "epoch": 1.4217721518987343,
      "grad_norm": 0.3871302469461973,
      "learning_rate": 3.595732597966167e-05,
      "loss": 0.556,
      "step": 351
    },
    {
      "epoch": 1.4258227848101266,
      "grad_norm": 0.4220062781783028,
      "learning_rate": 3.592304568223542e-05,
      "loss": 0.5498,
      "step": 352
    },
    {
      "epoch": 1.4298734177215189,
      "grad_norm": 0.37026485661659675,
      "learning_rate": 3.588863714278826e-05,
      "loss": 0.5628,
      "step": 353
    },
    {
      "epoch": 1.4339240506329114,
      "grad_norm": 0.4398208528718663,
      "learning_rate": 3.585410063844186e-05,
      "loss": 0.5539,
      "step": 354
    },
    {
      "epoch": 1.437974683544304,
      "grad_norm": 0.45705723161686324,
      "learning_rate": 3.581943644734846e-05,
      "loss": 0.5636,
      "step": 355
    },
    {
      "epoch": 1.4420253164556962,
      "grad_norm": 0.41972681418342017,
      "learning_rate": 3.578464484868869e-05,
      "loss": 0.559,
      "step": 356
    },
    {
      "epoch": 1.4460759493670885,
      "grad_norm": 0.46119604263496566,
      "learning_rate": 3.5749726122669316e-05,
      "loss": 0.5577,
      "step": 357
    },
    {
      "epoch": 1.450126582278481,
      "grad_norm": 0.3802277194516473,
      "learning_rate": 3.5714680550520943e-05,
      "loss": 0.552,
      "step": 358
    },
    {
      "epoch": 1.4541772151898735,
      "grad_norm": 0.42677314916200587,
      "learning_rate": 3.5679508414495794e-05,
      "loss": 0.5552,
      "step": 359
    },
    {
      "epoch": 1.4582278481012658,
      "grad_norm": 0.4500683720020329,
      "learning_rate": 3.564420999786543e-05,
      "loss": 0.5667,
      "step": 360
    },
    {
      "epoch": 1.4622784810126581,
      "grad_norm": 0.4427368782735112,
      "learning_rate": 3.560878558491842e-05,
      "loss": 0.5526,
      "step": 361
    },
    {
      "epoch": 1.4663291139240506,
      "grad_norm": 0.4686123229098893,
      "learning_rate": 3.5573235460958145e-05,
      "loss": 0.5571,
      "step": 362
    },
    {
      "epoch": 1.4703797468354431,
      "grad_norm": 0.5389534703370309,
      "learning_rate": 3.553755991230039e-05,
      "loss": 0.551,
      "step": 363
    },
    {
      "epoch": 1.4744303797468354,
      "grad_norm": 0.42262065040834745,
      "learning_rate": 3.5501759226271144e-05,
      "loss": 0.5633,
      "step": 364
    },
    {
      "epoch": 1.4784810126582277,
      "grad_norm": 0.4291820558563475,
      "learning_rate": 3.546583369120419e-05,
      "loss": 0.5508,
      "step": 365
    },
    {
      "epoch": 1.4825316455696202,
      "grad_norm": 0.5584590431385019,
      "learning_rate": 3.5429783596438864e-05,
      "loss": 0.5547,
      "step": 366
    },
    {
      "epoch": 1.4865822784810128,
      "grad_norm": 0.4852888584056597,
      "learning_rate": 3.539360923231766e-05,
      "loss": 0.5533,
      "step": 367
    },
    {
      "epoch": 1.490632911392405,
      "grad_norm": 0.33386083419398505,
      "learning_rate": 3.535731089018394e-05,
      "loss": 0.5446,
      "step": 368
    },
    {
      "epoch": 1.4946835443037974,
      "grad_norm": 0.4362985045883858,
      "learning_rate": 3.532088886237956e-05,
      "loss": 0.552,
      "step": 369
    },
    {
      "epoch": 1.4987341772151899,
      "grad_norm": 0.44307349459706413,
      "learning_rate": 3.528434344224253e-05,
      "loss": 0.5587,
      "step": 370
    },
    {
      "epoch": 1.5027848101265824,
      "grad_norm": 0.40699057506321584,
      "learning_rate": 3.524767492410464e-05,
      "loss": 0.5581,
      "step": 371
    },
    {
      "epoch": 1.5068354430379747,
      "grad_norm": 0.3500946801068946,
      "learning_rate": 3.521088360328908e-05,
      "loss": 0.5369,
      "step": 372
    },
    {
      "epoch": 1.510886075949367,
      "grad_norm": 0.39245438605735583,
      "learning_rate": 3.517396977610811e-05,
      "loss": 0.5564,
      "step": 373
    },
    {
      "epoch": 1.5149367088607595,
      "grad_norm": 0.4936810062662944,
      "learning_rate": 3.5136933739860595e-05,
      "loss": 0.5655,
      "step": 374
    },
    {
      "epoch": 1.518987341772152,
      "grad_norm": 0.4931450693544562,
      "learning_rate": 3.509977579282971e-05,
      "loss": 0.5549,
      "step": 375
    },
    {
      "epoch": 1.5230379746835443,
      "grad_norm": 0.32128900718642733,
      "learning_rate": 3.5062496234280424e-05,
      "loss": 0.5463,
      "step": 376
    },
    {
      "epoch": 1.5270886075949366,
      "grad_norm": 0.3541596329393044,
      "learning_rate": 3.502509536445719e-05,
      "loss": 0.5466,
      "step": 377
    },
    {
      "epoch": 1.531139240506329,
      "grad_norm": 0.4197765712242005,
      "learning_rate": 3.498757348458147e-05,
      "loss": 0.5605,
      "step": 378
    },
    {
      "epoch": 1.5351898734177216,
      "grad_norm": 0.4887713658240785,
      "learning_rate": 3.4949930896849324e-05,
      "loss": 0.5574,
      "step": 379
    },
    {
      "epoch": 1.539240506329114,
      "grad_norm": 0.3744638983395964,
      "learning_rate": 3.491216790442899e-05,
      "loss": 0.5503,
      "step": 380
    },
    {
      "epoch": 1.5432911392405062,
      "grad_norm": 0.4057635713642805,
      "learning_rate": 3.487428481145839e-05,
      "loss": 0.5604,
      "step": 381
    },
    {
      "epoch": 1.5473417721518987,
      "grad_norm": 0.41577114599811055,
      "learning_rate": 3.483628192304278e-05,
      "loss": 0.5587,
      "step": 382
    },
    {
      "epoch": 1.5513924050632912,
      "grad_norm": 0.31850504209470026,
      "learning_rate": 3.479815954525219e-05,
      "loss": 0.5485,
      "step": 383
    },
    {
      "epoch": 1.5554430379746835,
      "grad_norm": 0.4154167642181542,
      "learning_rate": 3.475991798511899e-05,
      "loss": 0.559,
      "step": 384
    },
    {
      "epoch": 1.5594936708860758,
      "grad_norm": 0.356830268547951,
      "learning_rate": 3.4721557550635464e-05,
      "loss": 0.5657,
      "step": 385
    },
    {
      "epoch": 1.5635443037974683,
      "grad_norm": 0.35835683754607245,
      "learning_rate": 3.468307855075128e-05,
      "loss": 0.528,
      "step": 386
    },
    {
      "epoch": 1.5675949367088609,
      "grad_norm": 0.3459719965572832,
      "learning_rate": 3.4644481295371005e-05,
      "loss": 0.5537,
      "step": 387
    },
    {
      "epoch": 1.5716455696202531,
      "grad_norm": 0.31912788155459076,
      "learning_rate": 3.460576609535163e-05,
      "loss": 0.5612,
      "step": 388
    },
    {
      "epoch": 1.5756962025316454,
      "grad_norm": 0.37288051450424514,
      "learning_rate": 3.456693326250006e-05,
      "loss": 0.5478,
      "step": 389
    },
    {
      "epoch": 1.579746835443038,
      "grad_norm": 0.3904420272992536,
      "learning_rate": 3.452798310957058e-05,
      "loss": 0.5584,
      "step": 390
    },
    {
      "epoch": 1.5837974683544305,
      "grad_norm": 0.4334862256334808,
      "learning_rate": 3.4488915950262386e-05,
      "loss": 0.5572,
      "step": 391
    },
    {
      "epoch": 1.5878481012658228,
      "grad_norm": 0.3594630445742829,
      "learning_rate": 3.4449732099216985e-05,
      "loss": 0.5705,
      "step": 392
    },
    {
      "epoch": 1.591898734177215,
      "grad_norm": 0.38311596985783347,
      "learning_rate": 3.441043187201574e-05,
      "loss": 0.5573,
      "step": 393
    },
    {
      "epoch": 1.5959493670886076,
      "grad_norm": 0.3686530430136156,
      "learning_rate": 3.437101558517728e-05,
      "loss": 0.5586,
      "step": 394
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.4496345971957146,
      "learning_rate": 3.433148355615496e-05,
      "loss": 0.5496,
      "step": 395
    },
    {
      "epoch": 1.6040506329113924,
      "grad_norm": 0.3704216420913592,
      "learning_rate": 3.4291836103334294e-05,
      "loss": 0.5494,
      "step": 396
    },
    {
      "epoch": 1.6081012658227847,
      "grad_norm": 0.3870730103445187,
      "learning_rate": 3.425207354603043e-05,
      "loss": 0.5484,
      "step": 397
    },
    {
      "epoch": 1.6121518987341772,
      "grad_norm": 0.40983323909761443,
      "learning_rate": 3.421219620448553e-05,
      "loss": 0.5609,
      "step": 398
    },
    {
      "epoch": 1.6162025316455697,
      "grad_norm": 0.4489676597331663,
      "learning_rate": 3.417220439986623e-05,
      "loss": 0.5632,
      "step": 399
    },
    {
      "epoch": 1.620253164556962,
      "grad_norm": 0.5118096816138852,
      "learning_rate": 3.4132098454261024e-05,
      "loss": 0.5485,
      "step": 400
    },
    {
      "epoch": 1.6243037974683543,
      "grad_norm": 0.33197751824459554,
      "learning_rate": 3.4091878690677676e-05,
      "loss": 0.5546,
      "step": 401
    },
    {
      "epoch": 1.6283544303797468,
      "grad_norm": 0.43822760559801344,
      "learning_rate": 3.405154543304065e-05,
      "loss": 0.5528,
      "step": 402
    },
    {
      "epoch": 1.6324050632911393,
      "grad_norm": 0.4440612546698338,
      "learning_rate": 3.401109900618843e-05,
      "loss": 0.5543,
      "step": 403
    },
    {
      "epoch": 1.6364556962025316,
      "grad_norm": 0.48587003555557196,
      "learning_rate": 3.3970539735870996e-05,
      "loss": 0.5407,
      "step": 404
    },
    {
      "epoch": 1.640506329113924,
      "grad_norm": 0.3878017438003019,
      "learning_rate": 3.392986794874714e-05,
      "loss": 0.5619,
      "step": 405
    },
    {
      "epoch": 1.6445569620253164,
      "grad_norm": 0.4791696357375201,
      "learning_rate": 3.388908397238184e-05,
      "loss": 0.5632,
      "step": 406
    },
    {
      "epoch": 1.648607594936709,
      "grad_norm": 0.4629310198127877,
      "learning_rate": 3.384818813524362e-05,
      "loss": 0.5477,
      "step": 407
    },
    {
      "epoch": 1.6526582278481012,
      "grad_norm": 0.35921721614507574,
      "learning_rate": 3.380718076670195e-05,
      "loss": 0.5459,
      "step": 408
    },
    {
      "epoch": 1.6567088607594935,
      "grad_norm": 0.4535341790280869,
      "learning_rate": 3.376606219702454e-05,
      "loss": 0.5602,
      "step": 409
    },
    {
      "epoch": 1.660759493670886,
      "grad_norm": 0.38527180222546026,
      "learning_rate": 3.372483275737468e-05,
      "loss": 0.5511,
      "step": 410
    },
    {
      "epoch": 1.6648101265822786,
      "grad_norm": 0.38977292577527717,
      "learning_rate": 3.368349277980861e-05,
      "loss": 0.5599,
      "step": 411
    },
    {
      "epoch": 1.6688607594936709,
      "grad_norm": 0.4445553661124202,
      "learning_rate": 3.3642042597272844e-05,
      "loss": 0.5657,
      "step": 412
    },
    {
      "epoch": 1.6729113924050631,
      "grad_norm": 0.3804461592029394,
      "learning_rate": 3.360048254360144e-05,
      "loss": 0.5463,
      "step": 413
    },
    {
      "epoch": 1.6769620253164557,
      "grad_norm": 0.4324523659105559,
      "learning_rate": 3.355881295351336e-05,
      "loss": 0.5432,
      "step": 414
    },
    {
      "epoch": 1.6810126582278482,
      "grad_norm": 0.45741533700289905,
      "learning_rate": 3.351703416260975e-05,
      "loss": 0.5519,
      "step": 415
    },
    {
      "epoch": 1.6850632911392405,
      "grad_norm": 0.43028951688461764,
      "learning_rate": 3.347514650737126e-05,
      "loss": 0.5523,
      "step": 416
    },
    {
      "epoch": 1.6891139240506328,
      "grad_norm": 0.406894301683445,
      "learning_rate": 3.3433150325155295e-05,
      "loss": 0.5602,
      "step": 417
    },
    {
      "epoch": 1.6931645569620253,
      "grad_norm": 0.5107444161824117,
      "learning_rate": 3.339104595419334e-05,
      "loss": 0.558,
      "step": 418
    },
    {
      "epoch": 1.6972151898734178,
      "grad_norm": 0.4057794129668358,
      "learning_rate": 3.3348833733588204e-05,
      "loss": 0.5638,
      "step": 419
    },
    {
      "epoch": 1.70126582278481,
      "grad_norm": 0.45722807899723117,
      "learning_rate": 3.3306514003311305e-05,
      "loss": 0.5619,
      "step": 420
    },
    {
      "epoch": 1.7053164556962024,
      "grad_norm": 0.5098839352046093,
      "learning_rate": 3.326408710419996e-05,
      "loss": 0.5596,
      "step": 421
    },
    {
      "epoch": 1.709367088607595,
      "grad_norm": 0.3472595893563582,
      "learning_rate": 3.322155337795454e-05,
      "loss": 0.5613,
      "step": 422
    },
    {
      "epoch": 1.7134177215189874,
      "grad_norm": 0.3785902114397688,
      "learning_rate": 3.317891316713587e-05,
      "loss": 0.5614,
      "step": 423
    },
    {
      "epoch": 1.7174683544303797,
      "grad_norm": 0.3999167394487493,
      "learning_rate": 3.313616681516231e-05,
      "loss": 0.5464,
      "step": 424
    },
    {
      "epoch": 1.721518987341772,
      "grad_norm": 0.328602600378993,
      "learning_rate": 3.309331466630713e-05,
      "loss": 0.5509,
      "step": 425
    },
    {
      "epoch": 1.7255696202531645,
      "grad_norm": 0.4717406926838388,
      "learning_rate": 3.305035706569563e-05,
      "loss": 0.5515,
      "step": 426
    },
    {
      "epoch": 1.729620253164557,
      "grad_norm": 0.4069885866098542,
      "learning_rate": 3.3007294359302433e-05,
      "loss": 0.5605,
      "step": 427
    },
    {
      "epoch": 1.7336708860759493,
      "grad_norm": 0.32473583698658115,
      "learning_rate": 3.296412689394864e-05,
      "loss": 0.5551,
      "step": 428
    },
    {
      "epoch": 1.7377215189873416,
      "grad_norm": 0.40315100763093503,
      "learning_rate": 3.292085501729909e-05,
      "loss": 0.5527,
      "step": 429
    },
    {
      "epoch": 1.7417721518987341,
      "grad_norm": 0.40146642378597847,
      "learning_rate": 3.2877479077859534e-05,
      "loss": 0.5498,
      "step": 430
    },
    {
      "epoch": 1.7458227848101266,
      "grad_norm": 0.3835526421579511,
      "learning_rate": 3.283399942497381e-05,
      "loss": 0.5575,
      "step": 431
    },
    {
      "epoch": 1.749873417721519,
      "grad_norm": 0.3781079642934701,
      "learning_rate": 3.279041640882108e-05,
      "loss": 0.5429,
      "step": 432
    },
    {
      "epoch": 1.7539240506329112,
      "grad_norm": 0.37414585707301634,
      "learning_rate": 3.2746730380412964e-05,
      "loss": 0.5501,
      "step": 433
    },
    {
      "epoch": 1.7579746835443038,
      "grad_norm": 0.4194137648217172,
      "learning_rate": 3.2702941691590726e-05,
      "loss": 0.5586,
      "step": 434
    },
    {
      "epoch": 1.7620253164556963,
      "grad_norm": 0.44486461920229337,
      "learning_rate": 3.265905069502244e-05,
      "loss": 0.5352,
      "step": 435
    },
    {
      "epoch": 1.7660759493670886,
      "grad_norm": 0.39796653729890713,
      "learning_rate": 3.261505774420016e-05,
      "loss": 0.5599,
      "step": 436
    },
    {
      "epoch": 1.7701265822784809,
      "grad_norm": 0.38088811279413076,
      "learning_rate": 3.257096319343707e-05,
      "loss": 0.5523,
      "step": 437
    },
    {
      "epoch": 1.7741772151898734,
      "grad_norm": 0.40535650818029045,
      "learning_rate": 3.2526767397864614e-05,
      "loss": 0.5519,
      "step": 438
    },
    {
      "epoch": 1.7782278481012659,
      "grad_norm": 0.33499386127019043,
      "learning_rate": 3.248247071342966e-05,
      "loss": 0.5495,
      "step": 439
    },
    {
      "epoch": 1.7822784810126582,
      "grad_norm": 0.413207335048157,
      "learning_rate": 3.243807349689161e-05,
      "loss": 0.5409,
      "step": 440
    },
    {
      "epoch": 1.7863291139240505,
      "grad_norm": 0.3799132792068668,
      "learning_rate": 3.2393576105819544e-05,
      "loss": 0.554,
      "step": 441
    },
    {
      "epoch": 1.790379746835443,
      "grad_norm": 0.3785118165459444,
      "learning_rate": 3.2348978898589333e-05,
      "loss": 0.555,
      "step": 442
    },
    {
      "epoch": 1.7944303797468355,
      "grad_norm": 0.44737161008111903,
      "learning_rate": 3.230428223438075e-05,
      "loss": 0.5519,
      "step": 443
    },
    {
      "epoch": 1.7984810126582278,
      "grad_norm": 0.34840753719995593,
      "learning_rate": 3.225948647317459e-05,
      "loss": 0.5579,
      "step": 444
    },
    {
      "epoch": 1.80253164556962,
      "grad_norm": 1.5568882656017846,
      "learning_rate": 3.2214591975749745e-05,
      "loss": 0.5583,
      "step": 445
    },
    {
      "epoch": 1.8065822784810126,
      "grad_norm": 0.35986765358993783,
      "learning_rate": 3.216959910368034e-05,
      "loss": 0.5546,
      "step": 446
    },
    {
      "epoch": 1.8106329113924051,
      "grad_norm": 0.3217379715183224,
      "learning_rate": 3.212450821933277e-05,
      "loss": 0.5481,
      "step": 447
    },
    {
      "epoch": 1.8146835443037974,
      "grad_norm": 0.4086549446478703,
      "learning_rate": 3.207931968586281e-05,
      "loss": 0.5507,
      "step": 448
    },
    {
      "epoch": 1.81873417721519,
      "grad_norm": 0.31522601975313713,
      "learning_rate": 3.203403386721272e-05,
      "loss": 0.5582,
      "step": 449
    },
    {
      "epoch": 1.8227848101265822,
      "grad_norm": 0.35603666871896755,
      "learning_rate": 3.1988651128108245e-05,
      "loss": 0.5382,
      "step": 450
    },
    {
      "epoch": 1.8268354430379747,
      "grad_norm": 0.37995040861181373,
      "learning_rate": 3.194317183405573e-05,
      "loss": 0.56,
      "step": 451
    },
    {
      "epoch": 1.8308860759493673,
      "grad_norm": 0.42316059996582917,
      "learning_rate": 3.189759635133914e-05,
      "loss": 0.5564,
      "step": 452
    },
    {
      "epoch": 1.8349367088607595,
      "grad_norm": 0.37119489529724103,
      "learning_rate": 3.185192504701718e-05,
      "loss": 0.5505,
      "step": 453
    },
    {
      "epoch": 1.8389873417721518,
      "grad_norm": 0.6140310844022236,
      "learning_rate": 3.1806158288920234e-05,
      "loss": 0.5579,
      "step": 454
    },
    {
      "epoch": 1.8430379746835444,
      "grad_norm": 0.4005974615746777,
      "learning_rate": 3.1760296445647477e-05,
      "loss": 0.5521,
      "step": 455
    },
    {
      "epoch": 1.8470886075949369,
      "grad_norm": 0.480327027780718,
      "learning_rate": 3.1714339886563896e-05,
      "loss": 0.5483,
      "step": 456
    },
    {
      "epoch": 1.8511392405063292,
      "grad_norm": 0.5066076723298434,
      "learning_rate": 3.166828898179731e-05,
      "loss": 0.5493,
      "step": 457
    },
    {
      "epoch": 1.8551898734177215,
      "grad_norm": 0.421774286973629,
      "learning_rate": 3.162214410223536e-05,
      "loss": 0.558,
      "step": 458
    },
    {
      "epoch": 1.859240506329114,
      "grad_norm": 0.39011805786156156,
      "learning_rate": 3.157590561952257e-05,
      "loss": 0.5507,
      "step": 459
    },
    {
      "epoch": 1.8632911392405065,
      "grad_norm": 0.44626576552868896,
      "learning_rate": 3.152957390605732e-05,
      "loss": 0.5457,
      "step": 460
    },
    {
      "epoch": 1.8673417721518988,
      "grad_norm": 0.46268719896908356,
      "learning_rate": 3.148314933498886e-05,
      "loss": 0.5464,
      "step": 461
    },
    {
      "epoch": 1.871392405063291,
      "grad_norm": 0.3672551810541897,
      "learning_rate": 3.143663228021431e-05,
      "loss": 0.553,
      "step": 462
    },
    {
      "epoch": 1.8754430379746836,
      "grad_norm": 0.42280353289622374,
      "learning_rate": 3.1390023116375624e-05,
      "loss": 0.5406,
      "step": 463
    },
    {
      "epoch": 1.879493670886076,
      "grad_norm": 0.47928939040554835,
      "learning_rate": 3.134332221885661e-05,
      "loss": 0.5475,
      "step": 464
    },
    {
      "epoch": 1.8835443037974684,
      "grad_norm": 0.32903022717307884,
      "learning_rate": 3.129652996377987e-05,
      "loss": 0.5362,
      "step": 465
    },
    {
      "epoch": 1.8875949367088607,
      "grad_norm": 0.4141333974792075,
      "learning_rate": 3.12496467280038e-05,
      "loss": 0.5633,
      "step": 466
    },
    {
      "epoch": 1.8916455696202532,
      "grad_norm": 0.4857119369648965,
      "learning_rate": 3.120267288911952e-05,
      "loss": 0.5472,
      "step": 467
    },
    {
      "epoch": 1.8956962025316457,
      "grad_norm": 0.44568408737494597,
      "learning_rate": 3.11556088254479e-05,
      "loss": 0.5367,
      "step": 468
    },
    {
      "epoch": 1.899746835443038,
      "grad_norm": 0.4521030595411068,
      "learning_rate": 3.11084549160364e-05,
      "loss": 0.5598,
      "step": 469
    },
    {
      "epoch": 1.9037974683544303,
      "grad_norm": 0.6007995013471572,
      "learning_rate": 3.106121154065615e-05,
      "loss": 0.5476,
      "step": 470
    },
    {
      "epoch": 1.9078481012658228,
      "grad_norm": 0.3699592020522783,
      "learning_rate": 3.1013879079798805e-05,
      "loss": 0.5579,
      "step": 471
    },
    {
      "epoch": 1.9118987341772153,
      "grad_norm": 0.38731079740797963,
      "learning_rate": 3.096645791467348e-05,
      "loss": 0.5556,
      "step": 472
    },
    {
      "epoch": 1.9159493670886076,
      "grad_norm": 0.40466780924840534,
      "learning_rate": 3.091894842720373e-05,
      "loss": 0.5512,
      "step": 473
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.7545867339454393,
      "learning_rate": 3.0871351000024425e-05,
      "loss": 0.5408,
      "step": 474
    },
    {
      "epoch": 1.9240506329113924,
      "grad_norm": 0.3895015606195963,
      "learning_rate": 3.0823666016478716e-05,
      "loss": 0.5481,
      "step": 475
    },
    {
      "epoch": 1.928101265822785,
      "grad_norm": 0.47253230708767807,
      "learning_rate": 3.0775893860614896e-05,
      "loss": 0.5479,
      "step": 476
    },
    {
      "epoch": 1.9321518987341773,
      "grad_norm": 0.49637984264049784,
      "learning_rate": 3.0728034917183336e-05,
      "loss": 0.5642,
      "step": 477
    },
    {
      "epoch": 1.9362025316455695,
      "grad_norm": 0.456515947833048,
      "learning_rate": 3.06800895716334e-05,
      "loss": 0.5468,
      "step": 478
    },
    {
      "epoch": 1.940253164556962,
      "grad_norm": 0.4396441553232154,
      "learning_rate": 3.063205821011029e-05,
      "loss": 0.5568,
      "step": 479
    },
    {
      "epoch": 1.9443037974683546,
      "grad_norm": 0.32412418608667054,
      "learning_rate": 3.0583941219452016e-05,
      "loss": 0.5473,
      "step": 480
    },
    {
      "epoch": 1.9483544303797469,
      "grad_norm": 0.4961422198472992,
      "learning_rate": 3.053573898718618e-05,
      "loss": 0.5399,
      "step": 481
    },
    {
      "epoch": 1.9524050632911392,
      "grad_norm": 0.3408432794950519,
      "learning_rate": 3.0487451901526956e-05,
      "loss": 0.5539,
      "step": 482
    },
    {
      "epoch": 1.9564556962025317,
      "grad_norm": 0.41564962050408244,
      "learning_rate": 3.0439080351371875e-05,
      "loss": 0.5523,
      "step": 483
    },
    {
      "epoch": 1.9605063291139242,
      "grad_norm": 0.34682696253986167,
      "learning_rate": 3.0390624726298764e-05,
      "loss": 0.5461,
      "step": 484
    },
    {
      "epoch": 1.9645569620253165,
      "grad_norm": 0.39229371798949425,
      "learning_rate": 3.034208541656255e-05,
      "loss": 0.5536,
      "step": 485
    },
    {
      "epoch": 1.9686075949367088,
      "grad_norm": 0.4191945669221421,
      "learning_rate": 3.029346281309218e-05,
      "loss": 0.5549,
      "step": 486
    },
    {
      "epoch": 1.9726582278481013,
      "grad_norm": 0.3856287999939338,
      "learning_rate": 3.0244757307487415e-05,
      "loss": 0.5596,
      "step": 487
    },
    {
      "epoch": 1.9767088607594938,
      "grad_norm": 0.3880291392677673,
      "learning_rate": 3.019596929201569e-05,
      "loss": 0.5451,
      "step": 488
    },
    {
      "epoch": 1.980759493670886,
      "grad_norm": 0.4190330170063433,
      "learning_rate": 3.0147099159608985e-05,
      "loss": 0.5431,
      "step": 489
    },
    {
      "epoch": 1.9848101265822784,
      "grad_norm": 0.3767857080041659,
      "learning_rate": 3.0098147303860616e-05,
      "loss": 0.5485,
      "step": 490
    },
    {
      "epoch": 1.988860759493671,
      "grad_norm": 0.4375561091639583,
      "learning_rate": 3.0049114119022117e-05,
      "loss": 0.5338,
      "step": 491
    },
    {
      "epoch": 1.9929113924050634,
      "grad_norm": 0.40473453362874384,
      "learning_rate": 3.0000000000000004e-05,
      "loss": 0.5499,
      "step": 492
    },
    {
      "epoch": 1.9969620253164557,
      "grad_norm": 0.4396097525195346,
      "learning_rate": 2.995080534235264e-05,
      "loss": 0.5501,
      "step": 493
    },
    {
      "epoch": 2.001012658227848,
      "grad_norm": 0.37391599160087563,
      "learning_rate": 2.9901530542287044e-05,
      "loss": 0.5322,
      "step": 494
    },
    {
      "epoch": 2.0050632911392405,
      "grad_norm": 0.49122131638104555,
      "learning_rate": 2.9852175996655676e-05,
      "loss": 0.4947,
      "step": 495
    },
    {
      "epoch": 2.009113924050633,
      "grad_norm": 0.4912953483232416,
      "learning_rate": 2.980274210295326e-05,
      "loss": 0.4856,
      "step": 496
    },
    {
      "epoch": 2.013164556962025,
      "grad_norm": 0.5147763397556768,
      "learning_rate": 2.9753229259313578e-05,
      "loss": 0.4769,
      "step": 497
    },
    {
      "epoch": 2.0172151898734176,
      "grad_norm": 0.46846565691333575,
      "learning_rate": 2.9703637864506274e-05,
      "loss": 0.4893,
      "step": 498
    },
    {
      "epoch": 2.02126582278481,
      "grad_norm": 0.4004294264255696,
      "learning_rate": 2.965396831793362e-05,
      "loss": 0.4755,
      "step": 499
    },
    {
      "epoch": 2.0253164556962027,
      "grad_norm": 0.39594669728031423,
      "learning_rate": 2.9604221019627316e-05,
      "loss": 0.4756,
      "step": 500
    },
    {
      "epoch": 2.0293670886075947,
      "grad_norm": 0.3967409533654247,
      "learning_rate": 2.955439637024526e-05,
      "loss": 0.4867,
      "step": 501
    },
    {
      "epoch": 2.0334177215189873,
      "grad_norm": 0.40025467941750364,
      "learning_rate": 2.9504494771068334e-05,
      "loss": 0.4772,
      "step": 502
    },
    {
      "epoch": 2.0374683544303798,
      "grad_norm": 0.3464599114814506,
      "learning_rate": 2.9454516623997156e-05,
      "loss": 0.4931,
      "step": 503
    },
    {
      "epoch": 2.0415189873417723,
      "grad_norm": 0.39815745661294777,
      "learning_rate": 2.9404462331548847e-05,
      "loss": 0.4968,
      "step": 504
    },
    {
      "epoch": 2.0455696202531644,
      "grad_norm": 0.44784201942688134,
      "learning_rate": 2.93543322968538e-05,
      "loss": 0.4793,
      "step": 505
    },
    {
      "epoch": 2.049620253164557,
      "grad_norm": 0.3682949200255592,
      "learning_rate": 2.9304126923652428e-05,
      "loss": 0.4796,
      "step": 506
    },
    {
      "epoch": 2.0536708860759494,
      "grad_norm": 0.4572238422709273,
      "learning_rate": 2.9253846616291896e-05,
      "loss": 0.4928,
      "step": 507
    },
    {
      "epoch": 2.057721518987342,
      "grad_norm": 0.35424157030358006,
      "learning_rate": 2.9203491779722896e-05,
      "loss": 0.4659,
      "step": 508
    },
    {
      "epoch": 2.061772151898734,
      "grad_norm": 0.3883008720991813,
      "learning_rate": 2.9153062819496357e-05,
      "loss": 0.4743,
      "step": 509
    },
    {
      "epoch": 2.0658227848101265,
      "grad_norm": 0.4248717335770251,
      "learning_rate": 2.9102560141760178e-05,
      "loss": 0.495,
      "step": 510
    },
    {
      "epoch": 2.069873417721519,
      "grad_norm": 0.3222166607501827,
      "learning_rate": 2.9051984153256004e-05,
      "loss": 0.4802,
      "step": 511
    },
    {
      "epoch": 2.0739240506329115,
      "grad_norm": 0.40264151622294264,
      "learning_rate": 2.900133526131588e-05,
      "loss": 0.4769,
      "step": 512
    },
    {
      "epoch": 2.0779746835443036,
      "grad_norm": 0.3143923845582913,
      "learning_rate": 2.8950613873859025e-05,
      "loss": 0.4819,
      "step": 513
    },
    {
      "epoch": 2.082025316455696,
      "grad_norm": 0.3879055208429428,
      "learning_rate": 2.8899820399388515e-05,
      "loss": 0.4782,
      "step": 514
    },
    {
      "epoch": 2.0860759493670886,
      "grad_norm": 0.4080688136138641,
      "learning_rate": 2.8848955246988012e-05,
      "loss": 0.4835,
      "step": 515
    },
    {
      "epoch": 2.090126582278481,
      "grad_norm": 0.3471808507448825,
      "learning_rate": 2.879801882631847e-05,
      "loss": 0.4822,
      "step": 516
    },
    {
      "epoch": 2.094177215189873,
      "grad_norm": 0.36900076863745535,
      "learning_rate": 2.8747011547614808e-05,
      "loss": 0.4901,
      "step": 517
    },
    {
      "epoch": 2.0982278481012657,
      "grad_norm": 0.31815525328782956,
      "learning_rate": 2.8695933821682635e-05,
      "loss": 0.4908,
      "step": 518
    },
    {
      "epoch": 2.1022784810126582,
      "grad_norm": 0.3664017669027178,
      "learning_rate": 2.864478605989494e-05,
      "loss": 0.481,
      "step": 519
    },
    {
      "epoch": 2.1063291139240508,
      "grad_norm": 0.3425383754066873,
      "learning_rate": 2.8593568674188765e-05,
      "loss": 0.4892,
      "step": 520
    },
    {
      "epoch": 2.110379746835443,
      "grad_norm": 0.3136835992143645,
      "learning_rate": 2.8542282077061892e-05,
      "loss": 0.4857,
      "step": 521
    },
    {
      "epoch": 2.1144303797468353,
      "grad_norm": 0.3516905421218965,
      "learning_rate": 2.8490926681569523e-05,
      "loss": 0.4805,
      "step": 522
    },
    {
      "epoch": 2.118481012658228,
      "grad_norm": 0.36039819720193383,
      "learning_rate": 2.8439502901320956e-05,
      "loss": 0.4924,
      "step": 523
    },
    {
      "epoch": 2.1225316455696204,
      "grad_norm": 0.3362423938878692,
      "learning_rate": 2.8388011150476237e-05,
      "loss": 0.4893,
      "step": 524
    },
    {
      "epoch": 2.1265822784810124,
      "grad_norm": 0.6724013562784731,
      "learning_rate": 2.8336451843742866e-05,
      "loss": 0.4828,
      "step": 525
    },
    {
      "epoch": 2.130632911392405,
      "grad_norm": 0.34309763481416056,
      "learning_rate": 2.8284825396372387e-05,
      "loss": 0.4801,
      "step": 526
    },
    {
      "epoch": 2.1346835443037975,
      "grad_norm": 0.7213939565997735,
      "learning_rate": 2.8233132224157132e-05,
      "loss": 0.4891,
      "step": 527
    },
    {
      "epoch": 2.13873417721519,
      "grad_norm": 0.3815011540538773,
      "learning_rate": 2.8181372743426805e-05,
      "loss": 0.478,
      "step": 528
    },
    {
      "epoch": 2.1427848101265825,
      "grad_norm": 0.3075526643321197,
      "learning_rate": 2.8129547371045128e-05,
      "loss": 0.4897,
      "step": 529
    },
    {
      "epoch": 2.1468354430379746,
      "grad_norm": 0.40011781034652194,
      "learning_rate": 2.8077656524406534e-05,
      "loss": 0.4901,
      "step": 530
    },
    {
      "epoch": 2.150886075949367,
      "grad_norm": 0.39939998464752,
      "learning_rate": 2.802570062143278e-05,
      "loss": 0.475,
      "step": 531
    },
    {
      "epoch": 2.1549367088607596,
      "grad_norm": 0.3082736403075437,
      "learning_rate": 2.7973680080569555e-05,
      "loss": 0.4758,
      "step": 532
    },
    {
      "epoch": 2.1589873417721517,
      "grad_norm": 0.459236522347348,
      "learning_rate": 2.792159532078314e-05,
      "loss": 0.4968,
      "step": 533
    },
    {
      "epoch": 2.163037974683544,
      "grad_norm": 0.455468053436245,
      "learning_rate": 2.7869446761557033e-05,
      "loss": 0.4689,
      "step": 534
    },
    {
      "epoch": 2.1670886075949367,
      "grad_norm": 0.366946167970664,
      "learning_rate": 2.781723482288857e-05,
      "loss": 0.4834,
      "step": 535
    },
    {
      "epoch": 2.1711392405063292,
      "grad_norm": 0.3749063755002824,
      "learning_rate": 2.7764959925285517e-05,
      "loss": 0.4819,
      "step": 536
    },
    {
      "epoch": 2.1751898734177217,
      "grad_norm": 0.37520501286882674,
      "learning_rate": 2.771262248976272e-05,
      "loss": 0.4931,
      "step": 537
    },
    {
      "epoch": 2.179240506329114,
      "grad_norm": 0.3520419187862786,
      "learning_rate": 2.7660222937838677e-05,
      "loss": 0.4809,
      "step": 538
    },
    {
      "epoch": 2.1832911392405063,
      "grad_norm": 0.3069975151071309,
      "learning_rate": 2.7607761691532186e-05,
      "loss": 0.478,
      "step": 539
    },
    {
      "epoch": 2.187341772151899,
      "grad_norm": 0.3377357916767123,
      "learning_rate": 2.7555239173358916e-05,
      "loss": 0.4838,
      "step": 540
    },
    {
      "epoch": 2.191392405063291,
      "grad_norm": 0.31632923565152593,
      "learning_rate": 2.7502655806328e-05,
      "loss": 0.4735,
      "step": 541
    },
    {
      "epoch": 2.1954430379746834,
      "grad_norm": 0.3478255075100954,
      "learning_rate": 2.7450012013938648e-05,
      "loss": 0.4805,
      "step": 542
    },
    {
      "epoch": 2.199493670886076,
      "grad_norm": 0.3306368997185833,
      "learning_rate": 2.739730822017673e-05,
      "loss": 0.4626,
      "step": 543
    },
    {
      "epoch": 2.2035443037974685,
      "grad_norm": 0.346618672744398,
      "learning_rate": 2.7344544849511355e-05,
      "loss": 0.488,
      "step": 544
    },
    {
      "epoch": 2.207594936708861,
      "grad_norm": 0.3402565028658384,
      "learning_rate": 2.7291722326891456e-05,
      "loss": 0.494,
      "step": 545
    },
    {
      "epoch": 2.211645569620253,
      "grad_norm": 0.3128305778057646,
      "learning_rate": 2.723884107774236e-05,
      "loss": 0.4938,
      "step": 546
    },
    {
      "epoch": 2.2156962025316456,
      "grad_norm": 0.37399356171374426,
      "learning_rate": 2.718590152796239e-05,
      "loss": 0.4734,
      "step": 547
    },
    {
      "epoch": 2.219746835443038,
      "grad_norm": 0.2864463926809224,
      "learning_rate": 2.71329041039194e-05,
      "loss": 0.4898,
      "step": 548
    },
    {
      "epoch": 2.22379746835443,
      "grad_norm": 0.3878206764000815,
      "learning_rate": 2.7079849232447357e-05,
      "loss": 0.4842,
      "step": 549
    },
    {
      "epoch": 2.2278481012658227,
      "grad_norm": 0.30004830228595986,
      "learning_rate": 2.7026737340842895e-05,
      "loss": 0.4847,
      "step": 550
    },
    {
      "epoch": 2.231898734177215,
      "grad_norm": 0.3471536595576494,
      "learning_rate": 2.697356885686189e-05,
      "loss": 0.4666,
      "step": 551
    },
    {
      "epoch": 2.2359493670886077,
      "grad_norm": 0.35900108172368134,
      "learning_rate": 2.6920344208716014e-05,
      "loss": 0.4889,
      "step": 552
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.36801343148551374,
      "learning_rate": 2.6867063825069252e-05,
      "loss": 0.4969,
      "step": 553
    },
    {
      "epoch": 2.2440506329113923,
      "grad_norm": 0.37229883756087906,
      "learning_rate": 2.6813728135034494e-05,
      "loss": 0.4699,
      "step": 554
    },
    {
      "epoch": 2.248101265822785,
      "grad_norm": 0.39559864132909733,
      "learning_rate": 2.6760337568170056e-05,
      "loss": 0.4812,
      "step": 555
    },
    {
      "epoch": 2.2521518987341773,
      "grad_norm": 0.3254354608161272,
      "learning_rate": 2.6706892554476226e-05,
      "loss": 0.4861,
      "step": 556
    },
    {
      "epoch": 2.2562025316455694,
      "grad_norm": 0.3986451019380668,
      "learning_rate": 2.6653393524391795e-05,
      "loss": 0.4783,
      "step": 557
    },
    {
      "epoch": 2.260253164556962,
      "grad_norm": 0.4230310617704062,
      "learning_rate": 2.6599840908790592e-05,
      "loss": 0.4866,
      "step": 558
    },
    {
      "epoch": 2.2643037974683544,
      "grad_norm": 0.3163478833560352,
      "learning_rate": 2.6546235138978028e-05,
      "loss": 0.4848,
      "step": 559
    },
    {
      "epoch": 2.268354430379747,
      "grad_norm": 0.3460501050989648,
      "learning_rate": 2.6492576646687597e-05,
      "loss": 0.4891,
      "step": 560
    },
    {
      "epoch": 2.2724050632911394,
      "grad_norm": 0.35591988177797335,
      "learning_rate": 2.6438865864077425e-05,
      "loss": 0.4948,
      "step": 561
    },
    {
      "epoch": 2.2764556962025315,
      "grad_norm": 0.31233153122822277,
      "learning_rate": 2.6385103223726766e-05,
      "loss": 0.481,
      "step": 562
    },
    {
      "epoch": 2.280506329113924,
      "grad_norm": 0.4089281963756193,
      "learning_rate": 2.6331289158632537e-05,
      "loss": 0.484,
      "step": 563
    },
    {
      "epoch": 2.2845569620253166,
      "grad_norm": 0.362270426437326,
      "learning_rate": 2.6277424102205817e-05,
      "loss": 0.491,
      "step": 564
    },
    {
      "epoch": 2.2886075949367086,
      "grad_norm": 0.3931915307642271,
      "learning_rate": 2.6223508488268374e-05,
      "loss": 0.4809,
      "step": 565
    },
    {
      "epoch": 2.292658227848101,
      "grad_norm": 0.5003313448518608,
      "learning_rate": 2.6169542751049148e-05,
      "loss": 0.4858,
      "step": 566
    },
    {
      "epoch": 2.2967088607594937,
      "grad_norm": 0.30533130478469167,
      "learning_rate": 2.6115527325180754e-05,
      "loss": 0.4805,
      "step": 567
    },
    {
      "epoch": 2.300759493670886,
      "grad_norm": 0.4266751868107556,
      "learning_rate": 2.606146264569603e-05,
      "loss": 0.4935,
      "step": 568
    },
    {
      "epoch": 2.3048101265822787,
      "grad_norm": 0.37420395259078654,
      "learning_rate": 2.6007349148024447e-05,
      "loss": 0.4998,
      "step": 569
    },
    {
      "epoch": 2.3088607594936708,
      "grad_norm": 0.3371435844743191,
      "learning_rate": 2.5953187267988694e-05,
      "loss": 0.4971,
      "step": 570
    },
    {
      "epoch": 2.3129113924050633,
      "grad_norm": 0.4430142427784052,
      "learning_rate": 2.5898977441801097e-05,
      "loss": 0.4916,
      "step": 571
    },
    {
      "epoch": 2.316962025316456,
      "grad_norm": 0.3052966006592978,
      "learning_rate": 2.584472010606015e-05,
      "loss": 0.4772,
      "step": 572
    },
    {
      "epoch": 2.321012658227848,
      "grad_norm": 0.3646795072339449,
      "learning_rate": 2.5790415697746976e-05,
      "loss": 0.4714,
      "step": 573
    },
    {
      "epoch": 2.3250632911392404,
      "grad_norm": 0.31376391359299516,
      "learning_rate": 2.5736064654221808e-05,
      "loss": 0.4788,
      "step": 574
    },
    {
      "epoch": 2.329113924050633,
      "grad_norm": 0.3532241114913221,
      "learning_rate": 2.568166741322048e-05,
      "loss": 0.4865,
      "step": 575
    },
    {
      "epoch": 2.3331645569620254,
      "grad_norm": 0.374648674808917,
      "learning_rate": 2.56272244128509e-05,
      "loss": 0.4822,
      "step": 576
    },
    {
      "epoch": 2.337215189873418,
      "grad_norm": 0.33279505671367426,
      "learning_rate": 2.55727360915895e-05,
      "loss": 0.4767,
      "step": 577
    },
    {
      "epoch": 2.34126582278481,
      "grad_norm": 0.3718466898900236,
      "learning_rate": 2.5518202888277734e-05,
      "loss": 0.4758,
      "step": 578
    },
    {
      "epoch": 2.3453164556962025,
      "grad_norm": 0.38148051883046163,
      "learning_rate": 2.5463625242118523e-05,
      "loss": 0.4789,
      "step": 579
    },
    {
      "epoch": 2.349367088607595,
      "grad_norm": 0.3193400916610199,
      "learning_rate": 2.5409003592672723e-05,
      "loss": 0.4875,
      "step": 580
    },
    {
      "epoch": 2.353417721518987,
      "grad_norm": 0.3756648906771208,
      "learning_rate": 2.535433837985559e-05,
      "loss": 0.4778,
      "step": 581
    },
    {
      "epoch": 2.3574683544303796,
      "grad_norm": 0.3444267909830687,
      "learning_rate": 2.529963004393324e-05,
      "loss": 0.4842,
      "step": 582
    },
    {
      "epoch": 2.361518987341772,
      "grad_norm": 0.30641270620778593,
      "learning_rate": 2.524487902551908e-05,
      "loss": 0.48,
      "step": 583
    },
    {
      "epoch": 2.3655696202531646,
      "grad_norm": 0.3371221807735811,
      "learning_rate": 2.519008576557029e-05,
      "loss": 0.4891,
      "step": 584
    },
    {
      "epoch": 2.369620253164557,
      "grad_norm": 0.3531072441619311,
      "learning_rate": 2.5135250705384254e-05,
      "loss": 0.4824,
      "step": 585
    },
    {
      "epoch": 2.3736708860759492,
      "grad_norm": 0.3444892349903489,
      "learning_rate": 2.5080374286595007e-05,
      "loss": 0.483,
      "step": 586
    },
    {
      "epoch": 2.3777215189873417,
      "grad_norm": 0.35545597294655396,
      "learning_rate": 2.5025456951169677e-05,
      "loss": 0.4873,
      "step": 587
    },
    {
      "epoch": 2.3817721518987343,
      "grad_norm": 0.30409201446435624,
      "learning_rate": 2.4970499141404942e-05,
      "loss": 0.4885,
      "step": 588
    },
    {
      "epoch": 2.3858227848101268,
      "grad_norm": 0.41116719524083234,
      "learning_rate": 2.491550129992345e-05,
      "loss": 0.5108,
      "step": 589
    },
    {
      "epoch": 2.389873417721519,
      "grad_norm": 0.2678404787344525,
      "learning_rate": 2.486046386967024e-05,
      "loss": 0.4824,
      "step": 590
    },
    {
      "epoch": 2.3939240506329114,
      "grad_norm": 0.37622991846503545,
      "learning_rate": 2.4805387293909214e-05,
      "loss": 0.4876,
      "step": 591
    },
    {
      "epoch": 2.397974683544304,
      "grad_norm": 0.2877824160920245,
      "learning_rate": 2.4750272016219552e-05,
      "loss": 0.4862,
      "step": 592
    },
    {
      "epoch": 2.4020253164556964,
      "grad_norm": 0.3407753712384016,
      "learning_rate": 2.4695118480492114e-05,
      "loss": 0.4861,
      "step": 593
    },
    {
      "epoch": 2.4060759493670885,
      "grad_norm": 0.29345549486205685,
      "learning_rate": 2.4639927130925898e-05,
      "loss": 0.4925,
      "step": 594
    },
    {
      "epoch": 2.410126582278481,
      "grad_norm": 0.29123765717843236,
      "learning_rate": 2.458469841202444e-05,
      "loss": 0.4753,
      "step": 595
    },
    {
      "epoch": 2.4141772151898735,
      "grad_norm": 0.3150939099354683,
      "learning_rate": 2.452943276859226e-05,
      "loss": 0.4948,
      "step": 596
    },
    {
      "epoch": 2.418227848101266,
      "grad_norm": 0.3272504677755435,
      "learning_rate": 2.447413064573125e-05,
      "loss": 0.484,
      "step": 597
    },
    {
      "epoch": 2.422278481012658,
      "grad_norm": 0.29114977264989156,
      "learning_rate": 2.4418792488837095e-05,
      "loss": 0.4877,
      "step": 598
    },
    {
      "epoch": 2.4263291139240506,
      "grad_norm": 0.3134290362006714,
      "learning_rate": 2.4363418743595713e-05,
      "loss": 0.4847,
      "step": 599
    },
    {
      "epoch": 2.430379746835443,
      "grad_norm": 0.3157689663933256,
      "learning_rate": 2.430800985597963e-05,
      "loss": 0.4847,
      "step": 600
    },
    {
      "epoch": 2.4344303797468356,
      "grad_norm": 0.3592982945575709,
      "learning_rate": 2.4252566272244415e-05,
      "loss": 0.4962,
      "step": 601
    },
    {
      "epoch": 2.4384810126582277,
      "grad_norm": 0.30519033274569624,
      "learning_rate": 2.4197088438925063e-05,
      "loss": 0.4887,
      "step": 602
    },
    {
      "epoch": 2.44253164556962,
      "grad_norm": 0.31142735143630723,
      "learning_rate": 2.4141576802832417e-05,
      "loss": 0.48,
      "step": 603
    },
    {
      "epoch": 2.4465822784810127,
      "grad_norm": 1.785923107028605,
      "learning_rate": 2.408603181104957e-05,
      "loss": 0.5318,
      "step": 604
    },
    {
      "epoch": 2.4506329113924052,
      "grad_norm": 0.4767376417697447,
      "learning_rate": 2.4030453910928245e-05,
      "loss": 0.492,
      "step": 605
    },
    {
      "epoch": 2.4546835443037973,
      "grad_norm": 0.30466547518685516,
      "learning_rate": 2.397484355008521e-05,
      "loss": 0.4902,
      "step": 606
    },
    {
      "epoch": 2.45873417721519,
      "grad_norm": 0.3987732871615423,
      "learning_rate": 2.3919201176398662e-05,
      "loss": 0.4858,
      "step": 607
    },
    {
      "epoch": 2.4627848101265823,
      "grad_norm": 0.3769550506115508,
      "learning_rate": 2.3863527238004633e-05,
      "loss": 0.4905,
      "step": 608
    },
    {
      "epoch": 2.466835443037975,
      "grad_norm": 0.3416506828551855,
      "learning_rate": 2.380782218329337e-05,
      "loss": 0.4722,
      "step": 609
    },
    {
      "epoch": 2.470886075949367,
      "grad_norm": 0.39651135183976155,
      "learning_rate": 2.3752086460905725e-05,
      "loss": 0.4877,
      "step": 610
    },
    {
      "epoch": 2.4749367088607594,
      "grad_norm": 0.3267204693417352,
      "learning_rate": 2.3696320519729544e-05,
      "loss": 0.486,
      "step": 611
    },
    {
      "epoch": 2.478987341772152,
      "grad_norm": 0.30811421533717825,
      "learning_rate": 2.3640524808896045e-05,
      "loss": 0.4963,
      "step": 612
    },
    {
      "epoch": 2.4830379746835445,
      "grad_norm": 0.3148397013995186,
      "learning_rate": 2.3584699777776222e-05,
      "loss": 0.4778,
      "step": 613
    },
    {
      "epoch": 2.4870886075949366,
      "grad_norm": 0.30420982313346984,
      "learning_rate": 2.3528845875977195e-05,
      "loss": 0.5008,
      "step": 614
    },
    {
      "epoch": 2.491139240506329,
      "grad_norm": 2.8360338523187987,
      "learning_rate": 2.3472963553338614e-05,
      "loss": 0.5259,
      "step": 615
    },
    {
      "epoch": 2.4951898734177216,
      "grad_norm": 0.390227285284103,
      "learning_rate": 2.341705325992901e-05,
      "loss": 0.4851,
      "step": 616
    },
    {
      "epoch": 2.499240506329114,
      "grad_norm": 0.3026014201626577,
      "learning_rate": 2.336111544604222e-05,
      "loss": 0.4786,
      "step": 617
    },
    {
      "epoch": 2.503291139240506,
      "grad_norm": 0.3685982233286342,
      "learning_rate": 2.33051505621937e-05,
      "loss": 0.5022,
      "step": 618
    },
    {
      "epoch": 2.5073417721518987,
      "grad_norm": 0.3159739765877513,
      "learning_rate": 2.324915905911693e-05,
      "loss": 0.4832,
      "step": 619
    },
    {
      "epoch": 2.511392405063291,
      "grad_norm": 0.34338647086775387,
      "learning_rate": 2.319314138775977e-05,
      "loss": 0.4901,
      "step": 620
    },
    {
      "epoch": 2.5154430379746833,
      "grad_norm": 0.31185483506750616,
      "learning_rate": 2.3137097999280856e-05,
      "loss": 0.4829,
      "step": 621
    },
    {
      "epoch": 2.519493670886076,
      "grad_norm": 0.3427037014737789,
      "learning_rate": 2.308102934504593e-05,
      "loss": 0.4827,
      "step": 622
    },
    {
      "epoch": 2.5235443037974683,
      "grad_norm": 0.3070091639383764,
      "learning_rate": 2.3024935876624222e-05,
      "loss": 0.4826,
      "step": 623
    },
    {
      "epoch": 2.527594936708861,
      "grad_norm": 0.35237493064773545,
      "learning_rate": 2.2968818045784813e-05,
      "loss": 0.487,
      "step": 624
    },
    {
      "epoch": 2.5316455696202533,
      "grad_norm": 0.3358983790147517,
      "learning_rate": 2.2912676304493006e-05,
      "loss": 0.4853,
      "step": 625
    },
    {
      "epoch": 2.5356962025316454,
      "grad_norm": 0.2923021951327282,
      "learning_rate": 2.2856511104906668e-05,
      "loss": 0.48,
      "step": 626
    },
    {
      "epoch": 2.539746835443038,
      "grad_norm": 0.37264219494039313,
      "learning_rate": 2.2800322899372586e-05,
      "loss": 0.4892,
      "step": 627
    },
    {
      "epoch": 2.5437974683544304,
      "grad_norm": 0.26753436268210296,
      "learning_rate": 2.2744112140422844e-05,
      "loss": 0.4898,
      "step": 628
    },
    {
      "epoch": 2.547848101265823,
      "grad_norm": 0.3201917185629326,
      "learning_rate": 2.2687879280771177e-05,
      "loss": 0.4863,
      "step": 629
    },
    {
      "epoch": 2.5518987341772155,
      "grad_norm": 0.28853679868720905,
      "learning_rate": 2.26316247733093e-05,
      "loss": 0.484,
      "step": 630
    },
    {
      "epoch": 2.5559493670886075,
      "grad_norm": 0.37787062653726905,
      "learning_rate": 2.257534907110328e-05,
      "loss": 0.4943,
      "step": 631
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.3590548248321571,
      "learning_rate": 2.2519052627389882e-05,
      "loss": 0.4846,
      "step": 632
    },
    {
      "epoch": 2.5640506329113926,
      "grad_norm": 0.345107623897901,
      "learning_rate": 2.246273589557294e-05,
      "loss": 0.4856,
      "step": 633
    },
    {
      "epoch": 2.5681012658227846,
      "grad_norm": 0.3450933398494853,
      "learning_rate": 2.240639932921966e-05,
      "loss": 0.496,
      "step": 634
    },
    {
      "epoch": 2.572151898734177,
      "grad_norm": 0.30999446539277403,
      "learning_rate": 2.2350043382056995e-05,
      "loss": 0.4849,
      "step": 635
    },
    {
      "epoch": 2.5762025316455697,
      "grad_norm": 0.27892022721020954,
      "learning_rate": 2.2293668507968015e-05,
      "loss": 0.4784,
      "step": 636
    },
    {
      "epoch": 2.580253164556962,
      "grad_norm": 0.2705659478264623,
      "learning_rate": 2.2237275160988186e-05,
      "loss": 0.495,
      "step": 637
    },
    {
      "epoch": 2.5843037974683547,
      "grad_norm": 0.27489496856103884,
      "learning_rate": 2.2180863795301787e-05,
      "loss": 0.4881,
      "step": 638
    },
    {
      "epoch": 2.5883544303797468,
      "grad_norm": 1.1181093924768495,
      "learning_rate": 2.212443486523819e-05,
      "loss": 0.4927,
      "step": 639
    },
    {
      "epoch": 2.5924050632911393,
      "grad_norm": 0.2832937086796022,
      "learning_rate": 2.2067988825268243e-05,
      "loss": 0.4775,
      "step": 640
    },
    {
      "epoch": 2.596455696202532,
      "grad_norm": 0.3055676496939581,
      "learning_rate": 2.2011526130000596e-05,
      "loss": 0.4997,
      "step": 641
    },
    {
      "epoch": 2.600506329113924,
      "grad_norm": 0.33211817679521766,
      "learning_rate": 2.1955047234178038e-05,
      "loss": 0.484,
      "step": 642
    },
    {
      "epoch": 2.6045569620253164,
      "grad_norm": 0.31952107854812045,
      "learning_rate": 2.1898552592673825e-05,
      "loss": 0.4878,
      "step": 643
    },
    {
      "epoch": 2.608607594936709,
      "grad_norm": 0.3430578011864111,
      "learning_rate": 2.184204266048803e-05,
      "loss": 0.4713,
      "step": 644
    },
    {
      "epoch": 2.6126582278481014,
      "grad_norm": 0.3172929237384558,
      "learning_rate": 2.1785517892743887e-05,
      "loss": 0.4859,
      "step": 645
    },
    {
      "epoch": 2.616708860759494,
      "grad_norm": 0.29410861823564055,
      "learning_rate": 2.17289787446841e-05,
      "loss": 0.4849,
      "step": 646
    },
    {
      "epoch": 2.620759493670886,
      "grad_norm": 0.2987399906792195,
      "learning_rate": 2.1672425671667198e-05,
      "loss": 0.4892,
      "step": 647
    },
    {
      "epoch": 2.6248101265822785,
      "grad_norm": 0.2704890166645887,
      "learning_rate": 2.161585912916385e-05,
      "loss": 0.4938,
      "step": 648
    },
    {
      "epoch": 2.628860759493671,
      "grad_norm": 0.30131236093141617,
      "learning_rate": 2.1559279572753214e-05,
      "loss": 0.4785,
      "step": 649
    },
    {
      "epoch": 2.632911392405063,
      "grad_norm": 0.2946595263035566,
      "learning_rate": 2.1502687458119268e-05,
      "loss": 0.4865,
      "step": 650
    },
    {
      "epoch": 2.6369620253164556,
      "grad_norm": 0.2827921701435336,
      "learning_rate": 2.1446083241047116e-05,
      "loss": 0.4873,
      "step": 651
    },
    {
      "epoch": 2.641012658227848,
      "grad_norm": 0.27531959708369647,
      "learning_rate": 2.1389467377419333e-05,
      "loss": 0.4966,
      "step": 652
    },
    {
      "epoch": 2.6450632911392407,
      "grad_norm": 0.2629550403757396,
      "learning_rate": 2.133284032321232e-05,
      "loss": 0.4786,
      "step": 653
    },
    {
      "epoch": 2.649113924050633,
      "grad_norm": 0.27169380745886135,
      "learning_rate": 2.1276202534492566e-05,
      "loss": 0.4921,
      "step": 654
    },
    {
      "epoch": 2.6531645569620252,
      "grad_norm": 0.2822062950775314,
      "learning_rate": 2.121955446741306e-05,
      "loss": 0.4778,
      "step": 655
    },
    {
      "epoch": 2.6572151898734178,
      "grad_norm": 0.26480766583270837,
      "learning_rate": 2.1162896578209517e-05,
      "loss": 0.4835,
      "step": 656
    },
    {
      "epoch": 2.6612658227848103,
      "grad_norm": 0.3145987165234338,
      "learning_rate": 2.1106229323196813e-05,
      "loss": 0.4901,
      "step": 657
    },
    {
      "epoch": 2.6653164556962023,
      "grad_norm": 0.27802965121351736,
      "learning_rate": 2.1049553158765214e-05,
      "loss": 0.4872,
      "step": 658
    },
    {
      "epoch": 2.669367088607595,
      "grad_norm": 0.2772190167733706,
      "learning_rate": 2.0992868541376764e-05,
      "loss": 0.4848,
      "step": 659
    },
    {
      "epoch": 2.6734177215189874,
      "grad_norm": 0.2964133861305068,
      "learning_rate": 2.093617592756158e-05,
      "loss": 0.4835,
      "step": 660
    },
    {
      "epoch": 2.67746835443038,
      "grad_norm": 0.2686631235428652,
      "learning_rate": 2.0879475773914167e-05,
      "loss": 0.4886,
      "step": 661
    },
    {
      "epoch": 2.6815189873417724,
      "grad_norm": 0.28738821835992645,
      "learning_rate": 2.082276853708978e-05,
      "loss": 0.4798,
      "step": 662
    },
    {
      "epoch": 2.6855696202531645,
      "grad_norm": 0.27760265820499647,
      "learning_rate": 2.076605467380071e-05,
      "loss": 0.483,
      "step": 663
    },
    {
      "epoch": 2.689620253164557,
      "grad_norm": 0.34910188057473646,
      "learning_rate": 2.0709334640812613e-05,
      "loss": 0.48,
      "step": 664
    },
    {
      "epoch": 2.6936708860759495,
      "grad_norm": 0.2728465097891045,
      "learning_rate": 2.0652608894940824e-05,
      "loss": 0.4809,
      "step": 665
    },
    {
      "epoch": 2.6977215189873416,
      "grad_norm": 0.3392138872895886,
      "learning_rate": 2.0595877893046722e-05,
      "loss": 0.4858,
      "step": 666
    },
    {
      "epoch": 2.701772151898734,
      "grad_norm": 0.2761217937643811,
      "learning_rate": 2.0539142092033985e-05,
      "loss": 0.4841,
      "step": 667
    },
    {
      "epoch": 2.7058227848101266,
      "grad_norm": 0.31627534439623306,
      "learning_rate": 2.048240194884496e-05,
      "loss": 0.4806,
      "step": 668
    },
    {
      "epoch": 2.709873417721519,
      "grad_norm": 0.28813821414847485,
      "learning_rate": 2.042565792045695e-05,
      "loss": 0.4819,
      "step": 669
    },
    {
      "epoch": 2.7139240506329116,
      "grad_norm": 0.31444176251809336,
      "learning_rate": 2.036891046387857e-05,
      "loss": 0.4801,
      "step": 670
    },
    {
      "epoch": 2.7179746835443037,
      "grad_norm": 0.3295465533384432,
      "learning_rate": 2.0312160036146036e-05,
      "loss": 0.4775,
      "step": 671
    },
    {
      "epoch": 2.7220253164556962,
      "grad_norm": 0.34616549962784826,
      "learning_rate": 2.025540709431948e-05,
      "loss": 0.4888,
      "step": 672
    },
    {
      "epoch": 2.7260759493670887,
      "grad_norm": 0.3639111621097783,
      "learning_rate": 2.0198652095479298e-05,
      "loss": 0.4884,
      "step": 673
    },
    {
      "epoch": 2.730126582278481,
      "grad_norm": 0.29755922748279007,
      "learning_rate": 2.014189549672245e-05,
      "loss": 0.4865,
      "step": 674
    },
    {
      "epoch": 2.7341772151898733,
      "grad_norm": 0.3671932349452924,
      "learning_rate": 2.0085137755158776e-05,
      "loss": 0.4842,
      "step": 675
    },
    {
      "epoch": 2.738227848101266,
      "grad_norm": 0.32123330047225557,
      "learning_rate": 2.0028379327907327e-05,
      "loss": 0.4731,
      "step": 676
    },
    {
      "epoch": 2.7422784810126584,
      "grad_norm": 0.2992772869323405,
      "learning_rate": 1.9971620672092676e-05,
      "loss": 0.4793,
      "step": 677
    },
    {
      "epoch": 2.746329113924051,
      "grad_norm": 0.361224135945958,
      "learning_rate": 1.991486224484123e-05,
      "loss": 0.4907,
      "step": 678
    },
    {
      "epoch": 2.750379746835443,
      "grad_norm": 0.2773942838221015,
      "learning_rate": 1.985810450327756e-05,
      "loss": 0.4883,
      "step": 679
    },
    {
      "epoch": 2.7544303797468355,
      "grad_norm": 0.38140933617605177,
      "learning_rate": 1.9801347904520706e-05,
      "loss": 0.4789,
      "step": 680
    },
    {
      "epoch": 2.758481012658228,
      "grad_norm": 0.32110231249604904,
      "learning_rate": 1.974459290568053e-05,
      "loss": 0.4845,
      "step": 681
    },
    {
      "epoch": 2.76253164556962,
      "grad_norm": 0.28981794909734204,
      "learning_rate": 1.968783996385397e-05,
      "loss": 0.481,
      "step": 682
    },
    {
      "epoch": 2.7665822784810126,
      "grad_norm": 0.273012818100641,
      "learning_rate": 1.963108953612143e-05,
      "loss": 0.4877,
      "step": 683
    },
    {
      "epoch": 2.770632911392405,
      "grad_norm": 0.3086207143894874,
      "learning_rate": 1.9574342079543056e-05,
      "loss": 0.4755,
      "step": 684
    },
    {
      "epoch": 2.7746835443037976,
      "grad_norm": 0.2443632238240474,
      "learning_rate": 1.9517598051155046e-05,
      "loss": 0.4827,
      "step": 685
    },
    {
      "epoch": 2.77873417721519,
      "grad_norm": 0.30799952542679887,
      "learning_rate": 1.9460857907966025e-05,
      "loss": 0.4847,
      "step": 686
    },
    {
      "epoch": 2.782784810126582,
      "grad_norm": 0.2516019513986894,
      "learning_rate": 1.9404122106953285e-05,
      "loss": 0.4847,
      "step": 687
    },
    {
      "epoch": 2.7868354430379747,
      "grad_norm": 0.2964116959629949,
      "learning_rate": 1.9347391105059176e-05,
      "loss": 0.4841,
      "step": 688
    },
    {
      "epoch": 2.790886075949367,
      "grad_norm": 0.25347340844388294,
      "learning_rate": 1.92906653591874e-05,
      "loss": 0.4845,
      "step": 689
    },
    {
      "epoch": 2.7949367088607593,
      "grad_norm": 0.2836574678143199,
      "learning_rate": 1.9233945326199295e-05,
      "loss": 0.4815,
      "step": 690
    },
    {
      "epoch": 2.798987341772152,
      "grad_norm": 0.2805218942185103,
      "learning_rate": 1.917723146291022e-05,
      "loss": 0.4866,
      "step": 691
    },
    {
      "epoch": 2.8030379746835443,
      "grad_norm": 0.30495633686631485,
      "learning_rate": 1.912052422608584e-05,
      "loss": 0.48,
      "step": 692
    },
    {
      "epoch": 2.807088607594937,
      "grad_norm": 0.3150851250437663,
      "learning_rate": 1.9063824072438428e-05,
      "loss": 0.492,
      "step": 693
    },
    {
      "epoch": 2.8111392405063294,
      "grad_norm": 0.29010676970049976,
      "learning_rate": 1.9007131458623246e-05,
      "loss": 0.4901,
      "step": 694
    },
    {
      "epoch": 2.8151898734177214,
      "grad_norm": 0.2953077913123613,
      "learning_rate": 1.895044684123479e-05,
      "loss": 0.4965,
      "step": 695
    },
    {
      "epoch": 2.819240506329114,
      "grad_norm": 0.31830367813157284,
      "learning_rate": 1.8893770676803194e-05,
      "loss": 0.4743,
      "step": 696
    },
    {
      "epoch": 2.8232911392405065,
      "grad_norm": 0.267123075832414,
      "learning_rate": 1.8837103421790486e-05,
      "loss": 0.493,
      "step": 697
    },
    {
      "epoch": 2.8273417721518985,
      "grad_norm": 0.3329165954478567,
      "learning_rate": 1.8780445532586952e-05,
      "loss": 0.4972,
      "step": 698
    },
    {
      "epoch": 2.831392405063291,
      "grad_norm": 0.2662305165169784,
      "learning_rate": 1.872379746550743e-05,
      "loss": 0.4885,
      "step": 699
    },
    {
      "epoch": 2.8354430379746836,
      "grad_norm": 0.2769901302375608,
      "learning_rate": 1.866715967678769e-05,
      "loss": 0.4849,
      "step": 700
    },
    {
      "epoch": 2.839493670886076,
      "grad_norm": 0.2636340034964746,
      "learning_rate": 1.861053262258067e-05,
      "loss": 0.4828,
      "step": 701
    },
    {
      "epoch": 2.8435443037974686,
      "grad_norm": 0.2888655252666192,
      "learning_rate": 1.8553916758952897e-05,
      "loss": 0.4845,
      "step": 702
    },
    {
      "epoch": 2.8475949367088607,
      "grad_norm": 0.2667284271146545,
      "learning_rate": 1.8497312541880735e-05,
      "loss": 0.4886,
      "step": 703
    },
    {
      "epoch": 2.851645569620253,
      "grad_norm": 1.205661023018657,
      "learning_rate": 1.8440720427246786e-05,
      "loss": 0.4866,
      "step": 704
    },
    {
      "epoch": 2.8556962025316457,
      "grad_norm": 0.26793782091307466,
      "learning_rate": 1.8384140870836157e-05,
      "loss": 0.4826,
      "step": 705
    },
    {
      "epoch": 2.8597468354430378,
      "grad_norm": 0.2646774200420843,
      "learning_rate": 1.8327574328332806e-05,
      "loss": 0.4757,
      "step": 706
    },
    {
      "epoch": 2.8637974683544303,
      "grad_norm": 0.29511502110860743,
      "learning_rate": 1.8271021255315906e-05,
      "loss": 0.4854,
      "step": 707
    },
    {
      "epoch": 2.867848101265823,
      "grad_norm": 0.2620343175837188,
      "learning_rate": 1.8214482107256117e-05,
      "loss": 0.4751,
      "step": 708
    },
    {
      "epoch": 2.8718987341772153,
      "grad_norm": 0.31184893434331473,
      "learning_rate": 1.8157957339511968e-05,
      "loss": 0.4848,
      "step": 709
    },
    {
      "epoch": 2.875949367088608,
      "grad_norm": 0.2738927015664275,
      "learning_rate": 1.8101447407326182e-05,
      "loss": 0.4906,
      "step": 710
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.2884356108459456,
      "learning_rate": 1.8044952765821966e-05,
      "loss": 0.4837,
      "step": 711
    },
    {
      "epoch": 2.8840506329113924,
      "grad_norm": 0.2854217826336115,
      "learning_rate": 1.7988473869999407e-05,
      "loss": 0.4894,
      "step": 712
    },
    {
      "epoch": 2.888101265822785,
      "grad_norm": 0.29537569060405305,
      "learning_rate": 1.7932011174731764e-05,
      "loss": 0.4871,
      "step": 713
    },
    {
      "epoch": 2.892151898734177,
      "grad_norm": 0.32964839327771983,
      "learning_rate": 1.7875565134761817e-05,
      "loss": 0.4778,
      "step": 714
    },
    {
      "epoch": 2.8962025316455695,
      "grad_norm": 0.27102480455772054,
      "learning_rate": 1.7819136204698226e-05,
      "loss": 0.4821,
      "step": 715
    },
    {
      "epoch": 2.900253164556962,
      "grad_norm": 0.3005883176827633,
      "learning_rate": 1.776272483901182e-05,
      "loss": 0.4832,
      "step": 716
    },
    {
      "epoch": 2.9043037974683545,
      "grad_norm": 0.24401022395539515,
      "learning_rate": 1.7706331492031995e-05,
      "loss": 0.4838,
      "step": 717
    },
    {
      "epoch": 2.908354430379747,
      "grad_norm": 0.26807330513533106,
      "learning_rate": 1.764995661794301e-05,
      "loss": 0.4753,
      "step": 718
    },
    {
      "epoch": 2.912405063291139,
      "grad_norm": 0.2576020841164344,
      "learning_rate": 1.759360067078035e-05,
      "loss": 0.4924,
      "step": 719
    },
    {
      "epoch": 2.9164556962025316,
      "grad_norm": 0.345016955109111,
      "learning_rate": 1.7537264104427064e-05,
      "loss": 0.4814,
      "step": 720
    },
    {
      "epoch": 2.920506329113924,
      "grad_norm": 0.26465818894064275,
      "learning_rate": 1.748094737261012e-05,
      "loss": 0.4831,
      "step": 721
    },
    {
      "epoch": 2.9245569620253162,
      "grad_norm": 0.2831617282303903,
      "learning_rate": 1.7424650928896726e-05,
      "loss": 0.4867,
      "step": 722
    },
    {
      "epoch": 2.9286075949367087,
      "grad_norm": 0.24040862403016286,
      "learning_rate": 1.7368375226690712e-05,
      "loss": 0.4754,
      "step": 723
    },
    {
      "epoch": 2.9326582278481013,
      "grad_norm": 0.2932652997410923,
      "learning_rate": 1.731212071922883e-05,
      "loss": 0.5016,
      "step": 724
    },
    {
      "epoch": 2.9367088607594938,
      "grad_norm": 0.26783507858405764,
      "learning_rate": 1.7255887859577156e-05,
      "loss": 0.4754,
      "step": 725
    },
    {
      "epoch": 2.9407594936708863,
      "grad_norm": 0.2889945457425256,
      "learning_rate": 1.7199677100627427e-05,
      "loss": 0.4866,
      "step": 726
    },
    {
      "epoch": 2.9448101265822784,
      "grad_norm": 0.49287349000329683,
      "learning_rate": 1.7143488895093343e-05,
      "loss": 0.4882,
      "step": 727
    },
    {
      "epoch": 2.948860759493671,
      "grad_norm": 0.3091881003280438,
      "learning_rate": 1.7087323695506994e-05,
      "loss": 0.491,
      "step": 728
    },
    {
      "epoch": 2.9529113924050634,
      "grad_norm": 0.28505001013981873,
      "learning_rate": 1.7031181954215194e-05,
      "loss": 0.4856,
      "step": 729
    },
    {
      "epoch": 2.9569620253164555,
      "grad_norm": 0.2837269166738964,
      "learning_rate": 1.6975064123375788e-05,
      "loss": 0.4831,
      "step": 730
    },
    {
      "epoch": 2.961012658227848,
      "grad_norm": 0.2843811725285847,
      "learning_rate": 1.6918970654954084e-05,
      "loss": 0.4721,
      "step": 731
    },
    {
      "epoch": 2.9650632911392405,
      "grad_norm": 0.2664091391521519,
      "learning_rate": 1.686290200071915e-05,
      "loss": 0.4862,
      "step": 732
    },
    {
      "epoch": 2.969113924050633,
      "grad_norm": 0.27908496997583826,
      "learning_rate": 1.6806858612240234e-05,
      "loss": 0.4691,
      "step": 733
    },
    {
      "epoch": 2.9731645569620255,
      "grad_norm": 0.9498068905256418,
      "learning_rate": 1.6750840940883078e-05,
      "loss": 0.4973,
      "step": 734
    },
    {
      "epoch": 2.9772151898734176,
      "grad_norm": 0.276676956166554,
      "learning_rate": 1.6694849437806305e-05,
      "loss": 0.4906,
      "step": 735
    },
    {
      "epoch": 2.98126582278481,
      "grad_norm": 0.2593799756335559,
      "learning_rate": 1.663888455395778e-05,
      "loss": 0.4933,
      "step": 736
    },
    {
      "epoch": 2.9853164556962026,
      "grad_norm": 0.25268214382091786,
      "learning_rate": 1.6582946740070995e-05,
      "loss": 0.4767,
      "step": 737
    },
    {
      "epoch": 2.9893670886075947,
      "grad_norm": 0.26390462079042987,
      "learning_rate": 1.6527036446661396e-05,
      "loss": 0.4779,
      "step": 738
    },
    {
      "epoch": 2.993417721518987,
      "grad_norm": 0.3139703410031718,
      "learning_rate": 1.6471154124022818e-05,
      "loss": 0.4905,
      "step": 739
    },
    {
      "epoch": 2.9974683544303797,
      "grad_norm": 0.282338486050531,
      "learning_rate": 1.6415300222223788e-05,
      "loss": 0.4997,
      "step": 740
    },
    {
      "epoch": 3.012151898734177,
      "grad_norm": 0.46271056247070913,
      "learning_rate": 1.6359475191103958e-05,
      "loss": 0.411,
      "step": 741
    },
    {
      "epoch": 3.0162025316455696,
      "grad_norm": 0.3769785900926335,
      "learning_rate": 1.6303679480270466e-05,
      "loss": 0.4151,
      "step": 742
    },
    {
      "epoch": 3.020253164556962,
      "grad_norm": 0.6616417062407739,
      "learning_rate": 1.624791353909428e-05,
      "loss": 0.439,
      "step": 743
    },
    {
      "epoch": 3.0243037974683546,
      "grad_norm": 0.46514747763161207,
      "learning_rate": 1.619217781670663e-05,
      "loss": 0.4179,
      "step": 744
    },
    {
      "epoch": 3.0283544303797467,
      "grad_norm": 0.47452127716919495,
      "learning_rate": 1.6136472761995373e-05,
      "loss": 0.4219,
      "step": 745
    },
    {
      "epoch": 3.0324050632911392,
      "grad_norm": 0.4040789609172518,
      "learning_rate": 1.608079882360134e-05,
      "loss": 0.4193,
      "step": 746
    },
    {
      "epoch": 3.0364556962025318,
      "grad_norm": 0.4152422587799311,
      "learning_rate": 1.60251564499148e-05,
      "loss": 0.4359,
      "step": 747
    },
    {
      "epoch": 3.0405063291139243,
      "grad_norm": 0.45460022463133837,
      "learning_rate": 1.596954608907176e-05,
      "loss": 0.4229,
      "step": 748
    },
    {
      "epoch": 3.0445569620253163,
      "grad_norm": 0.39374878080608744,
      "learning_rate": 1.591396818895043e-05,
      "loss": 0.4281,
      "step": 749
    },
    {
      "epoch": 3.048607594936709,
      "grad_norm": 0.6641917273564537,
      "learning_rate": 1.585842319716759e-05,
      "loss": 0.4318,
      "step": 750
    },
    {
      "epoch": 3.0526582278481014,
      "grad_norm": 0.43760270925287303,
      "learning_rate": 1.5802911561074944e-05,
      "loss": 0.4178,
      "step": 751
    },
    {
      "epoch": 3.056708860759494,
      "grad_norm": 0.37833207000635005,
      "learning_rate": 1.5747433727755595e-05,
      "loss": 0.4276,
      "step": 752
    },
    {
      "epoch": 3.060759493670886,
      "grad_norm": 0.40177551596405753,
      "learning_rate": 1.5691990144020376e-05,
      "loss": 0.4122,
      "step": 753
    },
    {
      "epoch": 3.0648101265822785,
      "grad_norm": 0.3759581521668601,
      "learning_rate": 1.5636581256404297e-05,
      "loss": 0.4314,
      "step": 754
    },
    {
      "epoch": 3.068860759493671,
      "grad_norm": 0.3813527622443326,
      "learning_rate": 1.558120751116291e-05,
      "loss": 0.4228,
      "step": 755
    },
    {
      "epoch": 3.0729113924050635,
      "grad_norm": 2.0340128093818564,
      "learning_rate": 1.552586935426876e-05,
      "loss": 0.4238,
      "step": 756
    },
    {
      "epoch": 3.0769620253164556,
      "grad_norm": 0.4112959676796039,
      "learning_rate": 1.547056723140774e-05,
      "loss": 0.4204,
      "step": 757
    },
    {
      "epoch": 3.081012658227848,
      "grad_norm": 0.33583033878991936,
      "learning_rate": 1.5415301587975565e-05,
      "loss": 0.4308,
      "step": 758
    },
    {
      "epoch": 3.0850632911392406,
      "grad_norm": 0.42444723679105806,
      "learning_rate": 1.536007286907411e-05,
      "loss": 0.4153,
      "step": 759
    },
    {
      "epoch": 3.089113924050633,
      "grad_norm": 0.3223797717938031,
      "learning_rate": 1.5304881519507896e-05,
      "loss": 0.4082,
      "step": 760
    },
    {
      "epoch": 3.093164556962025,
      "grad_norm": 0.34442996407533544,
      "learning_rate": 1.5249727983780453e-05,
      "loss": 0.4339,
      "step": 761
    },
    {
      "epoch": 3.0972151898734177,
      "grad_norm": 0.35436722408268023,
      "learning_rate": 1.5194612706090786e-05,
      "loss": 0.4219,
      "step": 762
    },
    {
      "epoch": 3.1012658227848102,
      "grad_norm": 0.34119128418318,
      "learning_rate": 1.5139536130329771e-05,
      "loss": 0.4316,
      "step": 763
    },
    {
      "epoch": 3.1053164556962027,
      "grad_norm": 0.3603358310974989,
      "learning_rate": 1.508449870007656e-05,
      "loss": 0.4323,
      "step": 764
    },
    {
      "epoch": 3.109367088607595,
      "grad_norm": 0.3820832682637235,
      "learning_rate": 1.5029500858595056e-05,
      "loss": 0.4155,
      "step": 765
    },
    {
      "epoch": 3.1134177215189873,
      "grad_norm": 0.3552203130326848,
      "learning_rate": 1.4974543048830328e-05,
      "loss": 0.4136,
      "step": 766
    },
    {
      "epoch": 3.11746835443038,
      "grad_norm": 0.41119283656818006,
      "learning_rate": 1.4919625713405e-05,
      "loss": 0.4244,
      "step": 767
    },
    {
      "epoch": 3.1215189873417724,
      "grad_norm": 0.3445746292273152,
      "learning_rate": 1.4864749294615756e-05,
      "loss": 0.4185,
      "step": 768
    },
    {
      "epoch": 3.1255696202531644,
      "grad_norm": 0.3503949321870011,
      "learning_rate": 1.4809914234429716e-05,
      "loss": 0.4176,
      "step": 769
    },
    {
      "epoch": 3.129620253164557,
      "grad_norm": 0.2933933013756213,
      "learning_rate": 1.4755120974480923e-05,
      "loss": 0.4438,
      "step": 770
    },
    {
      "epoch": 3.1336708860759495,
      "grad_norm": 0.32760914197922003,
      "learning_rate": 1.4700369956066771e-05,
      "loss": 0.4119,
      "step": 771
    },
    {
      "epoch": 3.137721518987342,
      "grad_norm": 0.34982884796860414,
      "learning_rate": 1.4645661620144413e-05,
      "loss": 0.4157,
      "step": 772
    },
    {
      "epoch": 3.141772151898734,
      "grad_norm": 0.29184947323952015,
      "learning_rate": 1.4590996407327284e-05,
      "loss": 0.4299,
      "step": 773
    },
    {
      "epoch": 3.1458227848101266,
      "grad_norm": 0.36898025301046405,
      "learning_rate": 1.4536374757881487e-05,
      "loss": 0.4325,
      "step": 774
    },
    {
      "epoch": 3.149873417721519,
      "grad_norm": 0.4471971936272371,
      "learning_rate": 1.4481797111722271e-05,
      "loss": 0.4318,
      "step": 775
    },
    {
      "epoch": 3.1539240506329116,
      "grad_norm": 0.33202223615600995,
      "learning_rate": 1.4427263908410507e-05,
      "loss": 0.4404,
      "step": 776
    },
    {
      "epoch": 3.1579746835443037,
      "grad_norm": 0.370401601735636,
      "learning_rate": 1.4372775587149108e-05,
      "loss": 0.4297,
      "step": 777
    },
    {
      "epoch": 3.162025316455696,
      "grad_norm": 0.31409686188013625,
      "learning_rate": 1.4318332586779522e-05,
      "loss": 0.4309,
      "step": 778
    },
    {
      "epoch": 3.1660759493670887,
      "grad_norm": 0.34080870386782286,
      "learning_rate": 1.4263935345778202e-05,
      "loss": 0.4232,
      "step": 779
    },
    {
      "epoch": 3.170126582278481,
      "grad_norm": 0.2665867742795507,
      "learning_rate": 1.420958430225303e-05,
      "loss": 0.4267,
      "step": 780
    },
    {
      "epoch": 3.1741772151898733,
      "grad_norm": 0.32757826027708165,
      "learning_rate": 1.415527989393985e-05,
      "loss": 0.4181,
      "step": 781
    },
    {
      "epoch": 3.178227848101266,
      "grad_norm": 0.28368647884645404,
      "learning_rate": 1.410102255819891e-05,
      "loss": 0.4105,
      "step": 782
    },
    {
      "epoch": 3.1822784810126583,
      "grad_norm": 0.3035800439932441,
      "learning_rate": 1.404681273201131e-05,
      "loss": 0.4367,
      "step": 783
    },
    {
      "epoch": 3.186329113924051,
      "grad_norm": 0.29285947710968113,
      "learning_rate": 1.399265085197556e-05,
      "loss": 0.4099,
      "step": 784
    },
    {
      "epoch": 3.190379746835443,
      "grad_norm": 0.2578394088659349,
      "learning_rate": 1.393853735430398e-05,
      "loss": 0.4299,
      "step": 785
    },
    {
      "epoch": 3.1944303797468354,
      "grad_norm": 0.2893786486553056,
      "learning_rate": 1.3884472674819246e-05,
      "loss": 0.4205,
      "step": 786
    },
    {
      "epoch": 3.198481012658228,
      "grad_norm": 0.2660751157776481,
      "learning_rate": 1.3830457248950864e-05,
      "loss": 0.4222,
      "step": 787
    },
    {
      "epoch": 3.2025316455696204,
      "grad_norm": 0.2696844745045692,
      "learning_rate": 1.377649151173163e-05,
      "loss": 0.4273,
      "step": 788
    },
    {
      "epoch": 3.2065822784810125,
      "grad_norm": 0.26688314102680166,
      "learning_rate": 1.3722575897794181e-05,
      "loss": 0.4271,
      "step": 789
    },
    {
      "epoch": 3.210632911392405,
      "grad_norm": 0.2685500782671214,
      "learning_rate": 1.3668710841367472e-05,
      "loss": 0.4134,
      "step": 790
    },
    {
      "epoch": 3.2146835443037975,
      "grad_norm": 0.2602075037403734,
      "learning_rate": 1.361489677627324e-05,
      "loss": 0.4171,
      "step": 791
    },
    {
      "epoch": 3.21873417721519,
      "grad_norm": 0.26051790766595945,
      "learning_rate": 1.3561134135922585e-05,
      "loss": 0.4098,
      "step": 792
    },
    {
      "epoch": 3.222784810126582,
      "grad_norm": 0.2560972178175247,
      "learning_rate": 1.350742335331241e-05,
      "loss": 0.4204,
      "step": 793
    },
    {
      "epoch": 3.2268354430379746,
      "grad_norm": 0.27267021865241636,
      "learning_rate": 1.345376486102198e-05,
      "loss": 0.4184,
      "step": 794
    },
    {
      "epoch": 3.230886075949367,
      "grad_norm": 0.2651104137845879,
      "learning_rate": 1.3400159091209414e-05,
      "loss": 0.4041,
      "step": 795
    },
    {
      "epoch": 3.2349367088607597,
      "grad_norm": 0.2490642333447309,
      "learning_rate": 1.3346606475608216e-05,
      "loss": 0.4282,
      "step": 796
    },
    {
      "epoch": 3.2389873417721518,
      "grad_norm": 0.26531888517378727,
      "learning_rate": 1.3293107445523781e-05,
      "loss": 0.4171,
      "step": 797
    },
    {
      "epoch": 3.2430379746835443,
      "grad_norm": 0.31118121326070813,
      "learning_rate": 1.3239662431829949e-05,
      "loss": 0.4175,
      "step": 798
    },
    {
      "epoch": 3.247088607594937,
      "grad_norm": 0.24079594369390647,
      "learning_rate": 1.3186271864965509e-05,
      "loss": 0.41,
      "step": 799
    },
    {
      "epoch": 3.2511392405063293,
      "grad_norm": 0.2902619776182376,
      "learning_rate": 1.3132936174930756e-05,
      "loss": 0.4071,
      "step": 800
    },
    {
      "epoch": 3.2551898734177214,
      "grad_norm": 0.2983835784396681,
      "learning_rate": 1.3079655791283995e-05,
      "loss": 0.4299,
      "step": 801
    },
    {
      "epoch": 3.259240506329114,
      "grad_norm": 0.23858029167967648,
      "learning_rate": 1.3026431143138108e-05,
      "loss": 0.4191,
      "step": 802
    },
    {
      "epoch": 3.2632911392405064,
      "grad_norm": 0.2943146108746643,
      "learning_rate": 1.2973262659157114e-05,
      "loss": 0.4277,
      "step": 803
    },
    {
      "epoch": 3.267341772151899,
      "grad_norm": 0.2851968746927365,
      "learning_rate": 1.2920150767552651e-05,
      "loss": 0.4069,
      "step": 804
    },
    {
      "epoch": 3.271392405063291,
      "grad_norm": 0.23498131439642939,
      "learning_rate": 1.2867095896080607e-05,
      "loss": 0.4218,
      "step": 805
    },
    {
      "epoch": 3.2754430379746835,
      "grad_norm": 0.2621601998024635,
      "learning_rate": 1.2814098472037612e-05,
      "loss": 0.4216,
      "step": 806
    },
    {
      "epoch": 3.279493670886076,
      "grad_norm": 0.2377669877362962,
      "learning_rate": 1.276115892225764e-05,
      "loss": 0.426,
      "step": 807
    },
    {
      "epoch": 3.2835443037974685,
      "grad_norm": 0.23152531731899248,
      "learning_rate": 1.2708277673108555e-05,
      "loss": 0.4162,
      "step": 808
    },
    {
      "epoch": 3.2875949367088606,
      "grad_norm": 0.2786823661860218,
      "learning_rate": 1.2655455150488649e-05,
      "loss": 0.411,
      "step": 809
    },
    {
      "epoch": 3.291645569620253,
      "grad_norm": 0.2557489178777418,
      "learning_rate": 1.2602691779823272e-05,
      "loss": 0.4203,
      "step": 810
    },
    {
      "epoch": 3.2956962025316456,
      "grad_norm": 0.2489135376030936,
      "learning_rate": 1.2549987986061355e-05,
      "loss": 0.4117,
      "step": 811
    },
    {
      "epoch": 3.299746835443038,
      "grad_norm": 0.24689005388548824,
      "learning_rate": 1.2497344193672005e-05,
      "loss": 0.4215,
      "step": 812
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.23580251736890603,
      "learning_rate": 1.2444760826641092e-05,
      "loss": 0.4321,
      "step": 813
    },
    {
      "epoch": 3.3078481012658227,
      "grad_norm": 0.2479709566572627,
      "learning_rate": 1.2392238308467817e-05,
      "loss": 0.4247,
      "step": 814
    },
    {
      "epoch": 3.3118987341772153,
      "grad_norm": 0.2326504578856396,
      "learning_rate": 1.2339777062161326e-05,
      "loss": 0.4229,
      "step": 815
    },
    {
      "epoch": 3.3159493670886078,
      "grad_norm": 0.2337149635367473,
      "learning_rate": 1.2287377510237293e-05,
      "loss": 0.4215,
      "step": 816
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.2567508772311241,
      "learning_rate": 1.2235040074714488e-05,
      "loss": 0.4053,
      "step": 817
    },
    {
      "epoch": 3.3240506329113924,
      "grad_norm": 0.2433977903324384,
      "learning_rate": 1.2182765177111434e-05,
      "loss": 0.4176,
      "step": 818
    },
    {
      "epoch": 3.328101265822785,
      "grad_norm": 0.24083852521639906,
      "learning_rate": 1.213055323844297e-05,
      "loss": 0.4028,
      "step": 819
    },
    {
      "epoch": 3.3321518987341774,
      "grad_norm": 0.23528648701087132,
      "learning_rate": 1.2078404679216864e-05,
      "loss": 0.4309,
      "step": 820
    },
    {
      "epoch": 3.3362025316455695,
      "grad_norm": 0.2800352726064461,
      "learning_rate": 1.2026319919430458e-05,
      "loss": 0.4082,
      "step": 821
    },
    {
      "epoch": 3.340253164556962,
      "grad_norm": 0.22931776297569073,
      "learning_rate": 1.1974299378567227e-05,
      "loss": 0.4325,
      "step": 822
    },
    {
      "epoch": 3.3443037974683545,
      "grad_norm": 0.2828620018047522,
      "learning_rate": 1.1922343475593462e-05,
      "loss": 0.4313,
      "step": 823
    },
    {
      "epoch": 3.348354430379747,
      "grad_norm": 0.2559634666934956,
      "learning_rate": 1.187045262895488e-05,
      "loss": 0.426,
      "step": 824
    },
    {
      "epoch": 3.352405063291139,
      "grad_norm": 0.2572190056059979,
      "learning_rate": 1.1818627256573203e-05,
      "loss": 0.4298,
      "step": 825
    },
    {
      "epoch": 3.3564556962025316,
      "grad_norm": 0.2570334103316047,
      "learning_rate": 1.1766867775842864e-05,
      "loss": 0.4157,
      "step": 826
    },
    {
      "epoch": 3.360506329113924,
      "grad_norm": 0.2708721722303622,
      "learning_rate": 1.1715174603627615e-05,
      "loss": 0.4228,
      "step": 827
    },
    {
      "epoch": 3.3645569620253166,
      "grad_norm": 1.0896679231606945,
      "learning_rate": 1.1663548156257147e-05,
      "loss": 0.4241,
      "step": 828
    },
    {
      "epoch": 3.3686075949367087,
      "grad_norm": 0.25463415716414484,
      "learning_rate": 1.161198884952377e-05,
      "loss": 0.414,
      "step": 829
    },
    {
      "epoch": 3.372658227848101,
      "grad_norm": 0.2477953419195249,
      "learning_rate": 1.1560497098679056e-05,
      "loss": 0.4362,
      "step": 830
    },
    {
      "epoch": 3.3767088607594937,
      "grad_norm": 0.24159114455478997,
      "learning_rate": 1.1509073318430479e-05,
      "loss": 0.423,
      "step": 831
    },
    {
      "epoch": 3.3807594936708862,
      "grad_norm": 0.25846328891063225,
      "learning_rate": 1.1457717922938116e-05,
      "loss": 0.4238,
      "step": 832
    },
    {
      "epoch": 3.3848101265822783,
      "grad_norm": 0.24597648116450482,
      "learning_rate": 1.1406431325811233e-05,
      "loss": 0.436,
      "step": 833
    },
    {
      "epoch": 3.388860759493671,
      "grad_norm": 0.23965766702065666,
      "learning_rate": 1.135521394010506e-05,
      "loss": 0.4196,
      "step": 834
    },
    {
      "epoch": 3.3929113924050633,
      "grad_norm": 0.2455040846014678,
      "learning_rate": 1.1304066178317367e-05,
      "loss": 0.419,
      "step": 835
    },
    {
      "epoch": 3.396962025316456,
      "grad_norm": 0.24099022361769915,
      "learning_rate": 1.1252988452385199e-05,
      "loss": 0.4072,
      "step": 836
    },
    {
      "epoch": 3.401012658227848,
      "grad_norm": 0.24448484103093243,
      "learning_rate": 1.1201981173681536e-05,
      "loss": 0.4201,
      "step": 837
    },
    {
      "epoch": 3.4050632911392404,
      "grad_norm": 0.2379758445952789,
      "learning_rate": 1.1151044753011991e-05,
      "loss": 0.4176,
      "step": 838
    },
    {
      "epoch": 3.409113924050633,
      "grad_norm": 0.23881676781525452,
      "learning_rate": 1.1100179600611491e-05,
      "loss": 0.4374,
      "step": 839
    },
    {
      "epoch": 3.4131645569620255,
      "grad_norm": 0.23098992974927474,
      "learning_rate": 1.1049386126140985e-05,
      "loss": 0.4205,
      "step": 840
    },
    {
      "epoch": 3.4172151898734175,
      "grad_norm": 0.23074277706368596,
      "learning_rate": 1.0998664738684128e-05,
      "loss": 0.4113,
      "step": 841
    },
    {
      "epoch": 3.42126582278481,
      "grad_norm": 0.23322694506393277,
      "learning_rate": 1.0948015846744e-05,
      "loss": 0.4179,
      "step": 842
    },
    {
      "epoch": 3.4253164556962026,
      "grad_norm": 0.23386947456834106,
      "learning_rate": 1.0897439858239832e-05,
      "loss": 0.4348,
      "step": 843
    },
    {
      "epoch": 3.429367088607595,
      "grad_norm": 0.30240774299787954,
      "learning_rate": 1.0846937180503652e-05,
      "loss": 0.4264,
      "step": 844
    },
    {
      "epoch": 3.433417721518987,
      "grad_norm": 0.3494401598055955,
      "learning_rate": 1.0796508220277117e-05,
      "loss": 0.4201,
      "step": 845
    },
    {
      "epoch": 3.4374683544303797,
      "grad_norm": 0.3009827812867687,
      "learning_rate": 1.0746153383708107e-05,
      "loss": 0.4204,
      "step": 846
    },
    {
      "epoch": 3.441518987341772,
      "grad_norm": 0.2366114582985829,
      "learning_rate": 1.0695873076347579e-05,
      "loss": 0.4302,
      "step": 847
    },
    {
      "epoch": 3.4455696202531647,
      "grad_norm": 0.24069963304279532,
      "learning_rate": 1.0645667703146205e-05,
      "loss": 0.4058,
      "step": 848
    },
    {
      "epoch": 3.449620253164557,
      "grad_norm": 0.2510415642162749,
      "learning_rate": 1.0595537668451161e-05,
      "loss": 0.4178,
      "step": 849
    },
    {
      "epoch": 3.4536708860759493,
      "grad_norm": 0.23411466415042598,
      "learning_rate": 1.0545483376002854e-05,
      "loss": 0.4186,
      "step": 850
    },
    {
      "epoch": 3.457721518987342,
      "grad_norm": 0.2321852794317087,
      "learning_rate": 1.0495505228931676e-05,
      "loss": 0.4117,
      "step": 851
    },
    {
      "epoch": 3.4617721518987343,
      "grad_norm": 0.23505093207693728,
      "learning_rate": 1.044560362975474e-05,
      "loss": 0.4163,
      "step": 852
    },
    {
      "epoch": 3.4658227848101264,
      "grad_norm": 0.22583485497974692,
      "learning_rate": 1.0395778980372695e-05,
      "loss": 0.4229,
      "step": 853
    },
    {
      "epoch": 3.469873417721519,
      "grad_norm": 0.23981870913744224,
      "learning_rate": 1.0346031682066381e-05,
      "loss": 0.4153,
      "step": 854
    },
    {
      "epoch": 3.4739240506329114,
      "grad_norm": 0.2243794154338694,
      "learning_rate": 1.0296362135493724e-05,
      "loss": 0.419,
      "step": 855
    },
    {
      "epoch": 3.477974683544304,
      "grad_norm": 0.24366240649991436,
      "learning_rate": 1.0246770740686422e-05,
      "loss": 0.4367,
      "step": 856
    },
    {
      "epoch": 3.482025316455696,
      "grad_norm": 0.26199017408586867,
      "learning_rate": 1.0197257897046743e-05,
      "loss": 0.4263,
      "step": 857
    },
    {
      "epoch": 3.4860759493670885,
      "grad_norm": 0.2299882592651752,
      "learning_rate": 1.014782400334433e-05,
      "loss": 0.4232,
      "step": 858
    },
    {
      "epoch": 3.490126582278481,
      "grad_norm": 0.2573938610767819,
      "learning_rate": 1.009846945771296e-05,
      "loss": 0.4381,
      "step": 859
    },
    {
      "epoch": 3.4941772151898736,
      "grad_norm": 0.24588768481105688,
      "learning_rate": 1.0049194657647363e-05,
      "loss": 0.4173,
      "step": 860
    },
    {
      "epoch": 3.4982278481012656,
      "grad_norm": 0.22634947824150042,
      "learning_rate": 1.0000000000000006e-05,
      "loss": 0.419,
      "step": 861
    },
    {
      "epoch": 3.502278481012658,
      "grad_norm": 0.24279115059750403,
      "learning_rate": 9.950885880977891e-06,
      "loss": 0.4255,
      "step": 862
    },
    {
      "epoch": 3.5063291139240507,
      "grad_norm": 0.23871677871906605,
      "learning_rate": 9.901852696139382e-06,
      "loss": 0.41,
      "step": 863
    },
    {
      "epoch": 3.510379746835443,
      "grad_norm": 0.21447194498076322,
      "learning_rate": 9.852900840391027e-06,
      "loss": 0.427,
      "step": 864
    },
    {
      "epoch": 3.5144303797468357,
      "grad_norm": 0.22213194672768355,
      "learning_rate": 9.804030707984313e-06,
      "loss": 0.4282,
      "step": 865
    },
    {
      "epoch": 3.5184810126582278,
      "grad_norm": 0.25553569900242534,
      "learning_rate": 9.755242692512599e-06,
      "loss": 0.4348,
      "step": 866
    },
    {
      "epoch": 3.5225316455696203,
      "grad_norm": 0.2194646296025119,
      "learning_rate": 9.70653718690782e-06,
      "loss": 0.4261,
      "step": 867
    },
    {
      "epoch": 3.526582278481013,
      "grad_norm": 0.2348225958900731,
      "learning_rate": 9.657914583437454e-06,
      "loss": 0.4129,
      "step": 868
    },
    {
      "epoch": 3.530632911392405,
      "grad_norm": 0.3270151593035121,
      "learning_rate": 9.609375273701246e-06,
      "loss": 0.4217,
      "step": 869
    },
    {
      "epoch": 3.5346835443037974,
      "grad_norm": 0.23924194339916596,
      "learning_rate": 9.560919648628133e-06,
      "loss": 0.4156,
      "step": 870
    },
    {
      "epoch": 3.53873417721519,
      "grad_norm": 0.23434353641723846,
      "learning_rate": 9.512548098473047e-06,
      "loss": 0.433,
      "step": 871
    },
    {
      "epoch": 3.5427848101265824,
      "grad_norm": 0.22615926833173414,
      "learning_rate": 9.464261012813825e-06,
      "loss": 0.4221,
      "step": 872
    },
    {
      "epoch": 3.546835443037975,
      "grad_norm": 0.23146025977503246,
      "learning_rate": 9.416058780547987e-06,
      "loss": 0.4257,
      "step": 873
    },
    {
      "epoch": 3.550886075949367,
      "grad_norm": 0.23031124270298375,
      "learning_rate": 9.367941789889714e-06,
      "loss": 0.4184,
      "step": 874
    },
    {
      "epoch": 3.5549367088607595,
      "grad_norm": 0.24094870265328855,
      "learning_rate": 9.319910428366607e-06,
      "loss": 0.4274,
      "step": 875
    },
    {
      "epoch": 3.558987341772152,
      "grad_norm": 0.2440406571655205,
      "learning_rate": 9.271965082816667e-06,
      "loss": 0.4312,
      "step": 876
    },
    {
      "epoch": 3.563037974683544,
      "grad_norm": 0.22958472843991795,
      "learning_rate": 9.224106139385111e-06,
      "loss": 0.4136,
      "step": 877
    },
    {
      "epoch": 3.5670886075949366,
      "grad_norm": 0.2248178425596016,
      "learning_rate": 9.176333983521291e-06,
      "loss": 0.429,
      "step": 878
    },
    {
      "epoch": 3.571139240506329,
      "grad_norm": 0.2356770118609031,
      "learning_rate": 9.12864899997558e-06,
      "loss": 0.4178,
      "step": 879
    },
    {
      "epoch": 3.5751898734177217,
      "grad_norm": 0.23236335249799916,
      "learning_rate": 9.08105157279628e-06,
      "loss": 0.4173,
      "step": 880
    },
    {
      "epoch": 3.579240506329114,
      "grad_norm": 0.23326899759029968,
      "learning_rate": 9.03354208532653e-06,
      "loss": 0.413,
      "step": 881
    },
    {
      "epoch": 3.5832911392405062,
      "grad_norm": 0.23924953118271688,
      "learning_rate": 8.986120920201205e-06,
      "loss": 0.4315,
      "step": 882
    },
    {
      "epoch": 3.5873417721518988,
      "grad_norm": 0.22727108248315017,
      "learning_rate": 8.938788459343852e-06,
      "loss": 0.4132,
      "step": 883
    },
    {
      "epoch": 3.5913924050632913,
      "grad_norm": 0.23090546748849877,
      "learning_rate": 8.8915450839636e-06,
      "loss": 0.4135,
      "step": 884
    },
    {
      "epoch": 3.5954430379746833,
      "grad_norm": 0.23876347330272485,
      "learning_rate": 8.844391174552116e-06,
      "loss": 0.4258,
      "step": 885
    },
    {
      "epoch": 3.599493670886076,
      "grad_norm": 0.23101065631271356,
      "learning_rate": 8.797327110880479e-06,
      "loss": 0.4034,
      "step": 886
    },
    {
      "epoch": 3.6035443037974684,
      "grad_norm": 0.22906178154798792,
      "learning_rate": 8.750353271996206e-06,
      "loss": 0.4289,
      "step": 887
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.23800843086560475,
      "learning_rate": 8.703470036220132e-06,
      "loss": 0.4232,
      "step": 888
    },
    {
      "epoch": 3.6116455696202534,
      "grad_norm": 0.2503756580992034,
      "learning_rate": 8.656677781143394e-06,
      "loss": 0.4007,
      "step": 889
    },
    {
      "epoch": 3.6156962025316455,
      "grad_norm": 0.21692416162912184,
      "learning_rate": 8.609976883624377e-06,
      "loss": 0.4126,
      "step": 890
    },
    {
      "epoch": 3.619746835443038,
      "grad_norm": 0.22921901957711707,
      "learning_rate": 8.563367719785698e-06,
      "loss": 0.4331,
      "step": 891
    },
    {
      "epoch": 3.6237974683544305,
      "grad_norm": 0.25441682745295835,
      "learning_rate": 8.516850665011138e-06,
      "loss": 0.4145,
      "step": 892
    },
    {
      "epoch": 3.6278481012658226,
      "grad_norm": 0.22727839010996545,
      "learning_rate": 8.47042609394269e-06,
      "loss": 0.4179,
      "step": 893
    },
    {
      "epoch": 3.631898734177215,
      "grad_norm": 0.22481265132415884,
      "learning_rate": 8.424094380477432e-06,
      "loss": 0.4235,
      "step": 894
    },
    {
      "epoch": 3.6359493670886076,
      "grad_norm": 0.22884510758967747,
      "learning_rate": 8.37785589776465e-06,
      "loss": 0.426,
      "step": 895
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.22219982441099315,
      "learning_rate": 8.331711018202694e-06,
      "loss": 0.4192,
      "step": 896
    },
    {
      "epoch": 3.6440506329113926,
      "grad_norm": 0.2105750790172259,
      "learning_rate": 8.285660113436104e-06,
      "loss": 0.4348,
      "step": 897
    },
    {
      "epoch": 3.6481012658227847,
      "grad_norm": 0.23309689460788074,
      "learning_rate": 8.239703554352527e-06,
      "loss": 0.432,
      "step": 898
    },
    {
      "epoch": 3.6521518987341772,
      "grad_norm": 0.23611889341366202,
      "learning_rate": 8.193841711079775e-06,
      "loss": 0.4188,
      "step": 899
    },
    {
      "epoch": 3.6562025316455697,
      "grad_norm": 0.22267810696847404,
      "learning_rate": 8.148074952982828e-06,
      "loss": 0.4227,
      "step": 900
    },
    {
      "epoch": 3.660253164556962,
      "grad_norm": 0.25847713019847957,
      "learning_rate": 8.102403648660859e-06,
      "loss": 0.4232,
      "step": 901
    },
    {
      "epoch": 3.6643037974683543,
      "grad_norm": 0.21778817525262326,
      "learning_rate": 8.056828165944282e-06,
      "loss": 0.4201,
      "step": 902
    },
    {
      "epoch": 3.668354430379747,
      "grad_norm": 0.2103910720034378,
      "learning_rate": 8.011348871891762e-06,
      "loss": 0.4282,
      "step": 903
    },
    {
      "epoch": 3.6724050632911394,
      "grad_norm": 1.6735956646212027,
      "learning_rate": 7.965966132787287e-06,
      "loss": 0.4918,
      "step": 904
    },
    {
      "epoch": 3.676455696202532,
      "grad_norm": 0.2538843659406464,
      "learning_rate": 7.920680314137189e-06,
      "loss": 0.408,
      "step": 905
    },
    {
      "epoch": 3.680506329113924,
      "grad_norm": 0.22920359035664697,
      "learning_rate": 7.875491780667246e-06,
      "loss": 0.4252,
      "step": 906
    },
    {
      "epoch": 3.6845569620253165,
      "grad_norm": 0.23421256865475304,
      "learning_rate": 7.830400896319667e-06,
      "loss": 0.4327,
      "step": 907
    },
    {
      "epoch": 3.688607594936709,
      "grad_norm": 0.23618931790641112,
      "learning_rate": 7.785408024250259e-06,
      "loss": 0.4182,
      "step": 908
    },
    {
      "epoch": 3.692658227848101,
      "grad_norm": 0.230191315086708,
      "learning_rate": 7.74051352682542e-06,
      "loss": 0.4189,
      "step": 909
    },
    {
      "epoch": 3.6967088607594936,
      "grad_norm": 0.2382335315516925,
      "learning_rate": 7.695717765619257e-06,
      "loss": 0.4137,
      "step": 910
    },
    {
      "epoch": 3.700759493670886,
      "grad_norm": 0.2408392719392905,
      "learning_rate": 7.651021101410673e-06,
      "loss": 0.4268,
      "step": 911
    },
    {
      "epoch": 3.7048101265822786,
      "grad_norm": 0.22402491119210552,
      "learning_rate": 7.606423894180464e-06,
      "loss": 0.4228,
      "step": 912
    },
    {
      "epoch": 3.708860759493671,
      "grad_norm": 0.2689912990226994,
      "learning_rate": 7.56192650310839e-06,
      "loss": 0.404,
      "step": 913
    },
    {
      "epoch": 3.712911392405063,
      "grad_norm": 0.2377613431481652,
      "learning_rate": 7.517529286570349e-06,
      "loss": 0.4102,
      "step": 914
    },
    {
      "epoch": 3.7169620253164557,
      "grad_norm": 0.2154643191025324,
      "learning_rate": 7.473232602135387e-06,
      "loss": 0.4244,
      "step": 915
    },
    {
      "epoch": 3.721012658227848,
      "grad_norm": 0.23745058620404297,
      "learning_rate": 7.429036806562935e-06,
      "loss": 0.4176,
      "step": 916
    },
    {
      "epoch": 3.7250632911392403,
      "grad_norm": 0.254202058944723,
      "learning_rate": 7.3849422557998455e-06,
      "loss": 0.4049,
      "step": 917
    },
    {
      "epoch": 3.729113924050633,
      "grad_norm": 0.22246003964037758,
      "learning_rate": 7.340949304977567e-06,
      "loss": 0.4288,
      "step": 918
    },
    {
      "epoch": 3.7331645569620253,
      "grad_norm": 0.223945279415476,
      "learning_rate": 7.297058308409282e-06,
      "loss": 0.4296,
      "step": 919
    },
    {
      "epoch": 3.737215189873418,
      "grad_norm": 0.21650254324338156,
      "learning_rate": 7.25326961958704e-06,
      "loss": 0.4188,
      "step": 920
    },
    {
      "epoch": 3.7412658227848103,
      "grad_norm": 1.9775821826866777,
      "learning_rate": 7.209583591178921e-06,
      "loss": 0.4912,
      "step": 921
    },
    {
      "epoch": 3.7453164556962024,
      "grad_norm": 0.24980749721899653,
      "learning_rate": 7.1660005750261925e-06,
      "loss": 0.4266,
      "step": 922
    },
    {
      "epoch": 3.749367088607595,
      "grad_norm": 0.2178220813403478,
      "learning_rate": 7.1225209221404765e-06,
      "loss": 0.4233,
      "step": 923
    },
    {
      "epoch": 3.7534177215189874,
      "grad_norm": 0.23137539997532897,
      "learning_rate": 7.079144982700909e-06,
      "loss": 0.4055,
      "step": 924
    },
    {
      "epoch": 3.7574683544303795,
      "grad_norm": 0.24404562132281807,
      "learning_rate": 7.0358731060513695e-06,
      "loss": 0.4258,
      "step": 925
    },
    {
      "epoch": 3.761518987341772,
      "grad_norm": 0.2308821658568181,
      "learning_rate": 6.99270564069757e-06,
      "loss": 0.4286,
      "step": 926
    },
    {
      "epoch": 3.7655696202531646,
      "grad_norm": 0.24603206998301297,
      "learning_rate": 6.949642934304375e-06,
      "loss": 0.423,
      "step": 927
    },
    {
      "epoch": 3.769620253164557,
      "grad_norm": 0.24234276347412007,
      "learning_rate": 6.906685333692871e-06,
      "loss": 0.4173,
      "step": 928
    },
    {
      "epoch": 3.7736708860759496,
      "grad_norm": 0.23055012591471635,
      "learning_rate": 6.86383318483769e-06,
      "loss": 0.4073,
      "step": 929
    },
    {
      "epoch": 3.7777215189873417,
      "grad_norm": 0.2112327153657589,
      "learning_rate": 6.821086832864139e-06,
      "loss": 0.3993,
      "step": 930
    },
    {
      "epoch": 3.781772151898734,
      "grad_norm": 0.25047084963944427,
      "learning_rate": 6.77844662204546e-06,
      "loss": 0.4345,
      "step": 931
    },
    {
      "epoch": 3.7858227848101267,
      "grad_norm": 0.23165913097616259,
      "learning_rate": 6.7359128958000455e-06,
      "loss": 0.411,
      "step": 932
    },
    {
      "epoch": 3.7898734177215188,
      "grad_norm": 0.23832949844486742,
      "learning_rate": 6.693485996688695e-06,
      "loss": 0.4184,
      "step": 933
    },
    {
      "epoch": 3.7939240506329113,
      "grad_norm": 0.23535583034583993,
      "learning_rate": 6.651166266411801e-06,
      "loss": 0.4115,
      "step": 934
    },
    {
      "epoch": 3.797974683544304,
      "grad_norm": 0.24135943516118996,
      "learning_rate": 6.6089540458066725e-06,
      "loss": 0.4237,
      "step": 935
    },
    {
      "epoch": 3.8020253164556963,
      "grad_norm": 0.2519916506929559,
      "learning_rate": 6.566849674844711e-06,
      "loss": 0.4199,
      "step": 936
    },
    {
      "epoch": 3.806075949367089,
      "grad_norm": 0.2591247035552907,
      "learning_rate": 6.524853492628747e-06,
      "loss": 0.4253,
      "step": 937
    },
    {
      "epoch": 3.810126582278481,
      "grad_norm": 0.23400037038342505,
      "learning_rate": 6.4829658373902536e-06,
      "loss": 0.4296,
      "step": 938
    },
    {
      "epoch": 3.8141772151898734,
      "grad_norm": 0.2423697576001491,
      "learning_rate": 6.441187046486648e-06,
      "loss": 0.4166,
      "step": 939
    },
    {
      "epoch": 3.818227848101266,
      "grad_norm": 0.25604311424651016,
      "learning_rate": 6.399517456398567e-06,
      "loss": 0.4209,
      "step": 940
    },
    {
      "epoch": 3.822278481012658,
      "grad_norm": 0.21618490443177324,
      "learning_rate": 6.357957402727164e-06,
      "loss": 0.4084,
      "step": 941
    },
    {
      "epoch": 3.8263291139240505,
      "grad_norm": 0.22176989885677015,
      "learning_rate": 6.316507220191395e-06,
      "loss": 0.425,
      "step": 942
    },
    {
      "epoch": 3.830379746835443,
      "grad_norm": 0.2316102635959312,
      "learning_rate": 6.275167242625331e-06,
      "loss": 0.4223,
      "step": 943
    },
    {
      "epoch": 3.8344303797468355,
      "grad_norm": 0.2272011761352473,
      "learning_rate": 6.233937802975471e-06,
      "loss": 0.4271,
      "step": 944
    },
    {
      "epoch": 3.838481012658228,
      "grad_norm": 0.22925454573770132,
      "learning_rate": 6.192819233298046e-06,
      "loss": 0.397,
      "step": 945
    },
    {
      "epoch": 3.84253164556962,
      "grad_norm": 0.21160227757885955,
      "learning_rate": 6.151811864756383e-06,
      "loss": 0.445,
      "step": 946
    },
    {
      "epoch": 3.8465822784810126,
      "grad_norm": 0.22595578272305453,
      "learning_rate": 6.1109160276181655e-06,
      "loss": 0.4221,
      "step": 947
    },
    {
      "epoch": 3.850632911392405,
      "grad_norm": 0.24474355630509373,
      "learning_rate": 6.070132051252868e-06,
      "loss": 0.4262,
      "step": 948
    },
    {
      "epoch": 3.8546835443037972,
      "grad_norm": 0.22513450849772404,
      "learning_rate": 6.0294602641290034e-06,
      "loss": 0.4236,
      "step": 949
    },
    {
      "epoch": 3.8587341772151897,
      "grad_norm": 0.21892508992112145,
      "learning_rate": 5.988900993811575e-06,
      "loss": 0.4352,
      "step": 950
    },
    {
      "epoch": 3.8627848101265823,
      "grad_norm": 0.22818432626227747,
      "learning_rate": 5.948454566959363e-06,
      "loss": 0.4201,
      "step": 951
    },
    {
      "epoch": 3.8668354430379748,
      "grad_norm": 0.2387968891092459,
      "learning_rate": 5.908121309322328e-06,
      "loss": 0.4113,
      "step": 952
    },
    {
      "epoch": 3.8708860759493673,
      "grad_norm": 0.23003867088911048,
      "learning_rate": 5.867901545738976e-06,
      "loss": 0.4245,
      "step": 953
    },
    {
      "epoch": 3.8749367088607594,
      "grad_norm": 0.22473347468835092,
      "learning_rate": 5.827795600133774e-06,
      "loss": 0.4279,
      "step": 954
    },
    {
      "epoch": 3.878987341772152,
      "grad_norm": 0.23486963394810423,
      "learning_rate": 5.787803795514466e-06,
      "loss": 0.4329,
      "step": 955
    },
    {
      "epoch": 3.8830379746835444,
      "grad_norm": 0.24099716448780048,
      "learning_rate": 5.747926453969576e-06,
      "loss": 0.4162,
      "step": 956
    },
    {
      "epoch": 3.8870886075949365,
      "grad_norm": 0.22858020265308876,
      "learning_rate": 5.708163896665708e-06,
      "loss": 0.4417,
      "step": 957
    },
    {
      "epoch": 3.891139240506329,
      "grad_norm": 0.20865488209107805,
      "learning_rate": 5.668516443845047e-06,
      "loss": 0.4184,
      "step": 958
    },
    {
      "epoch": 3.8951898734177215,
      "grad_norm": 0.20680008946172895,
      "learning_rate": 5.6289844148227225e-06,
      "loss": 0.4336,
      "step": 959
    },
    {
      "epoch": 3.899240506329114,
      "grad_norm": 0.21563679569787086,
      "learning_rate": 5.5895681279842615e-06,
      "loss": 0.4164,
      "step": 960
    },
    {
      "epoch": 3.9032911392405065,
      "grad_norm": 0.2370873595903689,
      "learning_rate": 5.550267900783019e-06,
      "loss": 0.4314,
      "step": 961
    },
    {
      "epoch": 3.9073417721518986,
      "grad_norm": 0.21754567704438246,
      "learning_rate": 5.511084049737623e-06,
      "loss": 0.4274,
      "step": 962
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.21587131573865515,
      "learning_rate": 5.4720168904294215e-06,
      "loss": 0.4157,
      "step": 963
    },
    {
      "epoch": 3.9154430379746836,
      "grad_norm": 0.21708161250532598,
      "learning_rate": 5.433066737499948e-06,
      "loss": 0.4245,
      "step": 964
    },
    {
      "epoch": 3.9194936708860757,
      "grad_norm": 0.21877014431943853,
      "learning_rate": 5.394233904648376e-06,
      "loss": 0.4212,
      "step": 965
    },
    {
      "epoch": 3.923544303797468,
      "grad_norm": 0.2279027896976329,
      "learning_rate": 5.355518704628997e-06,
      "loss": 0.4339,
      "step": 966
    },
    {
      "epoch": 3.9275949367088607,
      "grad_norm": 0.2142607892432795,
      "learning_rate": 5.316921449248731e-06,
      "loss": 0.4424,
      "step": 967
    },
    {
      "epoch": 3.9316455696202532,
      "grad_norm": 0.2099445345723734,
      "learning_rate": 5.278442449364538e-06,
      "loss": 0.4107,
      "step": 968
    },
    {
      "epoch": 3.9356962025316458,
      "grad_norm": 0.22561584947937482,
      "learning_rate": 5.240082014881016e-06,
      "loss": 0.4258,
      "step": 969
    },
    {
      "epoch": 3.939746835443038,
      "grad_norm": 0.22509859237492635,
      "learning_rate": 5.201840454747822e-06,
      "loss": 0.4344,
      "step": 970
    },
    {
      "epoch": 3.9437974683544303,
      "grad_norm": 0.2073801207751303,
      "learning_rate": 5.163718076957223e-06,
      "loss": 0.4228,
      "step": 971
    },
    {
      "epoch": 3.947848101265823,
      "grad_norm": 0.2109994395148202,
      "learning_rate": 5.125715188541609e-06,
      "loss": 0.4291,
      "step": 972
    },
    {
      "epoch": 3.951898734177215,
      "grad_norm": 0.2008853328283793,
      "learning_rate": 5.087832095571021e-06,
      "loss": 0.4318,
      "step": 973
    },
    {
      "epoch": 3.9559493670886074,
      "grad_norm": 0.21223371675288188,
      "learning_rate": 5.0500691031506766e-06,
      "loss": 0.423,
      "step": 974
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.2107672709701527,
      "learning_rate": 5.01242651541854e-06,
      "loss": 0.4247,
      "step": 975
    },
    {
      "epoch": 3.9640506329113925,
      "grad_norm": 0.2108358469013863,
      "learning_rate": 4.974904635542815e-06,
      "loss": 0.4225,
      "step": 976
    },
    {
      "epoch": 3.968101265822785,
      "grad_norm": 0.21418235716938178,
      "learning_rate": 4.937503765719582e-06,
      "loss": 0.4171,
      "step": 977
    },
    {
      "epoch": 3.972151898734177,
      "grad_norm": 0.20926879118895375,
      "learning_rate": 4.900224207170299e-06,
      "loss": 0.4291,
      "step": 978
    },
    {
      "epoch": 3.9762025316455696,
      "grad_norm": 0.2238533311244171,
      "learning_rate": 4.8630662601394065e-06,
      "loss": 0.4191,
      "step": 979
    },
    {
      "epoch": 3.980253164556962,
      "grad_norm": 0.22161904360462842,
      "learning_rate": 4.8260302238918995e-06,
      "loss": 0.4178,
      "step": 980
    },
    {
      "epoch": 3.984303797468354,
      "grad_norm": 0.2295818340680339,
      "learning_rate": 4.789116396710924e-06,
      "loss": 0.4156,
      "step": 981
    },
    {
      "epoch": 3.9883544303797467,
      "grad_norm": 0.20934828656766138,
      "learning_rate": 4.752325075895368e-06,
      "loss": 0.4322,
      "step": 982
    },
    {
      "epoch": 3.992405063291139,
      "grad_norm": 0.19936081347292983,
      "learning_rate": 4.715656557757473e-06,
      "loss": 0.4124,
      "step": 983
    },
    {
      "epoch": 3.9964556962025317,
      "grad_norm": 0.21700762498765805,
      "learning_rate": 4.679111137620442e-06,
      "loss": 0.4062,
      "step": 984
    },
    {
      "epoch": 4.000506329113924,
      "grad_norm": 0.2243660503886236,
      "learning_rate": 4.6426891098160585e-06,
      "loss": 0.415,
      "step": 985
    },
    {
      "epoch": 4.004556962025316,
      "grad_norm": 0.3460916861867595,
      "learning_rate": 4.6063907676823474e-06,
      "loss": 0.4058,
      "step": 986
    },
    {
      "epoch": 4.008607594936709,
      "grad_norm": 0.31652914302672897,
      "learning_rate": 4.570216403561141e-06,
      "loss": 0.3731,
      "step": 987
    },
    {
      "epoch": 4.012658227848101,
      "grad_norm": 0.2370872982096638,
      "learning_rate": 4.534166308795815e-06,
      "loss": 0.3876,
      "step": 988
    },
    {
      "epoch": 4.016708860759493,
      "grad_norm": 0.3846196312455671,
      "learning_rate": 4.498240773728859e-06,
      "loss": 0.3815,
      "step": 989
    },
    {
      "epoch": 4.020759493670886,
      "grad_norm": 0.3865672690232331,
      "learning_rate": 4.462440087699609e-06,
      "loss": 0.3915,
      "step": 990
    },
    {
      "epoch": 4.024810126582278,
      "grad_norm": 0.27857736955207474,
      "learning_rate": 4.426764539041861e-06,
      "loss": 0.3726,
      "step": 991
    },
    {
      "epoch": 4.0288607594936705,
      "grad_norm": 0.28339561574407107,
      "learning_rate": 4.391214415081582e-06,
      "loss": 0.3674,
      "step": 992
    },
    {
      "epoch": 4.0329113924050635,
      "grad_norm": 0.3494118226780523,
      "learning_rate": 4.355790002134579e-06,
      "loss": 0.3809,
      "step": 993
    },
    {
      "epoch": 4.0369620253164555,
      "grad_norm": 0.32184716221405435,
      "learning_rate": 4.320491585504207e-06,
      "loss": 0.3694,
      "step": 994
    },
    {
      "epoch": 4.0410126582278485,
      "grad_norm": 0.24013643364441134,
      "learning_rate": 4.2853194494790615e-06,
      "loss": 0.39,
      "step": 995
    },
    {
      "epoch": 4.045063291139241,
      "grad_norm": 0.25865749589135845,
      "learning_rate": 4.250273877330691e-06,
      "loss": 0.3785,
      "step": 996
    },
    {
      "epoch": 4.049113924050633,
      "grad_norm": 0.3049640637221162,
      "learning_rate": 4.215355151311313e-06,
      "loss": 0.3865,
      "step": 997
    },
    {
      "epoch": 4.053164556962026,
      "grad_norm": 0.2848838869556188,
      "learning_rate": 4.180563552651542e-06,
      "loss": 0.3882,
      "step": 998
    },
    {
      "epoch": 4.057215189873418,
      "grad_norm": 0.24935288792120502,
      "learning_rate": 4.145899361558147e-06,
      "loss": 0.3944,
      "step": 999
    },
    {
      "epoch": 4.06126582278481,
      "grad_norm": 0.261813493251327,
      "learning_rate": 4.111362857211738e-06,
      "loss": 0.369,
      "step": 1000
    },
    {
      "epoch": 4.065316455696203,
      "grad_norm": 0.26435547988458036,
      "learning_rate": 4.076954317764592e-06,
      "loss": 0.3938,
      "step": 1001
    },
    {
      "epoch": 4.069367088607595,
      "grad_norm": 0.2628729838188856,
      "learning_rate": 4.042674020338335e-06,
      "loss": 0.3779,
      "step": 1002
    },
    {
      "epoch": 4.073417721518988,
      "grad_norm": 0.23163430244754457,
      "learning_rate": 4.0085222410217835e-06,
      "loss": 0.3978,
      "step": 1003
    },
    {
      "epoch": 4.07746835443038,
      "grad_norm": 0.22575688968828495,
      "learning_rate": 3.974499254868674e-06,
      "loss": 0.3695,
      "step": 1004
    },
    {
      "epoch": 4.081518987341772,
      "grad_norm": 0.25741948287492367,
      "learning_rate": 3.940605335895451e-06,
      "loss": 0.3905,
      "step": 1005
    },
    {
      "epoch": 4.085569620253165,
      "grad_norm": 0.23196544793945842,
      "learning_rate": 3.90684075707908e-06,
      "loss": 0.3706,
      "step": 1006
    },
    {
      "epoch": 4.089620253164557,
      "grad_norm": 0.21656550144656517,
      "learning_rate": 3.8732057903548505e-06,
      "loss": 0.378,
      "step": 1007
    },
    {
      "epoch": 4.093670886075949,
      "grad_norm": 0.23402703205037684,
      "learning_rate": 3.8397007066141375e-06,
      "loss": 0.3733,
      "step": 1008
    },
    {
      "epoch": 4.097721518987342,
      "grad_norm": 0.21949587355378522,
      "learning_rate": 3.806325775702304e-06,
      "loss": 0.3899,
      "step": 1009
    },
    {
      "epoch": 4.101772151898734,
      "grad_norm": 0.22952249311162207,
      "learning_rate": 3.773081266416434e-06,
      "loss": 0.386,
      "step": 1010
    },
    {
      "epoch": 4.105822784810127,
      "grad_norm": 0.21541626035582165,
      "learning_rate": 3.739967446503245e-06,
      "loss": 0.3738,
      "step": 1011
    },
    {
      "epoch": 4.109873417721519,
      "grad_norm": 0.21308486258687126,
      "learning_rate": 3.706984582656894e-06,
      "loss": 0.3764,
      "step": 1012
    },
    {
      "epoch": 4.113924050632911,
      "grad_norm": 0.22551361506293363,
      "learning_rate": 3.6741329405168237e-06,
      "loss": 0.3857,
      "step": 1013
    },
    {
      "epoch": 4.117974683544304,
      "grad_norm": 0.2230053772345858,
      "learning_rate": 3.641412784665648e-06,
      "loss": 0.365,
      "step": 1014
    },
    {
      "epoch": 4.122025316455696,
      "grad_norm": 0.21262979781773328,
      "learning_rate": 3.608824378627005e-06,
      "loss": 0.3601,
      "step": 1015
    },
    {
      "epoch": 4.126075949367088,
      "grad_norm": 1.8936166638168768,
      "learning_rate": 3.5763679848634337e-06,
      "loss": 0.4252,
      "step": 1016
    },
    {
      "epoch": 4.130126582278481,
      "grad_norm": 0.22569856563072507,
      "learning_rate": 3.544043864774269e-06,
      "loss": 0.3784,
      "step": 1017
    },
    {
      "epoch": 4.134177215189873,
      "grad_norm": 0.2317008621935235,
      "learning_rate": 3.5118522786935282e-06,
      "loss": 0.3796,
      "step": 1018
    },
    {
      "epoch": 4.138227848101266,
      "grad_norm": 0.20143088636148765,
      "learning_rate": 3.479793485887819e-06,
      "loss": 0.3894,
      "step": 1019
    },
    {
      "epoch": 4.142278481012658,
      "grad_norm": 0.21162937529848308,
      "learning_rate": 3.4478677445542653e-06,
      "loss": 0.3636,
      "step": 1020
    },
    {
      "epoch": 4.14632911392405,
      "grad_norm": 0.21422940534385293,
      "learning_rate": 3.4160753118183767e-06,
      "loss": 0.3628,
      "step": 1021
    },
    {
      "epoch": 4.150379746835443,
      "grad_norm": 0.22153180442845785,
      "learning_rate": 3.3844164437320527e-06,
      "loss": 0.3878,
      "step": 1022
    },
    {
      "epoch": 4.154430379746835,
      "grad_norm": 0.20665646667629994,
      "learning_rate": 3.3528913952714558e-06,
      "loss": 0.373,
      "step": 1023
    },
    {
      "epoch": 4.1584810126582274,
      "grad_norm": 0.21196207536796413,
      "learning_rate": 3.321500420335e-06,
      "loss": 0.3789,
      "step": 1024
    },
    {
      "epoch": 4.16253164556962,
      "grad_norm": 0.23294051876282157,
      "learning_rate": 3.290243771741275e-06,
      "loss": 0.3665,
      "step": 1025
    },
    {
      "epoch": 4.1665822784810125,
      "grad_norm": 0.22441982816207498,
      "learning_rate": 3.2591217012270325e-06,
      "loss": 0.3559,
      "step": 1026
    },
    {
      "epoch": 4.170632911392405,
      "grad_norm": 0.21516142695120039,
      "learning_rate": 3.228134459445149e-06,
      "loss": 0.3861,
      "step": 1027
    },
    {
      "epoch": 4.1746835443037975,
      "grad_norm": 0.20517328638064086,
      "learning_rate": 3.1972822959626205e-06,
      "loss": 0.3843,
      "step": 1028
    },
    {
      "epoch": 4.17873417721519,
      "grad_norm": 0.2142310179141372,
      "learning_rate": 3.166565459258513e-06,
      "loss": 0.376,
      "step": 1029
    },
    {
      "epoch": 4.1827848101265825,
      "grad_norm": 0.21256645003829247,
      "learning_rate": 3.1359841967220193e-06,
      "loss": 0.3727,
      "step": 1030
    },
    {
      "epoch": 4.186835443037975,
      "grad_norm": 0.21314043327838494,
      "learning_rate": 3.105538754650419e-06,
      "loss": 0.3817,
      "step": 1031
    },
    {
      "epoch": 4.190886075949367,
      "grad_norm": 0.20635515622124073,
      "learning_rate": 3.07522937824712e-06,
      "loss": 0.3723,
      "step": 1032
    },
    {
      "epoch": 4.19493670886076,
      "grad_norm": 0.20304822356081026,
      "learning_rate": 3.0450563116196697e-06,
      "loss": 0.3719,
      "step": 1033
    },
    {
      "epoch": 4.198987341772152,
      "grad_norm": 0.22051209884849207,
      "learning_rate": 3.0150197977778008e-06,
      "loss": 0.392,
      "step": 1034
    },
    {
      "epoch": 4.203037974683545,
      "grad_norm": 0.19499419958647007,
      "learning_rate": 2.985120078631465e-06,
      "loss": 0.4053,
      "step": 1035
    },
    {
      "epoch": 4.207088607594937,
      "grad_norm": 0.21342296079094844,
      "learning_rate": 2.9553573949888893e-06,
      "loss": 0.3889,
      "step": 1036
    },
    {
      "epoch": 4.211139240506329,
      "grad_norm": 0.20473247065547237,
      "learning_rate": 2.9257319865546384e-06,
      "loss": 0.3767,
      "step": 1037
    },
    {
      "epoch": 4.215189873417722,
      "grad_norm": 0.21089463748435652,
      "learning_rate": 2.896244091927678e-06,
      "loss": 0.3781,
      "step": 1038
    },
    {
      "epoch": 4.219240506329114,
      "grad_norm": 0.19858358725696498,
      "learning_rate": 2.8668939485994584e-06,
      "loss": 0.382,
      "step": 1039
    },
    {
      "epoch": 4.223291139240506,
      "grad_norm": 0.20191219290564008,
      "learning_rate": 2.837681792951994e-06,
      "loss": 0.3647,
      "step": 1040
    },
    {
      "epoch": 4.227341772151899,
      "grad_norm": 0.2019725030549801,
      "learning_rate": 2.808607860255981e-06,
      "loss": 0.3905,
      "step": 1041
    },
    {
      "epoch": 4.231392405063291,
      "grad_norm": 0.21349930948645793,
      "learning_rate": 2.7796723846688634e-06,
      "loss": 0.3872,
      "step": 1042
    },
    {
      "epoch": 4.235443037974684,
      "grad_norm": 1.0011117689013227,
      "learning_rate": 2.7508755992329937e-06,
      "loss": 0.3805,
      "step": 1043
    },
    {
      "epoch": 4.239493670886076,
      "grad_norm": 0.19977169959546123,
      "learning_rate": 2.722217735873718e-06,
      "loss": 0.384,
      "step": 1044
    },
    {
      "epoch": 4.243544303797468,
      "grad_norm": 0.1990730100722398,
      "learning_rate": 2.6936990253975315e-06,
      "loss": 0.405,
      "step": 1045
    },
    {
      "epoch": 4.247594936708861,
      "grad_norm": 0.20782409250256156,
      "learning_rate": 2.665319697490205e-06,
      "loss": 0.3769,
      "step": 1046
    },
    {
      "epoch": 4.251645569620253,
      "grad_norm": 0.20475771391222067,
      "learning_rate": 2.637079980714945e-06,
      "loss": 0.3884,
      "step": 1047
    },
    {
      "epoch": 4.255696202531645,
      "grad_norm": 0.19914421836134014,
      "learning_rate": 2.6089801025105453e-06,
      "loss": 0.3931,
      "step": 1048
    },
    {
      "epoch": 4.259746835443038,
      "grad_norm": 0.20293372854767877,
      "learning_rate": 2.581020289189571e-06,
      "loss": 0.3911,
      "step": 1049
    },
    {
      "epoch": 4.26379746835443,
      "grad_norm": 0.21419082195896083,
      "learning_rate": 2.553200765936501e-06,
      "loss": 0.379,
      "step": 1050
    },
    {
      "epoch": 4.267848101265823,
      "grad_norm": 0.20780153574693935,
      "learning_rate": 2.525521756805962e-06,
      "loss": 0.3693,
      "step": 1051
    },
    {
      "epoch": 4.271898734177215,
      "grad_norm": 0.2073689890929148,
      "learning_rate": 2.497983484720885e-06,
      "loss": 0.4053,
      "step": 1052
    },
    {
      "epoch": 4.275949367088607,
      "grad_norm": 0.2087190459201903,
      "learning_rate": 2.470586171470728e-06,
      "loss": 0.3757,
      "step": 1053
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.2081709792596039,
      "learning_rate": 2.4433300377096836e-06,
      "loss": 0.3752,
      "step": 1054
    },
    {
      "epoch": 4.284050632911392,
      "grad_norm": 0.21285273231659413,
      "learning_rate": 2.4162153029549073e-06,
      "loss": 0.3854,
      "step": 1055
    },
    {
      "epoch": 4.288101265822785,
      "grad_norm": 0.2063176332579796,
      "learning_rate": 2.3892421855847458e-06,
      "loss": 0.3683,
      "step": 1056
    },
    {
      "epoch": 4.292151898734177,
      "grad_norm": 0.2009579624622712,
      "learning_rate": 2.362410902836978e-06,
      "loss": 0.3771,
      "step": 1057
    },
    {
      "epoch": 4.296202531645569,
      "grad_norm": 0.21845461149615683,
      "learning_rate": 2.3357216708070653e-06,
      "loss": 0.366,
      "step": 1058
    },
    {
      "epoch": 4.300253164556962,
      "grad_norm": 0.2082619095615136,
      "learning_rate": 2.309174704446411e-06,
      "loss": 0.3779,
      "step": 1059
    },
    {
      "epoch": 4.3043037974683545,
      "grad_norm": 0.20473717564231894,
      "learning_rate": 2.2827702175606437e-06,
      "loss": 0.3767,
      "step": 1060
    },
    {
      "epoch": 4.3083544303797465,
      "grad_norm": 0.20490717467069527,
      "learning_rate": 2.256508422807855e-06,
      "loss": 0.3649,
      "step": 1061
    },
    {
      "epoch": 4.3124050632911395,
      "grad_norm": 0.21259504948226793,
      "learning_rate": 2.230389531696946e-06,
      "loss": 0.3767,
      "step": 1062
    },
    {
      "epoch": 4.3164556962025316,
      "grad_norm": 0.19412944627512393,
      "learning_rate": 2.204413754585857e-06,
      "loss": 0.3929,
      "step": 1063
    },
    {
      "epoch": 4.320506329113924,
      "grad_norm": 0.20619112925223457,
      "learning_rate": 2.1785813006799406e-06,
      "loss": 0.3826,
      "step": 1064
    },
    {
      "epoch": 4.324556962025317,
      "grad_norm": 0.20286053395387854,
      "learning_rate": 2.1528923780302224e-06,
      "loss": 0.384,
      "step": 1065
    },
    {
      "epoch": 4.328607594936709,
      "grad_norm": 0.19905618525868865,
      "learning_rate": 2.127347193531757e-06,
      "loss": 0.3808,
      "step": 1066
    },
    {
      "epoch": 4.332658227848102,
      "grad_norm": 0.20172286922571148,
      "learning_rate": 2.101945952921942e-06,
      "loss": 0.3841,
      "step": 1067
    },
    {
      "epoch": 4.336708860759494,
      "grad_norm": 0.19692791884677205,
      "learning_rate": 2.0766888607788906e-06,
      "loss": 0.3718,
      "step": 1068
    },
    {
      "epoch": 4.340759493670886,
      "grad_norm": 0.19119871679423692,
      "learning_rate": 2.0515761205197337e-06,
      "loss": 0.3889,
      "step": 1069
    },
    {
      "epoch": 4.344810126582279,
      "grad_norm": 0.20397425509843575,
      "learning_rate": 2.0266079343990453e-06,
      "loss": 0.3831,
      "step": 1070
    },
    {
      "epoch": 4.348860759493671,
      "grad_norm": 0.1976964686308105,
      "learning_rate": 2.0017845035071494e-06,
      "loss": 0.379,
      "step": 1071
    },
    {
      "epoch": 4.352911392405064,
      "grad_norm": 0.19589909779678172,
      "learning_rate": 1.9771060277685537e-06,
      "loss": 0.3987,
      "step": 1072
    },
    {
      "epoch": 4.356962025316456,
      "grad_norm": 0.20117677780396903,
      "learning_rate": 1.95257270594031e-06,
      "loss": 0.3772,
      "step": 1073
    },
    {
      "epoch": 4.361012658227848,
      "grad_norm": 0.1998060923730576,
      "learning_rate": 1.9281847356104188e-06,
      "loss": 0.3764,
      "step": 1074
    },
    {
      "epoch": 4.365063291139241,
      "grad_norm": 0.2164436723076192,
      "learning_rate": 1.9039423131962365e-06,
      "loss": 0.3762,
      "step": 1075
    },
    {
      "epoch": 4.369113924050633,
      "grad_norm": 0.21597743753265858,
      "learning_rate": 1.8798456339429027e-06,
      "loss": 0.3804,
      "step": 1076
    },
    {
      "epoch": 4.373164556962025,
      "grad_norm": 0.1971217703289722,
      "learning_rate": 1.8558948919217612e-06,
      "loss": 0.3836,
      "step": 1077
    },
    {
      "epoch": 4.377215189873418,
      "grad_norm": 0.2030170221838002,
      "learning_rate": 1.8320902800287954e-06,
      "loss": 0.3837,
      "step": 1078
    },
    {
      "epoch": 4.38126582278481,
      "grad_norm": 0.19896786387463775,
      "learning_rate": 1.8084319899830726e-06,
      "loss": 0.3694,
      "step": 1079
    },
    {
      "epoch": 4.385316455696202,
      "grad_norm": 0.19402710608292448,
      "learning_rate": 1.7849202123252097e-06,
      "loss": 0.3813,
      "step": 1080
    },
    {
      "epoch": 4.389367088607595,
      "grad_norm": 0.1986287146043745,
      "learning_rate": 1.7615551364158401e-06,
      "loss": 0.3959,
      "step": 1081
    },
    {
      "epoch": 4.393417721518987,
      "grad_norm": 0.21358687647550562,
      "learning_rate": 1.738336950434061e-06,
      "loss": 0.3828,
      "step": 1082
    },
    {
      "epoch": 4.39746835443038,
      "grad_norm": 0.19943419285920033,
      "learning_rate": 1.715265841375957e-06,
      "loss": 0.3635,
      "step": 1083
    },
    {
      "epoch": 4.401518987341772,
      "grad_norm": 0.19691193857127845,
      "learning_rate": 1.6923419950530684e-06,
      "loss": 0.3876,
      "step": 1084
    },
    {
      "epoch": 4.405569620253164,
      "grad_norm": 0.21193090938918144,
      "learning_rate": 1.6695655960909008e-06,
      "loss": 0.3785,
      "step": 1085
    },
    {
      "epoch": 4.409620253164557,
      "grad_norm": 0.1938815167938794,
      "learning_rate": 1.646936827927441e-06,
      "loss": 0.3883,
      "step": 1086
    },
    {
      "epoch": 4.413670886075949,
      "grad_norm": 0.19794760350049007,
      "learning_rate": 1.6244558728116766e-06,
      "loss": 0.3808,
      "step": 1087
    },
    {
      "epoch": 4.417721518987342,
      "grad_norm": 0.20510416422790625,
      "learning_rate": 1.6021229118021265e-06,
      "loss": 0.3619,
      "step": 1088
    },
    {
      "epoch": 4.421772151898734,
      "grad_norm": 0.19931820114863977,
      "learning_rate": 1.5799381247653967e-06,
      "loss": 0.3794,
      "step": 1089
    },
    {
      "epoch": 4.425822784810126,
      "grad_norm": 0.20114863757410634,
      "learning_rate": 1.5579016903747013e-06,
      "loss": 0.3718,
      "step": 1090
    },
    {
      "epoch": 4.429873417721519,
      "grad_norm": 0.2039038411281141,
      "learning_rate": 1.5360137861084656e-06,
      "loss": 0.3829,
      "step": 1091
    },
    {
      "epoch": 4.433924050632911,
      "grad_norm": 0.1907752140933656,
      "learning_rate": 1.5142745882488475e-06,
      "loss": 0.397,
      "step": 1092
    },
    {
      "epoch": 4.4379746835443035,
      "grad_norm": 0.19425827851980607,
      "learning_rate": 1.4926842718803691e-06,
      "loss": 0.3773,
      "step": 1093
    },
    {
      "epoch": 4.442025316455696,
      "grad_norm": 0.20264131715336695,
      "learning_rate": 1.4712430108884657e-06,
      "loss": 0.38,
      "step": 1094
    },
    {
      "epoch": 4.4460759493670885,
      "grad_norm": 0.20138005411747553,
      "learning_rate": 1.4499509779581078e-06,
      "loss": 0.372,
      "step": 1095
    },
    {
      "epoch": 4.450126582278481,
      "grad_norm": 0.20486147550668538,
      "learning_rate": 1.4288083445723988e-06,
      "loss": 0.3815,
      "step": 1096
    },
    {
      "epoch": 4.4541772151898735,
      "grad_norm": 0.20541601115885025,
      "learning_rate": 1.4078152810112045e-06,
      "loss": 0.3655,
      "step": 1097
    },
    {
      "epoch": 4.458227848101266,
      "grad_norm": 0.19240377973950584,
      "learning_rate": 1.3869719563497697e-06,
      "loss": 0.3871,
      "step": 1098
    },
    {
      "epoch": 4.462278481012659,
      "grad_norm": 0.20451273292622948,
      "learning_rate": 1.3662785384573663e-06,
      "loss": 0.3691,
      "step": 1099
    },
    {
      "epoch": 4.466329113924051,
      "grad_norm": 0.2057331854943753,
      "learning_rate": 1.3457351939959383e-06,
      "loss": 0.3594,
      "step": 1100
    },
    {
      "epoch": 4.470379746835443,
      "grad_norm": 0.2038658153533088,
      "learning_rate": 1.3253420884187551e-06,
      "loss": 0.3662,
      "step": 1101
    },
    {
      "epoch": 4.474430379746836,
      "grad_norm": 0.19228065517513465,
      "learning_rate": 1.3050993859690953e-06,
      "loss": 0.3927,
      "step": 1102
    },
    {
      "epoch": 4.478481012658228,
      "grad_norm": 0.1963315249597224,
      "learning_rate": 1.2850072496788869e-06,
      "loss": 0.3819,
      "step": 1103
    },
    {
      "epoch": 4.482531645569621,
      "grad_norm": 0.20546046611008886,
      "learning_rate": 1.2650658413674434e-06,
      "loss": 0.3688,
      "step": 1104
    },
    {
      "epoch": 4.486582278481013,
      "grad_norm": 0.2025546137224705,
      "learning_rate": 1.2452753216401226e-06,
      "loss": 0.374,
      "step": 1105
    },
    {
      "epoch": 4.490632911392405,
      "grad_norm": 0.19472641220239031,
      "learning_rate": 1.2256358498870503e-06,
      "loss": 0.3664,
      "step": 1106
    },
    {
      "epoch": 4.494683544303798,
      "grad_norm": 0.20326982767644153,
      "learning_rate": 1.2061475842818337e-06,
      "loss": 0.3946,
      "step": 1107
    },
    {
      "epoch": 4.49873417721519,
      "grad_norm": 0.20532966801251537,
      "learning_rate": 1.1868106817802816e-06,
      "loss": 0.3797,
      "step": 1108
    },
    {
      "epoch": 4.502784810126582,
      "grad_norm": 0.19805261876555277,
      "learning_rate": 1.1676252981191482e-06,
      "loss": 0.3667,
      "step": 1109
    },
    {
      "epoch": 4.506835443037975,
      "grad_norm": 0.21233715499080233,
      "learning_rate": 1.1485915878148823e-06,
      "loss": 0.3878,
      "step": 1110
    },
    {
      "epoch": 4.510886075949367,
      "grad_norm": 0.20007747864245404,
      "learning_rate": 1.1297097041623584e-06,
      "loss": 0.3967,
      "step": 1111
    },
    {
      "epoch": 4.514936708860759,
      "grad_norm": 0.19744548378800952,
      "learning_rate": 1.1109797992336847e-06,
      "loss": 0.3741,
      "step": 1112
    },
    {
      "epoch": 4.518987341772152,
      "grad_norm": 0.19813512644133535,
      "learning_rate": 1.092402023876933e-06,
      "loss": 0.3751,
      "step": 1113
    },
    {
      "epoch": 4.523037974683544,
      "grad_norm": 0.19980506773450968,
      "learning_rate": 1.0739765277149527e-06,
      "loss": 0.3825,
      "step": 1114
    },
    {
      "epoch": 4.527088607594937,
      "grad_norm": 0.19880553667323234,
      "learning_rate": 1.0557034591441596e-06,
      "loss": 0.378,
      "step": 1115
    },
    {
      "epoch": 4.531139240506329,
      "grad_norm": 0.1966548801517225,
      "learning_rate": 1.0375829653333324e-06,
      "loss": 0.3787,
      "step": 1116
    },
    {
      "epoch": 4.535189873417721,
      "grad_norm": 0.1989240657027985,
      "learning_rate": 1.0196151922224385e-06,
      "loss": 0.3603,
      "step": 1117
    },
    {
      "epoch": 4.539240506329114,
      "grad_norm": 0.19776329302369436,
      "learning_rate": 1.0018002845214526e-06,
      "loss": 0.3878,
      "step": 1118
    },
    {
      "epoch": 4.543291139240506,
      "grad_norm": 0.19688793698979973,
      "learning_rate": 9.841383857091947e-07,
      "loss": 0.383,
      "step": 1119
    },
    {
      "epoch": 4.547341772151899,
      "grad_norm": 0.20024035475448357,
      "learning_rate": 9.666296380321616e-07,
      "loss": 0.3769,
      "step": 1120
    },
    {
      "epoch": 4.551392405063291,
      "grad_norm": 0.2130665226507089,
      "learning_rate": 9.492741825034124e-07,
      "loss": 0.3786,
      "step": 1121
    },
    {
      "epoch": 4.555443037974683,
      "grad_norm": 0.1975443876465471,
      "learning_rate": 9.320721589013892e-07,
      "loss": 0.3746,
      "step": 1122
    },
    {
      "epoch": 4.559493670886076,
      "grad_norm": 0.19953500204467847,
      "learning_rate": 9.150237057688339e-07,
      "loss": 0.3692,
      "step": 1123
    },
    {
      "epoch": 4.563544303797468,
      "grad_norm": 0.196532987780736,
      "learning_rate": 8.981289604116328e-07,
      "loss": 0.3794,
      "step": 1124
    },
    {
      "epoch": 4.56759493670886,
      "grad_norm": 0.18997526349356722,
      "learning_rate": 8.813880588977542e-07,
      "loss": 0.391,
      "step": 1125
    },
    {
      "epoch": 4.571645569620253,
      "grad_norm": 0.20651138383705542,
      "learning_rate": 8.648011360561126e-07,
      "loss": 0.3821,
      "step": 1126
    },
    {
      "epoch": 4.575696202531645,
      "grad_norm": 0.1975895893207903,
      "learning_rate": 8.483683254755037e-07,
      "loss": 0.3812,
      "step": 1127
    },
    {
      "epoch": 4.5797468354430375,
      "grad_norm": 0.19175387156941356,
      "learning_rate": 8.320897595035227e-07,
      "loss": 0.3844,
      "step": 1128
    },
    {
      "epoch": 4.5837974683544305,
      "grad_norm": 0.19277939946725375,
      "learning_rate": 8.159655692455093e-07,
      "loss": 0.3846,
      "step": 1129
    },
    {
      "epoch": 4.5878481012658225,
      "grad_norm": 0.1955390786948655,
      "learning_rate": 7.999958845634648e-07,
      "loss": 0.3848,
      "step": 1130
    },
    {
      "epoch": 4.5918987341772155,
      "grad_norm": 0.20812047625998514,
      "learning_rate": 7.841808340750478e-07,
      "loss": 0.3708,
      "step": 1131
    },
    {
      "epoch": 4.595949367088608,
      "grad_norm": 0.2017598449588567,
      "learning_rate": 7.685205451524869e-07,
      "loss": 0.3835,
      "step": 1132
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.19469347293312775,
      "learning_rate": 7.530151439216027e-07,
      "loss": 0.3881,
      "step": 1133
    },
    {
      "epoch": 4.604050632911393,
      "grad_norm": 0.20060049958494575,
      "learning_rate": 7.376647552607675e-07,
      "loss": 0.3627,
      "step": 1134
    },
    {
      "epoch": 4.608101265822785,
      "grad_norm": 0.19843380031257696,
      "learning_rate": 7.224695027998963e-07,
      "loss": 0.3726,
      "step": 1135
    },
    {
      "epoch": 4.612151898734178,
      "grad_norm": 0.20524124371765606,
      "learning_rate": 7.07429508919466e-07,
      "loss": 0.3734,
      "step": 1136
    },
    {
      "epoch": 4.61620253164557,
      "grad_norm": 0.18971049653724,
      "learning_rate": 6.925448947495206e-07,
      "loss": 0.3781,
      "step": 1137
    },
    {
      "epoch": 4.620253164556962,
      "grad_norm": 0.20316574440464846,
      "learning_rate": 6.778157801686936e-07,
      "loss": 0.3914,
      "step": 1138
    },
    {
      "epoch": 4.624303797468355,
      "grad_norm": 0.2000601557775413,
      "learning_rate": 6.632422838032515e-07,
      "loss": 0.3855,
      "step": 1139
    },
    {
      "epoch": 4.628354430379747,
      "grad_norm": 0.2023687856635506,
      "learning_rate": 6.488245230261281e-07,
      "loss": 0.3881,
      "step": 1140
    },
    {
      "epoch": 4.632405063291139,
      "grad_norm": 0.19754507905299273,
      "learning_rate": 6.345626139559868e-07,
      "loss": 0.3649,
      "step": 1141
    },
    {
      "epoch": 4.636455696202532,
      "grad_norm": 0.8068431262460422,
      "learning_rate": 6.204566714562866e-07,
      "loss": 0.3749,
      "step": 1142
    },
    {
      "epoch": 4.640506329113924,
      "grad_norm": 0.2064000880330441,
      "learning_rate": 6.06506809134344e-07,
      "loss": 0.3909,
      "step": 1143
    },
    {
      "epoch": 4.644556962025316,
      "grad_norm": 0.19253784302665708,
      "learning_rate": 5.927131393404373e-07,
      "loss": 0.3826,
      "step": 1144
    },
    {
      "epoch": 4.648607594936709,
      "grad_norm": 0.19494067313292524,
      "learning_rate": 5.790757731668817e-07,
      "loss": 0.3749,
      "step": 1145
    },
    {
      "epoch": 4.652658227848101,
      "grad_norm": 0.21674559180904507,
      "learning_rate": 5.655948204471507e-07,
      "loss": 0.3805,
      "step": 1146
    },
    {
      "epoch": 4.656708860759494,
      "grad_norm": 0.18368054804716102,
      "learning_rate": 5.522703897549875e-07,
      "loss": 0.3808,
      "step": 1147
    },
    {
      "epoch": 4.660759493670886,
      "grad_norm": 0.20104073319528418,
      "learning_rate": 5.391025884035239e-07,
      "loss": 0.3713,
      "step": 1148
    },
    {
      "epoch": 4.664810126582278,
      "grad_norm": 0.19550075418580112,
      "learning_rate": 5.260915224444207e-07,
      "loss": 0.3785,
      "step": 1149
    },
    {
      "epoch": 4.668860759493671,
      "grad_norm": 0.19903675815402908,
      "learning_rate": 5.132372966670129e-07,
      "loss": 0.3782,
      "step": 1150
    },
    {
      "epoch": 4.672911392405063,
      "grad_norm": 0.18972617404555303,
      "learning_rate": 5.005400145974704e-07,
      "loss": 0.3855,
      "step": 1151
    },
    {
      "epoch": 4.676962025316456,
      "grad_norm": 0.19560226258719843,
      "learning_rate": 4.879997784979562e-07,
      "loss": 0.3858,
      "step": 1152
    },
    {
      "epoch": 4.681012658227848,
      "grad_norm": 0.2017500093610045,
      "learning_rate": 4.7561668936580984e-07,
      "loss": 0.376,
      "step": 1153
    },
    {
      "epoch": 4.68506329113924,
      "grad_norm": 0.19555050628854243,
      "learning_rate": 4.6339084693272306e-07,
      "loss": 0.3772,
      "step": 1154
    },
    {
      "epoch": 4.689113924050633,
      "grad_norm": 0.19610135378652715,
      "learning_rate": 4.5132234966395847e-07,
      "loss": 0.3827,
      "step": 1155
    },
    {
      "epoch": 4.693164556962025,
      "grad_norm": 0.1982516479361589,
      "learning_rate": 4.3941129475752795e-07,
      "loss": 0.3624,
      "step": 1156
    },
    {
      "epoch": 4.697215189873417,
      "grad_norm": 0.19226272468287492,
      "learning_rate": 4.27657778143431e-07,
      "loss": 0.3989,
      "step": 1157
    },
    {
      "epoch": 4.70126582278481,
      "grad_norm": 0.21262285906334377,
      "learning_rate": 4.1606189448287757e-07,
      "loss": 0.3681,
      "step": 1158
    },
    {
      "epoch": 4.705316455696202,
      "grad_norm": 0.19284927002030816,
      "learning_rate": 4.046237371675177e-07,
      "loss": 0.3602,
      "step": 1159
    },
    {
      "epoch": 4.7093670886075945,
      "grad_norm": 0.19375706796893907,
      "learning_rate": 3.9334339831869963e-07,
      "loss": 0.376,
      "step": 1160
    },
    {
      "epoch": 4.713417721518987,
      "grad_norm": 0.20122951362073185,
      "learning_rate": 3.8222096878671955e-07,
      "loss": 0.396,
      "step": 1161
    },
    {
      "epoch": 4.7174683544303795,
      "grad_norm": 0.18845126688084865,
      "learning_rate": 3.7125653815009545e-07,
      "loss": 0.3921,
      "step": 1162
    },
    {
      "epoch": 4.7215189873417724,
      "grad_norm": 0.1922787378085307,
      "learning_rate": 3.6045019471484974e-07,
      "loss": 0.3881,
      "step": 1163
    },
    {
      "epoch": 4.7255696202531645,
      "grad_norm": 0.19463034037202379,
      "learning_rate": 3.498020255137813e-07,
      "loss": 0.39,
      "step": 1164
    },
    {
      "epoch": 4.729620253164557,
      "grad_norm": 0.19303586110262358,
      "learning_rate": 3.393121163057811e-07,
      "loss": 0.3867,
      "step": 1165
    },
    {
      "epoch": 4.7336708860759495,
      "grad_norm": 0.19274600053929372,
      "learning_rate": 3.289805515751399e-07,
      "loss": 0.3884,
      "step": 1166
    },
    {
      "epoch": 4.737721518987342,
      "grad_norm": 0.19121504434322578,
      "learning_rate": 3.188074145308573e-07,
      "loss": 0.3739,
      "step": 1167
    },
    {
      "epoch": 4.741772151898735,
      "grad_norm": 0.1908193837460981,
      "learning_rate": 3.087927871059804e-07,
      "loss": 0.3675,
      "step": 1168
    },
    {
      "epoch": 4.745822784810127,
      "grad_norm": 0.19676860981052843,
      "learning_rate": 2.989367499569418e-07,
      "loss": 0.381,
      "step": 1169
    },
    {
      "epoch": 4.749873417721519,
      "grad_norm": 0.19204204288564558,
      "learning_rate": 2.8923938246290917e-07,
      "loss": 0.3781,
      "step": 1170
    },
    {
      "epoch": 4.753924050632912,
      "grad_norm": 0.19229476792484437,
      "learning_rate": 2.7970076272514804e-07,
      "loss": 0.3784,
      "step": 1171
    },
    {
      "epoch": 4.757974683544304,
      "grad_norm": 0.19343400394159502,
      "learning_rate": 2.703209675663887e-07,
      "loss": 0.3831,
      "step": 1172
    },
    {
      "epoch": 4.762025316455696,
      "grad_norm": 0.1905975391398447,
      "learning_rate": 2.6110007253021374e-07,
      "loss": 0.3862,
      "step": 1173
    },
    {
      "epoch": 4.766075949367089,
      "grad_norm": 0.19184982338520637,
      "learning_rate": 2.520381518804471e-07,
      "loss": 0.3881,
      "step": 1174
    },
    {
      "epoch": 4.770126582278481,
      "grad_norm": 0.19810991065255307,
      "learning_rate": 2.4313527860054585e-07,
      "loss": 0.3791,
      "step": 1175
    },
    {
      "epoch": 4.774177215189873,
      "grad_norm": 0.18841123920254044,
      "learning_rate": 2.343915243930317e-07,
      "loss": 0.3881,
      "step": 1176
    },
    {
      "epoch": 4.778227848101266,
      "grad_norm": 0.20571639425984445,
      "learning_rate": 2.2580695967889367e-07,
      "loss": 0.3691,
      "step": 1177
    },
    {
      "epoch": 4.782278481012658,
      "grad_norm": 0.21710032399969403,
      "learning_rate": 2.1738165359704189e-07,
      "loss": 0.3642,
      "step": 1178
    },
    {
      "epoch": 4.786329113924051,
      "grad_norm": 0.19478086349954932,
      "learning_rate": 2.0911567400373257e-07,
      "loss": 0.385,
      "step": 1179
    },
    {
      "epoch": 4.790379746835443,
      "grad_norm": 0.18960969789140925,
      "learning_rate": 2.0100908747202607e-07,
      "loss": 0.3894,
      "step": 1180
    },
    {
      "epoch": 4.794430379746835,
      "grad_norm": 0.1984131596290072,
      "learning_rate": 1.9306195929125638e-07,
      "loss": 0.3838,
      "step": 1181
    },
    {
      "epoch": 4.798481012658228,
      "grad_norm": 0.18903937674034524,
      "learning_rate": 1.8527435346650247e-07,
      "loss": 0.3894,
      "step": 1182
    },
    {
      "epoch": 4.80253164556962,
      "grad_norm": 0.1898107465194545,
      "learning_rate": 1.7764633271807108e-07,
      "loss": 0.3819,
      "step": 1183
    },
    {
      "epoch": 4.806582278481013,
      "grad_norm": 0.19904862067664766,
      "learning_rate": 1.7017795848099262e-07,
      "loss": 0.3805,
      "step": 1184
    },
    {
      "epoch": 4.810632911392405,
      "grad_norm": 0.1934024593265589,
      "learning_rate": 1.6286929090452596e-07,
      "loss": 0.3844,
      "step": 1185
    },
    {
      "epoch": 4.814683544303797,
      "grad_norm": 0.18847666813319658,
      "learning_rate": 1.557203888516745e-07,
      "loss": 0.3899,
      "step": 1186
    },
    {
      "epoch": 4.81873417721519,
      "grad_norm": 0.19505186287682147,
      "learning_rate": 1.487313098987131e-07,
      "loss": 0.3715,
      "step": 1187
    },
    {
      "epoch": 4.822784810126582,
      "grad_norm": 0.1898576935305802,
      "learning_rate": 1.4190211033472402e-07,
      "loss": 0.3768,
      "step": 1188
    },
    {
      "epoch": 4.826835443037975,
      "grad_norm": 0.1923491322060219,
      "learning_rate": 1.3523284516113955e-07,
      "loss": 0.3776,
      "step": 1189
    },
    {
      "epoch": 4.830886075949367,
      "grad_norm": 0.1911323939489998,
      "learning_rate": 1.2872356809130682e-07,
      "loss": 0.3746,
      "step": 1190
    },
    {
      "epoch": 4.834936708860759,
      "grad_norm": 0.1992678414765751,
      "learning_rate": 1.2237433155004807e-07,
      "loss": 0.3835,
      "step": 1191
    },
    {
      "epoch": 4.838987341772152,
      "grad_norm": 0.18723522973684825,
      "learning_rate": 1.1618518667323886e-07,
      "loss": 0.3916,
      "step": 1192
    },
    {
      "epoch": 4.843037974683544,
      "grad_norm": 0.18826014151664963,
      "learning_rate": 1.1015618330740385e-07,
      "loss": 0.3876,
      "step": 1193
    },
    {
      "epoch": 4.847088607594936,
      "grad_norm": 0.18712183864019108,
      "learning_rate": 1.042873700093061e-07,
      "loss": 0.3858,
      "step": 1194
    },
    {
      "epoch": 4.851139240506329,
      "grad_norm": 0.18463809802213257,
      "learning_rate": 9.857879404556291e-08,
      "loss": 0.3916,
      "step": 1195
    },
    {
      "epoch": 4.8551898734177215,
      "grad_norm": 0.18847629894987838,
      "learning_rate": 9.303050139225722e-08,
      "loss": 0.3805,
      "step": 1196
    },
    {
      "epoch": 4.8592405063291135,
      "grad_norm": 0.19561146682744515,
      "learning_rate": 8.76425367345779e-08,
      "loss": 0.3729,
      "step": 1197
    },
    {
      "epoch": 4.8632911392405065,
      "grad_norm": 0.1912599695719425,
      "learning_rate": 8.241494346644897e-08,
      "loss": 0.3894,
      "step": 1198
    },
    {
      "epoch": 4.867341772151899,
      "grad_norm": 0.19938593329387821,
      "learning_rate": 7.734776369019204e-08,
      "loss": 0.3632,
      "step": 1199
    },
    {
      "epoch": 4.8713924050632915,
      "grad_norm": 0.19035230581600482,
      "learning_rate": 7.244103821617332e-08,
      "loss": 0.385,
      "step": 1200
    },
    {
      "epoch": 4.875443037974684,
      "grad_norm": 0.19039197968180496,
      "learning_rate": 6.769480656248606e-08,
      "loss": 0.3652,
      "step": 1201
    },
    {
      "epoch": 4.879493670886076,
      "grad_norm": 0.19347597526932914,
      "learning_rate": 6.310910695462635e-08,
      "loss": 0.3669,
      "step": 1202
    },
    {
      "epoch": 4.883544303797469,
      "grad_norm": 0.4419184107553553,
      "learning_rate": 5.8683976325191185e-08,
      "loss": 0.3799,
      "step": 1203
    },
    {
      "epoch": 4.887594936708861,
      "grad_norm": 0.20139004914504005,
      "learning_rate": 5.4419450313571984e-08,
      "loss": 0.3703,
      "step": 1204
    },
    {
      "epoch": 4.891645569620254,
      "grad_norm": 0.18861960412520512,
      "learning_rate": 5.031556326567488e-08,
      "loss": 0.3903,
      "step": 1205
    },
    {
      "epoch": 4.895696202531646,
      "grad_norm": 0.19592989817924164,
      "learning_rate": 4.637234823364312e-08,
      "loss": 0.3683,
      "step": 1206
    },
    {
      "epoch": 4.899746835443038,
      "grad_norm": 0.19467321871094947,
      "learning_rate": 4.258983697558838e-08,
      "loss": 0.3647,
      "step": 1207
    },
    {
      "epoch": 4.903797468354431,
      "grad_norm": 0.1924231123223658,
      "learning_rate": 3.896805995533548e-08,
      "loss": 0.3811,
      "step": 1208
    },
    {
      "epoch": 4.907848101265823,
      "grad_norm": 0.19157462696180616,
      "learning_rate": 3.550704634218028e-08,
      "loss": 0.3778,
      "step": 1209
    },
    {
      "epoch": 4.911898734177215,
      "grad_norm": 0.20112458293696134,
      "learning_rate": 3.2206824010647676e-08,
      "loss": 0.3665,
      "step": 1210
    },
    {
      "epoch": 4.915949367088608,
      "grad_norm": 0.19379730801217673,
      "learning_rate": 2.9067419540278476e-08,
      "loss": 0.3769,
      "step": 1211
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.20096615449338293,
      "learning_rate": 2.6088858215400638e-08,
      "loss": 0.3704,
      "step": 1212
    },
    {
      "epoch": 4.924050632911392,
      "grad_norm": 0.19431433658902003,
      "learning_rate": 2.3271164024940564e-08,
      "loss": 0.3838,
      "step": 1213
    },
    {
      "epoch": 4.928101265822785,
      "grad_norm": 0.1907444995418682,
      "learning_rate": 2.061435966221881e-08,
      "loss": 0.3925,
      "step": 1214
    },
    {
      "epoch": 4.932151898734177,
      "grad_norm": 0.19218431425281005,
      "learning_rate": 1.811846652477245e-08,
      "loss": 0.3867,
      "step": 1215
    },
    {
      "epoch": 4.93620253164557,
      "grad_norm": 0.19008998738846286,
      "learning_rate": 1.5783504714184106e-08,
      "loss": 0.3668,
      "step": 1216
    },
    {
      "epoch": 4.940253164556962,
      "grad_norm": 0.18933173404834758,
      "learning_rate": 1.360949303591097e-08,
      "loss": 0.3947,
      "step": 1217
    },
    {
      "epoch": 4.944303797468354,
      "grad_norm": 0.20096633093539032,
      "learning_rate": 1.1596448999144916e-08,
      "loss": 0.3705,
      "step": 1218
    },
    {
      "epoch": 4.948354430379747,
      "grad_norm": 0.2051904916613494,
      "learning_rate": 9.744388816668172e-09,
      "loss": 0.3622,
      "step": 1219
    },
    {
      "epoch": 4.952405063291139,
      "grad_norm": 0.19508522625707883,
      "learning_rate": 8.05332740472009e-09,
      "loss": 0.3846,
      "step": 1220
    },
    {
      "epoch": 4.956455696202532,
      "grad_norm": 0.18963963483191118,
      "learning_rate": 6.523278382872811e-09,
      "loss": 0.3693,
      "step": 1221
    },
    {
      "epoch": 4.960506329113924,
      "grad_norm": 0.19074075013103045,
      "learning_rate": 5.15425407393133e-09,
      "loss": 0.3718,
      "step": 1222
    },
    {
      "epoch": 4.964556962025316,
      "grad_norm": 0.19999163440316045,
      "learning_rate": 3.94626550383137e-09,
      "loss": 0.3755,
      "step": 1223
    },
    {
      "epoch": 4.968607594936709,
      "grad_norm": 0.18579380178770835,
      "learning_rate": 2.899322401546112e-09,
      "loss": 0.3705,
      "step": 1224
    },
    {
      "epoch": 4.972658227848101,
      "grad_norm": 0.1972805734734112,
      "learning_rate": 2.013433199010706e-09,
      "loss": 0.3667,
      "step": 1225
    },
    {
      "epoch": 4.976708860759493,
      "grad_norm": 0.19713576960957435,
      "learning_rate": 1.2886050310556563e-09,
      "loss": 0.3734,
      "step": 1226
    },
    {
      "epoch": 4.980759493670886,
      "grad_norm": 0.1883470607002637,
      "learning_rate": 7.248437353468695e-10,
      "loss": 0.3853,
      "step": 1227
    },
    {
      "epoch": 4.984810126582278,
      "grad_norm": 0.19452079142537607,
      "learning_rate": 3.221538523412449e-10,
      "loss": 0.3755,
      "step": 1228
    },
    {
      "epoch": 4.9888607594936705,
      "grad_norm": 0.18501710608159033,
      "learning_rate": 8.053862524670663e-11,
      "loss": 0.3792,
      "step": 1229
    },
    {
      "epoch": 4.992911392405063,
      "grad_norm": 0.18813966745080327,
      "learning_rate": 0.0,
      "loss": 0.3929,
      "step": 1230
    },
    {
      "epoch": 4.992911392405063,
      "step": 1230,
      "total_flos": 5.974905031848952e+18,
      "train_loss": 0.15978388905040616,
      "train_runtime": 22311.7261,
      "train_samples_per_second": 7.081,
      "train_steps_per_second": 0.055
    }
  ],
  "logging_steps": 1,
  "max_steps": 1230,
  "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": 5.974905031848952e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}