{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.99964924587864,
  "global_step": 14250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07,
      "learning_rate": 4.982394366197183e-06,
      "loss": 1.0614,
      "step": 100
    },
    {
      "epoch": 0.07,
      "eval_accuracy": 0.4345000088214874,
      "eval_loss": 1.019625186920166,
      "eval_runtime": 18.8668,
      "eval_samples_per_second": 106.006,
      "eval_steps_per_second": 6.625,
      "step": 100
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.94718309859155e-06,
      "loss": 0.8601,
      "step": 200
    },
    {
      "epoch": 0.14,
      "eval_accuracy": 0.6460000276565552,
      "eval_loss": 0.756058931350708,
      "eval_runtime": 19.8113,
      "eval_samples_per_second": 100.953,
      "eval_steps_per_second": 6.31,
      "step": 200
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.911971830985916e-06,
      "loss": 0.734,
      "step": 300
    },
    {
      "epoch": 0.21,
      "eval_accuracy": 0.6955000162124634,
      "eval_loss": 0.6796478629112244,
      "eval_runtime": 19.6375,
      "eval_samples_per_second": 101.846,
      "eval_steps_per_second": 6.365,
      "step": 300
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.876760563380282e-06,
      "loss": 0.6753,
      "step": 400
    },
    {
      "epoch": 0.28,
      "eval_accuracy": 0.699999988079071,
      "eval_loss": 0.6520820260047913,
      "eval_runtime": 18.2758,
      "eval_samples_per_second": 109.434,
      "eval_steps_per_second": 6.84,
      "step": 400
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.841549295774649e-06,
      "loss": 0.6408,
      "step": 500
    },
    {
      "epoch": 0.35,
      "eval_accuracy": 0.7440000176429749,
      "eval_loss": 0.6119081974029541,
      "eval_runtime": 17.8217,
      "eval_samples_per_second": 112.223,
      "eval_steps_per_second": 7.014,
      "step": 500
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.806338028169015e-06,
      "loss": 0.5991,
      "step": 600
    },
    {
      "epoch": 0.42,
      "eval_accuracy": 0.7369999885559082,
      "eval_loss": 0.6033942699432373,
      "eval_runtime": 17.3933,
      "eval_samples_per_second": 114.986,
      "eval_steps_per_second": 7.187,
      "step": 600
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.771126760563381e-06,
      "loss": 0.6069,
      "step": 700
    },
    {
      "epoch": 0.49,
      "eval_accuracy": 0.737500011920929,
      "eval_loss": 0.597550630569458,
      "eval_runtime": 20.251,
      "eval_samples_per_second": 98.76,
      "eval_steps_per_second": 6.173,
      "step": 700
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.735915492957747e-06,
      "loss": 0.6122,
      "step": 800
    },
    {
      "epoch": 0.56,
      "eval_accuracy": 0.7425000071525574,
      "eval_loss": 0.5870603322982788,
      "eval_runtime": 21.2152,
      "eval_samples_per_second": 94.272,
      "eval_steps_per_second": 5.892,
      "step": 800
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.7007042253521126e-06,
      "loss": 0.5908,
      "step": 900
    },
    {
      "epoch": 0.63,
      "eval_accuracy": 0.7444999814033508,
      "eval_loss": 0.5935022234916687,
      "eval_runtime": 22.0881,
      "eval_samples_per_second": 90.546,
      "eval_steps_per_second": 5.659,
      "step": 900
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.665492957746479e-06,
      "loss": 0.5884,
      "step": 1000
    },
    {
      "epoch": 0.7,
      "eval_accuracy": 0.7519999742507935,
      "eval_loss": 0.5792337656021118,
      "eval_runtime": 17.6606,
      "eval_samples_per_second": 113.246,
      "eval_steps_per_second": 7.078,
      "step": 1000
    },
    {
      "epoch": 0.77,
      "learning_rate": 4.630281690140845e-06,
      "loss": 0.5839,
      "step": 1100
    },
    {
      "epoch": 0.77,
      "eval_accuracy": 0.7555000185966492,
      "eval_loss": 0.578044056892395,
      "eval_runtime": 15.9931,
      "eval_samples_per_second": 125.054,
      "eval_steps_per_second": 7.816,
      "step": 1100
    },
    {
      "epoch": 0.84,
      "learning_rate": 4.595070422535211e-06,
      "loss": 0.5772,
      "step": 1200
    },
    {
      "epoch": 0.84,
      "eval_accuracy": 0.7570000290870667,
      "eval_loss": 0.5727072954177856,
      "eval_runtime": 21.5937,
      "eval_samples_per_second": 92.62,
      "eval_steps_per_second": 5.789,
      "step": 1200
    },
    {
      "epoch": 0.91,
      "learning_rate": 4.559859154929578e-06,
      "loss": 0.5895,
      "step": 1300
    },
    {
      "epoch": 0.91,
      "eval_accuracy": 0.7549999952316284,
      "eval_loss": 0.5601378679275513,
      "eval_runtime": 20.1543,
      "eval_samples_per_second": 99.234,
      "eval_steps_per_second": 6.202,
      "step": 1300
    },
    {
      "epoch": 0.98,
      "learning_rate": 4.524647887323944e-06,
      "loss": 0.5757,
      "step": 1400
    },
    {
      "epoch": 0.98,
      "eval_accuracy": 0.7524999976158142,
      "eval_loss": 0.561326801776886,
      "eval_runtime": 15.0299,
      "eval_samples_per_second": 133.068,
      "eval_steps_per_second": 8.317,
      "step": 1400
    },
    {
      "epoch": 1.05,
      "learning_rate": 4.489436619718311e-06,
      "loss": 0.5121,
      "step": 1500
    },
    {
      "epoch": 1.05,
      "eval_accuracy": 0.7599999904632568,
      "eval_loss": 0.5866703987121582,
      "eval_runtime": 18.2735,
      "eval_samples_per_second": 109.448,
      "eval_steps_per_second": 6.841,
      "step": 1500
    },
    {
      "epoch": 1.12,
      "learning_rate": 4.454225352112677e-06,
      "loss": 0.5254,
      "step": 1600
    },
    {
      "epoch": 1.12,
      "eval_accuracy": 0.7630000114440918,
      "eval_loss": 0.5595362186431885,
      "eval_runtime": 21.9878,
      "eval_samples_per_second": 90.959,
      "eval_steps_per_second": 5.685,
      "step": 1600
    },
    {
      "epoch": 1.19,
      "learning_rate": 4.419014084507043e-06,
      "loss": 0.5074,
      "step": 1700
    },
    {
      "epoch": 1.19,
      "eval_accuracy": 0.7584999799728394,
      "eval_loss": 0.559354841709137,
      "eval_runtime": 19.9155,
      "eval_samples_per_second": 100.424,
      "eval_steps_per_second": 6.277,
      "step": 1700
    },
    {
      "epoch": 1.26,
      "learning_rate": 4.383802816901409e-06,
      "loss": 0.4947,
      "step": 1800
    },
    {
      "epoch": 1.26,
      "eval_accuracy": 0.7574999928474426,
      "eval_loss": 0.5696709156036377,
      "eval_runtime": 22.8735,
      "eval_samples_per_second": 87.438,
      "eval_steps_per_second": 5.465,
      "step": 1800
    },
    {
      "epoch": 1.33,
      "learning_rate": 4.3485915492957745e-06,
      "loss": 0.5019,
      "step": 1900
    },
    {
      "epoch": 1.33,
      "eval_accuracy": 0.7580000162124634,
      "eval_loss": 0.5664528608322144,
      "eval_runtime": 22.2002,
      "eval_samples_per_second": 90.089,
      "eval_steps_per_second": 5.631,
      "step": 1900
    },
    {
      "epoch": 1.4,
      "learning_rate": 4.313380281690141e-06,
      "loss": 0.5005,
      "step": 2000
    },
    {
      "epoch": 1.4,
      "eval_accuracy": 0.765500009059906,
      "eval_loss": 0.5484071969985962,
      "eval_runtime": 21.9941,
      "eval_samples_per_second": 90.934,
      "eval_steps_per_second": 5.683,
      "step": 2000
    },
    {
      "epoch": 1.47,
      "learning_rate": 4.278169014084507e-06,
      "loss": 0.5125,
      "step": 2100
    },
    {
      "epoch": 1.47,
      "eval_accuracy": 0.7605000138282776,
      "eval_loss": 0.5626400709152222,
      "eval_runtime": 17.7868,
      "eval_samples_per_second": 112.443,
      "eval_steps_per_second": 7.028,
      "step": 2100
    },
    {
      "epoch": 1.54,
      "learning_rate": 4.242957746478873e-06,
      "loss": 0.5241,
      "step": 2200
    },
    {
      "epoch": 1.54,
      "eval_accuracy": 0.7559999823570251,
      "eval_loss": 0.556066632270813,
      "eval_runtime": 18.9121,
      "eval_samples_per_second": 105.753,
      "eval_steps_per_second": 6.61,
      "step": 2200
    },
    {
      "epoch": 1.61,
      "learning_rate": 4.20774647887324e-06,
      "loss": 0.5198,
      "step": 2300
    },
    {
      "epoch": 1.61,
      "eval_accuracy": 0.7599999904632568,
      "eval_loss": 0.560243546962738,
      "eval_runtime": 20.0908,
      "eval_samples_per_second": 99.548,
      "eval_steps_per_second": 6.222,
      "step": 2300
    },
    {
      "epoch": 1.68,
      "learning_rate": 4.172535211267606e-06,
      "loss": 0.5124,
      "step": 2400
    },
    {
      "epoch": 1.68,
      "eval_accuracy": 0.7490000128746033,
      "eval_loss": 0.5654177665710449,
      "eval_runtime": 19.7883,
      "eval_samples_per_second": 101.07,
      "eval_steps_per_second": 6.317,
      "step": 2400
    },
    {
      "epoch": 1.75,
      "learning_rate": 4.137323943661972e-06,
      "loss": 0.5096,
      "step": 2500
    },
    {
      "epoch": 1.75,
      "eval_accuracy": 0.7515000104904175,
      "eval_loss": 0.5803455710411072,
      "eval_runtime": 22.0507,
      "eval_samples_per_second": 90.7,
      "eval_steps_per_second": 5.669,
      "step": 2500
    },
    {
      "epoch": 1.82,
      "learning_rate": 4.102112676056339e-06,
      "loss": 0.4885,
      "step": 2600
    },
    {
      "epoch": 1.82,
      "eval_accuracy": 0.75,
      "eval_loss": 0.5889333486557007,
      "eval_runtime": 21.1933,
      "eval_samples_per_second": 94.369,
      "eval_steps_per_second": 5.898,
      "step": 2600
    },
    {
      "epoch": 1.89,
      "learning_rate": 4.0669014084507045e-06,
      "loss": 0.5111,
      "step": 2700
    },
    {
      "epoch": 1.89,
      "eval_accuracy": 0.7664999961853027,
      "eval_loss": 0.5507832169532776,
      "eval_runtime": 20.2761,
      "eval_samples_per_second": 98.638,
      "eval_steps_per_second": 6.165,
      "step": 2700
    },
    {
      "epoch": 1.96,
      "learning_rate": 4.031690140845071e-06,
      "loss": 0.4868,
      "step": 2800
    },
    {
      "epoch": 1.96,
      "eval_accuracy": 0.7634999752044678,
      "eval_loss": 0.5621495842933655,
      "eval_runtime": 18.2341,
      "eval_samples_per_second": 109.685,
      "eval_steps_per_second": 6.855,
      "step": 2800
    },
    {
      "epoch": 2.04,
      "learning_rate": 3.996478873239437e-06,
      "loss": 0.4599,
      "step": 2900
    },
    {
      "epoch": 2.04,
      "eval_accuracy": 0.7615000009536743,
      "eval_loss": 0.5994852185249329,
      "eval_runtime": 17.787,
      "eval_samples_per_second": 112.442,
      "eval_steps_per_second": 7.028,
      "step": 2900
    },
    {
      "epoch": 2.11,
      "learning_rate": 3.961267605633803e-06,
      "loss": 0.4147,
      "step": 3000
    },
    {
      "epoch": 2.11,
      "eval_accuracy": 0.753000020980835,
      "eval_loss": 0.6202083230018616,
      "eval_runtime": 20.5139,
      "eval_samples_per_second": 97.495,
      "eval_steps_per_second": 6.093,
      "step": 3000
    },
    {
      "epoch": 2.18,
      "learning_rate": 3.926056338028169e-06,
      "loss": 0.4233,
      "step": 3100
    },
    {
      "epoch": 2.18,
      "eval_accuracy": 0.762499988079071,
      "eval_loss": 0.5875486135482788,
      "eval_runtime": 19.4349,
      "eval_samples_per_second": 102.908,
      "eval_steps_per_second": 6.432,
      "step": 3100
    },
    {
      "epoch": 2.25,
      "learning_rate": 3.890845070422535e-06,
      "loss": 0.4324,
      "step": 3200
    },
    {
      "epoch": 2.25,
      "eval_accuracy": 0.7609999775886536,
      "eval_loss": 0.5794370174407959,
      "eval_runtime": 18.3807,
      "eval_samples_per_second": 108.81,
      "eval_steps_per_second": 6.801,
      "step": 3200
    },
    {
      "epoch": 2.32,
      "learning_rate": 3.855633802816902e-06,
      "loss": 0.4141,
      "step": 3300
    },
    {
      "epoch": 2.32,
      "eval_accuracy": 0.7459999918937683,
      "eval_loss": 0.5901930928230286,
      "eval_runtime": 20.1122,
      "eval_samples_per_second": 99.442,
      "eval_steps_per_second": 6.215,
      "step": 3300
    },
    {
      "epoch": 2.39,
      "learning_rate": 3.820422535211268e-06,
      "loss": 0.4306,
      "step": 3400
    },
    {
      "epoch": 2.39,
      "eval_accuracy": 0.7544999718666077,
      "eval_loss": 0.6053192019462585,
      "eval_runtime": 20.0091,
      "eval_samples_per_second": 99.955,
      "eval_steps_per_second": 6.247,
      "step": 3400
    },
    {
      "epoch": 2.46,
      "learning_rate": 3.785211267605634e-06,
      "loss": 0.4266,
      "step": 3500
    },
    {
      "epoch": 2.46,
      "eval_accuracy": 0.7570000290870667,
      "eval_loss": 0.5978769659996033,
      "eval_runtime": 19.8772,
      "eval_samples_per_second": 100.618,
      "eval_steps_per_second": 6.289,
      "step": 3500
    },
    {
      "epoch": 2.53,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.4227,
      "step": 3600
    },
    {
      "epoch": 2.53,
      "eval_accuracy": 0.7649999856948853,
      "eval_loss": 0.5919951796531677,
      "eval_runtime": 21.1736,
      "eval_samples_per_second": 94.457,
      "eval_steps_per_second": 5.904,
      "step": 3600
    },
    {
      "epoch": 2.6,
      "learning_rate": 3.7147887323943665e-06,
      "loss": 0.4226,
      "step": 3700
    },
    {
      "epoch": 2.6,
      "eval_accuracy": 0.7455000281333923,
      "eval_loss": 0.6165611743927002,
      "eval_runtime": 20.5872,
      "eval_samples_per_second": 97.148,
      "eval_steps_per_second": 6.072,
      "step": 3700
    },
    {
      "epoch": 2.67,
      "learning_rate": 3.679577464788733e-06,
      "loss": 0.3978,
      "step": 3800
    },
    {
      "epoch": 2.67,
      "eval_accuracy": 0.7559999823570251,
      "eval_loss": 0.6125866770744324,
      "eval_runtime": 22.1514,
      "eval_samples_per_second": 90.288,
      "eval_steps_per_second": 5.643,
      "step": 3800
    },
    {
      "epoch": 2.74,
      "learning_rate": 3.644366197183099e-06,
      "loss": 0.3954,
      "step": 3900
    },
    {
      "epoch": 2.74,
      "eval_accuracy": 0.7549999952316284,
      "eval_loss": 0.615158200263977,
      "eval_runtime": 17.1074,
      "eval_samples_per_second": 116.908,
      "eval_steps_per_second": 7.307,
      "step": 3900
    },
    {
      "epoch": 2.81,
      "learning_rate": 3.609154929577465e-06,
      "loss": 0.4209,
      "step": 4000
    },
    {
      "epoch": 2.81,
      "eval_accuracy": 0.75,
      "eval_loss": 0.597953736782074,
      "eval_runtime": 22.6397,
      "eval_samples_per_second": 88.34,
      "eval_steps_per_second": 5.521,
      "step": 4000
    },
    {
      "epoch": 2.88,
      "learning_rate": 3.5739436619718315e-06,
      "loss": 0.3982,
      "step": 4100
    },
    {
      "epoch": 2.88,
      "eval_accuracy": 0.7490000128746033,
      "eval_loss": 0.6096097230911255,
      "eval_runtime": 22.0097,
      "eval_samples_per_second": 90.869,
      "eval_steps_per_second": 5.679,
      "step": 4100
    },
    {
      "epoch": 2.95,
      "learning_rate": 3.538732394366197e-06,
      "loss": 0.4016,
      "step": 4200
    },
    {
      "epoch": 2.95,
      "eval_accuracy": 0.7425000071525574,
      "eval_loss": 0.6540722846984863,
      "eval_runtime": 16.1358,
      "eval_samples_per_second": 123.948,
      "eval_steps_per_second": 7.747,
      "step": 4200
    },
    {
      "epoch": 3.02,
      "learning_rate": 3.5035211267605634e-06,
      "loss": 0.3966,
      "step": 4300
    },
    {
      "epoch": 3.02,
      "eval_accuracy": 0.7544999718666077,
      "eval_loss": 0.6377372145652771,
      "eval_runtime": 20.4003,
      "eval_samples_per_second": 98.038,
      "eval_steps_per_second": 6.127,
      "step": 4300
    },
    {
      "epoch": 3.09,
      "learning_rate": 3.4683098591549297e-06,
      "loss": 0.3074,
      "step": 4400
    },
    {
      "epoch": 3.09,
      "eval_accuracy": 0.75,
      "eval_loss": 0.6859884262084961,
      "eval_runtime": 20.6366,
      "eval_samples_per_second": 96.915,
      "eval_steps_per_second": 6.057,
      "step": 4400
    },
    {
      "epoch": 3.16,
      "learning_rate": 3.433098591549296e-06,
      "loss": 0.3551,
      "step": 4500
    },
    {
      "epoch": 3.16,
      "eval_accuracy": 0.7549999952316284,
      "eval_loss": 0.6160025596618652,
      "eval_runtime": 21.368,
      "eval_samples_per_second": 93.598,
      "eval_steps_per_second": 5.85,
      "step": 4500
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.397887323943662e-06,
      "loss": 0.3323,
      "step": 4600
    },
    {
      "epoch": 3.23,
      "eval_accuracy": 0.7519999742507935,
      "eval_loss": 0.6714155077934265,
      "eval_runtime": 21.1568,
      "eval_samples_per_second": 94.532,
      "eval_steps_per_second": 5.908,
      "step": 4600
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.3626760563380284e-06,
      "loss": 0.3171,
      "step": 4700
    },
    {
      "epoch": 3.3,
      "eval_accuracy": 0.7534999847412109,
      "eval_loss": 0.6537904739379883,
      "eval_runtime": 17.8954,
      "eval_samples_per_second": 111.76,
      "eval_steps_per_second": 6.985,
      "step": 4700
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.3274647887323947e-06,
      "loss": 0.3403,
      "step": 4800
    },
    {
      "epoch": 3.37,
      "eval_accuracy": 0.7465000152587891,
      "eval_loss": 0.677370548248291,
      "eval_runtime": 22.2854,
      "eval_samples_per_second": 89.745,
      "eval_steps_per_second": 5.609,
      "step": 4800
    },
    {
      "epoch": 3.44,
      "learning_rate": 3.292253521126761e-06,
      "loss": 0.3396,
      "step": 4900
    },
    {
      "epoch": 3.44,
      "eval_accuracy": 0.7465000152587891,
      "eval_loss": 0.6725812554359436,
      "eval_runtime": 20.4072,
      "eval_samples_per_second": 98.005,
      "eval_steps_per_second": 6.125,
      "step": 4900
    },
    {
      "epoch": 3.51,
      "learning_rate": 3.257042253521127e-06,
      "loss": 0.3259,
      "step": 5000
    },
    {
      "epoch": 3.51,
      "eval_accuracy": 0.7480000257492065,
      "eval_loss": 0.6465049982070923,
      "eval_runtime": 16.3957,
      "eval_samples_per_second": 121.983,
      "eval_steps_per_second": 7.624,
      "step": 5000
    },
    {
      "epoch": 3.58,
      "learning_rate": 3.2218309859154934e-06,
      "loss": 0.3392,
      "step": 5100
    },
    {
      "epoch": 3.58,
      "eval_accuracy": 0.7459999918937683,
      "eval_loss": 0.6860352754592896,
      "eval_runtime": 19.1797,
      "eval_samples_per_second": 104.277,
      "eval_steps_per_second": 6.517,
      "step": 5100
    },
    {
      "epoch": 3.65,
      "learning_rate": 3.1866197183098598e-06,
      "loss": 0.3251,
      "step": 5200
    },
    {
      "epoch": 3.65,
      "eval_accuracy": 0.7494999766349792,
      "eval_loss": 0.6696720123291016,
      "eval_runtime": 17.883,
      "eval_samples_per_second": 111.838,
      "eval_steps_per_second": 6.99,
      "step": 5200
    },
    {
      "epoch": 3.72,
      "learning_rate": 3.1514084507042257e-06,
      "loss": 0.3253,
      "step": 5300
    },
    {
      "epoch": 3.72,
      "eval_accuracy": 0.7429999709129333,
      "eval_loss": 0.6769505739212036,
      "eval_runtime": 19.8341,
      "eval_samples_per_second": 100.836,
      "eval_steps_per_second": 6.302,
      "step": 5300
    },
    {
      "epoch": 3.79,
      "learning_rate": 3.1161971830985916e-06,
      "loss": 0.3455,
      "step": 5400
    },
    {
      "epoch": 3.79,
      "eval_accuracy": 0.7360000014305115,
      "eval_loss": 0.7176979780197144,
      "eval_runtime": 17.9226,
      "eval_samples_per_second": 111.591,
      "eval_steps_per_second": 6.974,
      "step": 5400
    },
    {
      "epoch": 3.86,
      "learning_rate": 3.0809859154929576e-06,
      "loss": 0.3323,
      "step": 5500
    },
    {
      "epoch": 3.86,
      "eval_accuracy": 0.7400000095367432,
      "eval_loss": 0.6943067908287048,
      "eval_runtime": 21.4155,
      "eval_samples_per_second": 93.39,
      "eval_steps_per_second": 5.837,
      "step": 5500
    },
    {
      "epoch": 3.93,
      "learning_rate": 3.045774647887324e-06,
      "loss": 0.3335,
      "step": 5600
    },
    {
      "epoch": 3.93,
      "eval_accuracy": 0.7555000185966492,
      "eval_loss": 0.6506811380386353,
      "eval_runtime": 21.6099,
      "eval_samples_per_second": 92.55,
      "eval_steps_per_second": 5.784,
      "step": 5600
    },
    {
      "epoch": 4.0,
      "learning_rate": 3.0105633802816903e-06,
      "loss": 0.3368,
      "step": 5700
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7484999895095825,
      "eval_loss": 0.6580154895782471,
      "eval_runtime": 21.7515,
      "eval_samples_per_second": 91.948,
      "eval_steps_per_second": 5.747,
      "step": 5700
    },
    {
      "epoch": 4.07,
      "learning_rate": 2.9753521126760567e-06,
      "loss": 0.2479,
      "step": 5800
    },
    {
      "epoch": 4.07,
      "eval_accuracy": 0.7429999709129333,
      "eval_loss": 0.7666531801223755,
      "eval_runtime": 15.994,
      "eval_samples_per_second": 125.047,
      "eval_steps_per_second": 7.815,
      "step": 5800
    },
    {
      "epoch": 4.14,
      "learning_rate": 2.9401408450704226e-06,
      "loss": 0.2613,
      "step": 5900
    },
    {
      "epoch": 4.14,
      "eval_accuracy": 0.7505000233650208,
      "eval_loss": 0.751265823841095,
      "eval_runtime": 16.5258,
      "eval_samples_per_second": 121.023,
      "eval_steps_per_second": 7.564,
      "step": 5900
    },
    {
      "epoch": 4.21,
      "learning_rate": 2.904929577464789e-06,
      "loss": 0.2557,
      "step": 6000
    },
    {
      "epoch": 4.21,
      "eval_accuracy": 0.7484999895095825,
      "eval_loss": 0.7926999926567078,
      "eval_runtime": 20.2799,
      "eval_samples_per_second": 98.62,
      "eval_steps_per_second": 6.164,
      "step": 6000
    },
    {
      "epoch": 4.28,
      "learning_rate": 2.8697183098591553e-06,
      "loss": 0.243,
      "step": 6100
    },
    {
      "epoch": 4.28,
      "eval_accuracy": 0.7450000047683716,
      "eval_loss": 0.77916020154953,
      "eval_runtime": 19.6554,
      "eval_samples_per_second": 101.753,
      "eval_steps_per_second": 6.36,
      "step": 6100
    },
    {
      "epoch": 4.35,
      "learning_rate": 2.8345070422535217e-06,
      "loss": 0.2473,
      "step": 6200
    },
    {
      "epoch": 4.35,
      "eval_accuracy": 0.7354999780654907,
      "eval_loss": 0.8106710314750671,
      "eval_runtime": 14.7909,
      "eval_samples_per_second": 135.218,
      "eval_steps_per_second": 8.451,
      "step": 6200
    },
    {
      "epoch": 4.42,
      "learning_rate": 2.7992957746478876e-06,
      "loss": 0.2447,
      "step": 6300
    },
    {
      "epoch": 4.42,
      "eval_accuracy": 0.7369999885559082,
      "eval_loss": 0.7850819230079651,
      "eval_runtime": 15.145,
      "eval_samples_per_second": 132.057,
      "eval_steps_per_second": 8.254,
      "step": 6300
    },
    {
      "epoch": 4.49,
      "learning_rate": 2.764084507042254e-06,
      "loss": 0.2515,
      "step": 6400
    },
    {
      "epoch": 4.49,
      "eval_accuracy": 0.7465000152587891,
      "eval_loss": 0.7529160380363464,
      "eval_runtime": 14.1294,
      "eval_samples_per_second": 141.549,
      "eval_steps_per_second": 8.847,
      "step": 6400
    },
    {
      "epoch": 4.56,
      "learning_rate": 2.7288732394366203e-06,
      "loss": 0.274,
      "step": 6500
    },
    {
      "epoch": 4.56,
      "eval_accuracy": 0.7465000152587891,
      "eval_loss": 0.7389978170394897,
      "eval_runtime": 18.3936,
      "eval_samples_per_second": 108.733,
      "eval_steps_per_second": 6.796,
      "step": 6500
    },
    {
      "epoch": 4.63,
      "learning_rate": 2.693661971830986e-06,
      "loss": 0.2674,
      "step": 6600
    },
    {
      "epoch": 4.63,
      "eval_accuracy": 0.7459999918937683,
      "eval_loss": 0.7657651305198669,
      "eval_runtime": 19.2395,
      "eval_samples_per_second": 103.953,
      "eval_steps_per_second": 6.497,
      "step": 6600
    },
    {
      "epoch": 4.7,
      "learning_rate": 2.6584507042253522e-06,
      "loss": 0.2416,
      "step": 6700
    },
    {
      "epoch": 4.7,
      "eval_accuracy": 0.7484999895095825,
      "eval_loss": 0.7914510369300842,
      "eval_runtime": 17.7833,
      "eval_samples_per_second": 112.465,
      "eval_steps_per_second": 7.029,
      "step": 6700
    },
    {
      "epoch": 4.77,
      "learning_rate": 2.623239436619718e-06,
      "loss": 0.2432,
      "step": 6800
    },
    {
      "epoch": 4.77,
      "eval_accuracy": 0.7434999942779541,
      "eval_loss": 0.7988595962524414,
      "eval_runtime": 16.8516,
      "eval_samples_per_second": 118.683,
      "eval_steps_per_second": 7.418,
      "step": 6800
    },
    {
      "epoch": 4.84,
      "learning_rate": 2.5880281690140845e-06,
      "loss": 0.2595,
      "step": 6900
    },
    {
      "epoch": 4.84,
      "eval_accuracy": 0.7379999756813049,
      "eval_loss": 0.7850367426872253,
      "eval_runtime": 22.0351,
      "eval_samples_per_second": 90.764,
      "eval_steps_per_second": 5.673,
      "step": 6900
    },
    {
      "epoch": 4.91,
      "learning_rate": 2.552816901408451e-06,
      "loss": 0.2736,
      "step": 7000
    },
    {
      "epoch": 4.91,
      "eval_accuracy": 0.7394999861717224,
      "eval_loss": 0.7577053308486938,
      "eval_runtime": 22.4529,
      "eval_samples_per_second": 89.075,
      "eval_steps_per_second": 5.567,
      "step": 7000
    },
    {
      "epoch": 4.98,
      "learning_rate": 2.5176056338028172e-06,
      "loss": 0.2783,
      "step": 7100
    },
    {
      "epoch": 4.98,
      "eval_accuracy": 0.7404999732971191,
      "eval_loss": 0.7649760842323303,
      "eval_runtime": 18.1063,
      "eval_samples_per_second": 110.459,
      "eval_steps_per_second": 6.904,
      "step": 7100
    },
    {
      "epoch": 5.05,
      "learning_rate": 2.482394366197183e-06,
      "loss": 0.2304,
      "step": 7200
    },
    {
      "epoch": 5.05,
      "eval_accuracy": 0.7384999990463257,
      "eval_loss": 0.8541684746742249,
      "eval_runtime": 20.2711,
      "eval_samples_per_second": 98.663,
      "eval_steps_per_second": 6.166,
      "step": 7200
    },
    {
      "epoch": 5.12,
      "learning_rate": 2.4471830985915495e-06,
      "loss": 0.1937,
      "step": 7300
    },
    {
      "epoch": 5.12,
      "eval_accuracy": 0.734499990940094,
      "eval_loss": 0.8389941453933716,
      "eval_runtime": 21.6926,
      "eval_samples_per_second": 92.197,
      "eval_steps_per_second": 5.762,
      "step": 7300
    },
    {
      "epoch": 5.19,
      "learning_rate": 2.411971830985916e-06,
      "loss": 0.1878,
      "step": 7400
    },
    {
      "epoch": 5.19,
      "eval_accuracy": 0.7329999804496765,
      "eval_loss": 0.9149684906005859,
      "eval_runtime": 19.4205,
      "eval_samples_per_second": 102.984,
      "eval_steps_per_second": 6.436,
      "step": 7400
    },
    {
      "epoch": 5.26,
      "learning_rate": 2.376760563380282e-06,
      "loss": 0.1921,
      "step": 7500
    },
    {
      "epoch": 5.26,
      "eval_accuracy": 0.7404999732971191,
      "eval_loss": 0.8792451024055481,
      "eval_runtime": 20.264,
      "eval_samples_per_second": 98.697,
      "eval_steps_per_second": 6.169,
      "step": 7500
    },
    {
      "epoch": 5.33,
      "learning_rate": 2.341549295774648e-06,
      "loss": 0.1916,
      "step": 7600
    },
    {
      "epoch": 5.33,
      "eval_accuracy": 0.7409999966621399,
      "eval_loss": 0.8891890645027161,
      "eval_runtime": 20.8102,
      "eval_samples_per_second": 96.107,
      "eval_steps_per_second": 6.007,
      "step": 7600
    },
    {
      "epoch": 5.4,
      "learning_rate": 2.306338028169014e-06,
      "loss": 0.2011,
      "step": 7700
    },
    {
      "epoch": 5.4,
      "eval_accuracy": 0.7325000166893005,
      "eval_loss": 0.9012252688407898,
      "eval_runtime": 21.4132,
      "eval_samples_per_second": 93.4,
      "eval_steps_per_second": 5.838,
      "step": 7700
    },
    {
      "epoch": 5.47,
      "learning_rate": 2.2711267605633805e-06,
      "loss": 0.211,
      "step": 7800
    },
    {
      "epoch": 5.47,
      "eval_accuracy": 0.7419999837875366,
      "eval_loss": 0.8607960343360901,
      "eval_runtime": 17.8954,
      "eval_samples_per_second": 111.761,
      "eval_steps_per_second": 6.985,
      "step": 7800
    },
    {
      "epoch": 5.54,
      "learning_rate": 2.235915492957747e-06,
      "loss": 0.2194,
      "step": 7900
    },
    {
      "epoch": 5.54,
      "eval_accuracy": 0.7319999933242798,
      "eval_loss": 0.8851566314697266,
      "eval_runtime": 12.4785,
      "eval_samples_per_second": 160.275,
      "eval_steps_per_second": 10.017,
      "step": 7900
    },
    {
      "epoch": 5.61,
      "learning_rate": 2.200704225352113e-06,
      "loss": 0.205,
      "step": 8000
    },
    {
      "epoch": 5.61,
      "eval_accuracy": 0.7384999990463257,
      "eval_loss": 0.8803377151489258,
      "eval_runtime": 17.7919,
      "eval_samples_per_second": 112.41,
      "eval_steps_per_second": 7.026,
      "step": 8000
    },
    {
      "epoch": 5.68,
      "learning_rate": 2.1654929577464787e-06,
      "loss": 0.1981,
      "step": 8100
    },
    {
      "epoch": 5.68,
      "eval_accuracy": 0.7329999804496765,
      "eval_loss": 0.86810302734375,
      "eval_runtime": 15.0816,
      "eval_samples_per_second": 132.612,
      "eval_steps_per_second": 8.288,
      "step": 8100
    },
    {
      "epoch": 5.75,
      "learning_rate": 2.130281690140845e-06,
      "loss": 0.1908,
      "step": 8200
    },
    {
      "epoch": 5.75,
      "eval_accuracy": 0.7434999942779541,
      "eval_loss": 0.9019960761070251,
      "eval_runtime": 16.0702,
      "eval_samples_per_second": 124.454,
      "eval_steps_per_second": 7.778,
      "step": 8200
    },
    {
      "epoch": 5.82,
      "learning_rate": 2.0950704225352115e-06,
      "loss": 0.1942,
      "step": 8300
    },
    {
      "epoch": 5.82,
      "eval_accuracy": 0.7409999966621399,
      "eval_loss": 0.8780096173286438,
      "eval_runtime": 18.9162,
      "eval_samples_per_second": 105.73,
      "eval_steps_per_second": 6.608,
      "step": 8300
    },
    {
      "epoch": 5.89,
      "learning_rate": 2.059859154929578e-06,
      "loss": 0.1958,
      "step": 8400
    },
    {
      "epoch": 5.89,
      "eval_accuracy": 0.734499990940094,
      "eval_loss": 0.8936640620231628,
      "eval_runtime": 19.3712,
      "eval_samples_per_second": 103.246,
      "eval_steps_per_second": 6.453,
      "step": 8400
    },
    {
      "epoch": 5.96,
      "learning_rate": 2.0246478873239438e-06,
      "loss": 0.1883,
      "step": 8500
    },
    {
      "epoch": 5.96,
      "eval_accuracy": 0.7360000014305115,
      "eval_loss": 0.9120668172836304,
      "eval_runtime": 17.1206,
      "eval_samples_per_second": 116.818,
      "eval_steps_per_second": 7.301,
      "step": 8500
    },
    {
      "epoch": 6.04,
      "learning_rate": 1.98943661971831e-06,
      "loss": 0.1819,
      "step": 8600
    },
    {
      "epoch": 6.04,
      "eval_accuracy": 0.7429999709129333,
      "eval_loss": 0.94089674949646,
      "eval_runtime": 22.6258,
      "eval_samples_per_second": 88.395,
      "eval_steps_per_second": 5.525,
      "step": 8600
    },
    {
      "epoch": 6.11,
      "learning_rate": 1.954225352112676e-06,
      "loss": 0.145,
      "step": 8700
    },
    {
      "epoch": 6.11,
      "eval_accuracy": 0.7264999747276306,
      "eval_loss": 1.1389663219451904,
      "eval_runtime": 15.9579,
      "eval_samples_per_second": 125.33,
      "eval_steps_per_second": 7.833,
      "step": 8700
    },
    {
      "epoch": 6.18,
      "learning_rate": 1.9190140845070424e-06,
      "loss": 0.1696,
      "step": 8800
    },
    {
      "epoch": 6.18,
      "eval_accuracy": 0.7429999709129333,
      "eval_loss": 0.9188500046730042,
      "eval_runtime": 21.2758,
      "eval_samples_per_second": 94.003,
      "eval_steps_per_second": 5.875,
      "step": 8800
    },
    {
      "epoch": 6.25,
      "learning_rate": 1.8838028169014086e-06,
      "loss": 0.1488,
      "step": 8900
    },
    {
      "epoch": 6.25,
      "eval_accuracy": 0.7400000095367432,
      "eval_loss": 0.9717501401901245,
      "eval_runtime": 17.3967,
      "eval_samples_per_second": 114.964,
      "eval_steps_per_second": 7.185,
      "step": 8900
    },
    {
      "epoch": 6.32,
      "learning_rate": 1.848591549295775e-06,
      "loss": 0.1637,
      "step": 9000
    },
    {
      "epoch": 6.32,
      "eval_accuracy": 0.7450000047683716,
      "eval_loss": 0.9701842069625854,
      "eval_runtime": 19.9143,
      "eval_samples_per_second": 100.43,
      "eval_steps_per_second": 6.277,
      "step": 9000
    },
    {
      "epoch": 6.39,
      "learning_rate": 1.813380281690141e-06,
      "loss": 0.1547,
      "step": 9100
    },
    {
      "epoch": 6.39,
      "eval_accuracy": 0.7409999966621399,
      "eval_loss": 1.0032541751861572,
      "eval_runtime": 21.7985,
      "eval_samples_per_second": 91.749,
      "eval_steps_per_second": 5.734,
      "step": 9100
    },
    {
      "epoch": 6.46,
      "learning_rate": 1.7781690140845072e-06,
      "loss": 0.1605,
      "step": 9200
    },
    {
      "epoch": 6.46,
      "eval_accuracy": 0.7354999780654907,
      "eval_loss": 0.99726402759552,
      "eval_runtime": 21.688,
      "eval_samples_per_second": 92.217,
      "eval_steps_per_second": 5.764,
      "step": 9200
    },
    {
      "epoch": 6.53,
      "learning_rate": 1.7429577464788734e-06,
      "loss": 0.1552,
      "step": 9300
    },
    {
      "epoch": 6.53,
      "eval_accuracy": 0.7289999723434448,
      "eval_loss": 1.0491423606872559,
      "eval_runtime": 20.7168,
      "eval_samples_per_second": 96.54,
      "eval_steps_per_second": 6.034,
      "step": 9300
    },
    {
      "epoch": 6.6,
      "learning_rate": 1.7077464788732395e-06,
      "loss": 0.1731,
      "step": 9400
    },
    {
      "epoch": 6.6,
      "eval_accuracy": 0.7335000038146973,
      "eval_loss": 1.027091145515442,
      "eval_runtime": 21.2758,
      "eval_samples_per_second": 94.003,
      "eval_steps_per_second": 5.875,
      "step": 9400
    },
    {
      "epoch": 6.67,
      "learning_rate": 1.6725352112676057e-06,
      "loss": 0.1738,
      "step": 9500
    },
    {
      "epoch": 6.67,
      "eval_accuracy": 0.7429999709129333,
      "eval_loss": 0.9575192928314209,
      "eval_runtime": 18.6703,
      "eval_samples_per_second": 107.122,
      "eval_steps_per_second": 6.695,
      "step": 9500
    },
    {
      "epoch": 6.74,
      "learning_rate": 1.637323943661972e-06,
      "loss": 0.1669,
      "step": 9600
    },
    {
      "epoch": 6.74,
      "eval_accuracy": 0.7350000143051147,
      "eval_loss": 0.9613668322563171,
      "eval_runtime": 21.3968,
      "eval_samples_per_second": 93.472,
      "eval_steps_per_second": 5.842,
      "step": 9600
    },
    {
      "epoch": 6.81,
      "learning_rate": 1.6021126760563382e-06,
      "loss": 0.1347,
      "step": 9700
    },
    {
      "epoch": 6.81,
      "eval_accuracy": 0.7365000247955322,
      "eval_loss": 1.0263434648513794,
      "eval_runtime": 19.7329,
      "eval_samples_per_second": 101.353,
      "eval_steps_per_second": 6.335,
      "step": 9700
    },
    {
      "epoch": 6.88,
      "learning_rate": 1.5669014084507045e-06,
      "loss": 0.1593,
      "step": 9800
    },
    {
      "epoch": 6.88,
      "eval_accuracy": 0.7360000014305115,
      "eval_loss": 1.017268180847168,
      "eval_runtime": 16.8592,
      "eval_samples_per_second": 118.63,
      "eval_steps_per_second": 7.414,
      "step": 9800
    },
    {
      "epoch": 6.95,
      "learning_rate": 1.5316901408450705e-06,
      "loss": 0.1549,
      "step": 9900
    },
    {
      "epoch": 6.95,
      "eval_accuracy": 0.7350000143051147,
      "eval_loss": 1.0397531986236572,
      "eval_runtime": 18.6472,
      "eval_samples_per_second": 107.255,
      "eval_steps_per_second": 6.703,
      "step": 9900
    },
    {
      "epoch": 7.02,
      "learning_rate": 1.4964788732394366e-06,
      "loss": 0.1675,
      "step": 10000
    },
    {
      "epoch": 7.02,
      "eval_accuracy": 0.7379999756813049,
      "eval_loss": 0.9975456595420837,
      "eval_runtime": 19.9355,
      "eval_samples_per_second": 100.323,
      "eval_steps_per_second": 6.27,
      "step": 10000
    },
    {
      "epoch": 7.09,
      "learning_rate": 1.461267605633803e-06,
      "loss": 0.1182,
      "step": 10100
    },
    {
      "epoch": 7.09,
      "eval_accuracy": 0.7350000143051147,
      "eval_loss": 1.105920672416687,
      "eval_runtime": 22.1229,
      "eval_samples_per_second": 90.404,
      "eval_steps_per_second": 5.65,
      "step": 10100
    },
    {
      "epoch": 7.16,
      "learning_rate": 1.4260563380281691e-06,
      "loss": 0.1351,
      "step": 10200
    },
    {
      "epoch": 7.16,
      "eval_accuracy": 0.7400000095367432,
      "eval_loss": 1.093347191810608,
      "eval_runtime": 20.9448,
      "eval_samples_per_second": 95.489,
      "eval_steps_per_second": 5.968,
      "step": 10200
    },
    {
      "epoch": 7.23,
      "learning_rate": 1.3908450704225355e-06,
      "loss": 0.1496,
      "step": 10300
    },
    {
      "epoch": 7.23,
      "eval_accuracy": 0.7354999780654907,
      "eval_loss": 1.0731019973754883,
      "eval_runtime": 17.3295,
      "eval_samples_per_second": 115.41,
      "eval_steps_per_second": 7.213,
      "step": 10300
    },
    {
      "epoch": 7.3,
      "learning_rate": 1.3556338028169017e-06,
      "loss": 0.1197,
      "step": 10400
    },
    {
      "epoch": 7.3,
      "eval_accuracy": 0.7360000014305115,
      "eval_loss": 1.1089140176773071,
      "eval_runtime": 19.6723,
      "eval_samples_per_second": 101.666,
      "eval_steps_per_second": 6.354,
      "step": 10400
    },
    {
      "epoch": 7.37,
      "learning_rate": 1.3204225352112676e-06,
      "loss": 0.1111,
      "step": 10500
    },
    {
      "epoch": 7.37,
      "eval_accuracy": 0.7404999732971191,
      "eval_loss": 1.1381380558013916,
      "eval_runtime": 18.6584,
      "eval_samples_per_second": 107.191,
      "eval_steps_per_second": 6.699,
      "step": 10500
    },
    {
      "epoch": 7.44,
      "learning_rate": 1.285211267605634e-06,
      "loss": 0.1494,
      "step": 10600
    },
    {
      "epoch": 7.44,
      "eval_accuracy": 0.7425000071525574,
      "eval_loss": 1.0251615047454834,
      "eval_runtime": 20.1427,
      "eval_samples_per_second": 99.292,
      "eval_steps_per_second": 6.206,
      "step": 10600
    },
    {
      "epoch": 7.51,
      "learning_rate": 1.25e-06,
      "loss": 0.1235,
      "step": 10700
    },
    {
      "epoch": 7.51,
      "eval_accuracy": 0.7360000014305115,
      "eval_loss": 1.0906413793563843,
      "eval_runtime": 22.5849,
      "eval_samples_per_second": 88.555,
      "eval_steps_per_second": 5.535,
      "step": 10700
    },
    {
      "epoch": 7.58,
      "learning_rate": 1.2147887323943663e-06,
      "loss": 0.133,
      "step": 10800
    },
    {
      "epoch": 7.58,
      "eval_accuracy": 0.737500011920929,
      "eval_loss": 1.1796296834945679,
      "eval_runtime": 12.3686,
      "eval_samples_per_second": 161.699,
      "eval_steps_per_second": 10.106,
      "step": 10800
    },
    {
      "epoch": 7.65,
      "learning_rate": 1.1795774647887324e-06,
      "loss": 0.1248,
      "step": 10900
    },
    {
      "epoch": 7.65,
      "eval_accuracy": 0.7419999837875366,
      "eval_loss": 1.1331868171691895,
      "eval_runtime": 21.2537,
      "eval_samples_per_second": 94.101,
      "eval_steps_per_second": 5.881,
      "step": 10900
    },
    {
      "epoch": 7.72,
      "learning_rate": 1.1443661971830988e-06,
      "loss": 0.1268,
      "step": 11000
    },
    {
      "epoch": 7.72,
      "eval_accuracy": 0.7415000200271606,
      "eval_loss": 1.1304017305374146,
      "eval_runtime": 20.2304,
      "eval_samples_per_second": 98.861,
      "eval_steps_per_second": 6.179,
      "step": 11000
    },
    {
      "epoch": 7.79,
      "learning_rate": 1.109154929577465e-06,
      "loss": 0.1368,
      "step": 11100
    },
    {
      "epoch": 7.79,
      "eval_accuracy": 0.7379999756813049,
      "eval_loss": 1.1345131397247314,
      "eval_runtime": 21.0499,
      "eval_samples_per_second": 95.013,
      "eval_steps_per_second": 5.938,
      "step": 11100
    },
    {
      "epoch": 7.86,
      "learning_rate": 1.073943661971831e-06,
      "loss": 0.1228,
      "step": 11200
    },
    {
      "epoch": 7.86,
      "eval_accuracy": 0.7319999933242798,
      "eval_loss": 1.2018308639526367,
      "eval_runtime": 21.3555,
      "eval_samples_per_second": 93.653,
      "eval_steps_per_second": 5.853,
      "step": 11200
    },
    {
      "epoch": 7.93,
      "learning_rate": 1.0387323943661972e-06,
      "loss": 0.1281,
      "step": 11300
    },
    {
      "epoch": 7.93,
      "eval_accuracy": 0.7350000143051147,
      "eval_loss": 1.1884474754333496,
      "eval_runtime": 18.4277,
      "eval_samples_per_second": 108.532,
      "eval_steps_per_second": 6.783,
      "step": 11300
    },
    {
      "epoch": 8.0,
      "learning_rate": 1.0035211267605636e-06,
      "loss": 0.1449,
      "step": 11400
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.734499990940094,
      "eval_loss": 1.157057762145996,
      "eval_runtime": 16.3477,
      "eval_samples_per_second": 122.341,
      "eval_steps_per_second": 7.646,
      "step": 11400
    },
    {
      "epoch": 8.07,
      "learning_rate": 9.683098591549295e-07,
      "loss": 0.1025,
      "step": 11500
    },
    {
      "epoch": 8.07,
      "eval_accuracy": 0.734499990940094,
      "eval_loss": 1.153812289237976,
      "eval_runtime": 15.7861,
      "eval_samples_per_second": 126.694,
      "eval_steps_per_second": 7.918,
      "step": 11500
    },
    {
      "epoch": 8.14,
      "learning_rate": 9.330985915492959e-07,
      "loss": 0.1199,
      "step": 11600
    },
    {
      "epoch": 8.14,
      "eval_accuracy": 0.7390000224113464,
      "eval_loss": 1.2113364934921265,
      "eval_runtime": 16.1478,
      "eval_samples_per_second": 123.856,
      "eval_steps_per_second": 7.741,
      "step": 11600
    },
    {
      "epoch": 8.21,
      "learning_rate": 8.978873239436621e-07,
      "loss": 0.1016,
      "step": 11700
    },
    {
      "epoch": 8.21,
      "eval_accuracy": 0.7369999885559082,
      "eval_loss": 1.2881745100021362,
      "eval_runtime": 23.0414,
      "eval_samples_per_second": 86.8,
      "eval_steps_per_second": 5.425,
      "step": 11700
    },
    {
      "epoch": 8.28,
      "learning_rate": 8.626760563380282e-07,
      "loss": 0.114,
      "step": 11800
    },
    {
      "epoch": 8.28,
      "eval_accuracy": 0.7390000224113464,
      "eval_loss": 1.287195086479187,
      "eval_runtime": 19.2083,
      "eval_samples_per_second": 104.122,
      "eval_steps_per_second": 6.508,
      "step": 11800
    },
    {
      "epoch": 8.35,
      "learning_rate": 8.274647887323944e-07,
      "loss": 0.1019,
      "step": 11900
    },
    {
      "epoch": 8.35,
      "eval_accuracy": 0.7379999756813049,
      "eval_loss": 1.287625789642334,
      "eval_runtime": 19.2034,
      "eval_samples_per_second": 104.148,
      "eval_steps_per_second": 6.509,
      "step": 11900
    },
    {
      "epoch": 8.42,
      "learning_rate": 7.922535211267607e-07,
      "loss": 0.1142,
      "step": 12000
    },
    {
      "epoch": 8.42,
      "eval_accuracy": 0.7384999990463257,
      "eval_loss": 1.2790753841400146,
      "eval_runtime": 12.6779,
      "eval_samples_per_second": 157.755,
      "eval_steps_per_second": 9.86,
      "step": 12000
    },
    {
      "epoch": 8.49,
      "learning_rate": 7.570422535211268e-07,
      "loss": 0.1135,
      "step": 12100
    },
    {
      "epoch": 8.49,
      "eval_accuracy": 0.7379999756813049,
      "eval_loss": 1.2882863283157349,
      "eval_runtime": 22.9048,
      "eval_samples_per_second": 87.318,
      "eval_steps_per_second": 5.457,
      "step": 12100
    },
    {
      "epoch": 8.56,
      "learning_rate": 7.21830985915493e-07,
      "loss": 0.1139,
      "step": 12200
    },
    {
      "epoch": 8.56,
      "eval_accuracy": 0.7360000014305115,
      "eval_loss": 1.2828530073165894,
      "eval_runtime": 12.6066,
      "eval_samples_per_second": 158.647,
      "eval_steps_per_second": 9.915,
      "step": 12200
    },
    {
      "epoch": 8.63,
      "learning_rate": 6.866197183098592e-07,
      "loss": 0.1107,
      "step": 12300
    },
    {
      "epoch": 8.63,
      "eval_accuracy": 0.7365000247955322,
      "eval_loss": 1.269805669784546,
      "eval_runtime": 18.0397,
      "eval_samples_per_second": 110.866,
      "eval_steps_per_second": 6.929,
      "step": 12300
    },
    {
      "epoch": 8.7,
      "learning_rate": 6.514084507042254e-07,
      "loss": 0.1183,
      "step": 12400
    },
    {
      "epoch": 8.7,
      "eval_accuracy": 0.734499990940094,
      "eval_loss": 1.266024112701416,
      "eval_runtime": 18.8559,
      "eval_samples_per_second": 106.068,
      "eval_steps_per_second": 6.629,
      "step": 12400
    },
    {
      "epoch": 8.77,
      "learning_rate": 6.161971830985916e-07,
      "loss": 0.1064,
      "step": 12500
    },
    {
      "epoch": 8.77,
      "eval_accuracy": 0.7365000247955322,
      "eval_loss": 1.288902759552002,
      "eval_runtime": 20.1463,
      "eval_samples_per_second": 99.274,
      "eval_steps_per_second": 6.205,
      "step": 12500
    },
    {
      "epoch": 8.84,
      "learning_rate": 5.809859154929578e-07,
      "loss": 0.0895,
      "step": 12600
    },
    {
      "epoch": 8.84,
      "eval_accuracy": 0.7329999804496765,
      "eval_loss": 1.3480335474014282,
      "eval_runtime": 19.9481,
      "eval_samples_per_second": 100.26,
      "eval_steps_per_second": 6.266,
      "step": 12600
    },
    {
      "epoch": 8.91,
      "learning_rate": 5.457746478873239e-07,
      "loss": 0.1244,
      "step": 12700
    },
    {
      "epoch": 8.91,
      "eval_accuracy": 0.7325000166893005,
      "eval_loss": 1.2872198820114136,
      "eval_runtime": 17.773,
      "eval_samples_per_second": 112.53,
      "eval_steps_per_second": 7.033,
      "step": 12700
    },
    {
      "epoch": 8.98,
      "learning_rate": 5.105633802816902e-07,
      "loss": 0.1209,
      "step": 12800
    },
    {
      "epoch": 8.98,
      "eval_accuracy": 0.737500011920929,
      "eval_loss": 1.2680846452713013,
      "eval_runtime": 22.2614,
      "eval_samples_per_second": 89.842,
      "eval_steps_per_second": 5.615,
      "step": 12800
    },
    {
      "epoch": 9.05,
      "learning_rate": 4.7535211267605635e-07,
      "loss": 0.1144,
      "step": 12900
    },
    {
      "epoch": 9.05,
      "eval_accuracy": 0.7369999885559082,
      "eval_loss": 1.2711447477340698,
      "eval_runtime": 18.1651,
      "eval_samples_per_second": 110.101,
      "eval_steps_per_second": 6.881,
      "step": 12900
    },
    {
      "epoch": 9.12,
      "learning_rate": 4.4014084507042255e-07,
      "loss": 0.1034,
      "step": 13000
    },
    {
      "epoch": 9.12,
      "eval_accuracy": 0.7360000014305115,
      "eval_loss": 1.2800805568695068,
      "eval_runtime": 20.275,
      "eval_samples_per_second": 98.644,
      "eval_steps_per_second": 6.165,
      "step": 13000
    },
    {
      "epoch": 9.19,
      "learning_rate": 4.049295774647888e-07,
      "loss": 0.113,
      "step": 13100
    },
    {
      "epoch": 9.19,
      "eval_accuracy": 0.7350000143051147,
      "eval_loss": 1.2801427841186523,
      "eval_runtime": 19.6919,
      "eval_samples_per_second": 101.565,
      "eval_steps_per_second": 6.348,
      "step": 13100
    },
    {
      "epoch": 9.26,
      "learning_rate": 3.6971830985915495e-07,
      "loss": 0.0994,
      "step": 13200
    },
    {
      "epoch": 9.26,
      "eval_accuracy": 0.7360000014305115,
      "eval_loss": 1.2920359373092651,
      "eval_runtime": 18.5273,
      "eval_samples_per_second": 107.949,
      "eval_steps_per_second": 6.747,
      "step": 13200
    },
    {
      "epoch": 9.33,
      "learning_rate": 3.345070422535211e-07,
      "loss": 0.0966,
      "step": 13300
    },
    {
      "epoch": 9.33,
      "eval_accuracy": 0.7335000038146973,
      "eval_loss": 1.2760682106018066,
      "eval_runtime": 21.2516,
      "eval_samples_per_second": 94.11,
      "eval_steps_per_second": 5.882,
      "step": 13300
    },
    {
      "epoch": 9.4,
      "learning_rate": 2.992957746478873e-07,
      "loss": 0.0939,
      "step": 13400
    },
    {
      "epoch": 9.4,
      "eval_accuracy": 0.7365000247955322,
      "eval_loss": 1.2908642292022705,
      "eval_runtime": 17.0257,
      "eval_samples_per_second": 117.47,
      "eval_steps_per_second": 7.342,
      "step": 13400
    },
    {
      "epoch": 9.47,
      "learning_rate": 2.6408450704225356e-07,
      "loss": 0.0975,
      "step": 13500
    },
    {
      "epoch": 9.47,
      "eval_accuracy": 0.7360000014305115,
      "eval_loss": 1.2952693700790405,
      "eval_runtime": 20.8537,
      "eval_samples_per_second": 95.906,
      "eval_steps_per_second": 5.994,
      "step": 13500
    },
    {
      "epoch": 9.54,
      "learning_rate": 2.2887323943661974e-07,
      "loss": 0.0842,
      "step": 13600
    },
    {
      "epoch": 9.54,
      "eval_accuracy": 0.7335000038146973,
      "eval_loss": 1.3179150819778442,
      "eval_runtime": 20.9509,
      "eval_samples_per_second": 95.461,
      "eval_steps_per_second": 5.966,
      "step": 13600
    },
    {
      "epoch": 9.61,
      "learning_rate": 1.936619718309859e-07,
      "loss": 0.0871,
      "step": 13700
    },
    {
      "epoch": 9.61,
      "eval_accuracy": 0.7384999990463257,
      "eval_loss": 1.314935564994812,
      "eval_runtime": 20.6989,
      "eval_samples_per_second": 96.623,
      "eval_steps_per_second": 6.039,
      "step": 13700
    },
    {
      "epoch": 9.68,
      "learning_rate": 1.5845070422535212e-07,
      "loss": 0.1162,
      "step": 13800
    },
    {
      "epoch": 9.68,
      "eval_accuracy": 0.7350000143051147,
      "eval_loss": 1.3124284744262695,
      "eval_runtime": 21.7736,
      "eval_samples_per_second": 91.854,
      "eval_steps_per_second": 5.741,
      "step": 13800
    },
    {
      "epoch": 9.75,
      "learning_rate": 1.2323943661971832e-07,
      "loss": 0.085,
      "step": 13900
    },
    {
      "epoch": 9.75,
      "eval_accuracy": 0.7354999780654907,
      "eval_loss": 1.3206626176834106,
      "eval_runtime": 18.1587,
      "eval_samples_per_second": 110.14,
      "eval_steps_per_second": 6.884,
      "step": 13900
    },
    {
      "epoch": 9.82,
      "learning_rate": 8.802816901408452e-08,
      "loss": 0.0966,
      "step": 14000
    },
    {
      "epoch": 9.82,
      "eval_accuracy": 0.7335000038146973,
      "eval_loss": 1.3247543573379517,
      "eval_runtime": 16.4512,
      "eval_samples_per_second": 121.572,
      "eval_steps_per_second": 7.598,
      "step": 14000
    },
    {
      "epoch": 9.89,
      "learning_rate": 5.281690140845071e-08,
      "loss": 0.1064,
      "step": 14100
    },
    {
      "epoch": 9.89,
      "eval_accuracy": 0.7335000038146973,
      "eval_loss": 1.3260987997055054,
      "eval_runtime": 22.2976,
      "eval_samples_per_second": 89.696,
      "eval_steps_per_second": 5.606,
      "step": 14100
    },
    {
      "epoch": 9.96,
      "learning_rate": 1.7605633802816902e-08,
      "loss": 0.1046,
      "step": 14200
    },
    {
      "epoch": 9.96,
      "eval_accuracy": 0.7360000014305115,
      "eval_loss": 1.3255339860916138,
      "eval_runtime": 19.2834,
      "eval_samples_per_second": 103.716,
      "eval_steps_per_second": 6.482,
      "step": 14200
    },
    {
      "epoch": 10.0,
      "step": 14250,
      "total_flos": 2.1254540922139392e+17,
      "train_loss": 0.2872312853629129,
      "train_runtime": 13159.309,
      "train_samples_per_second": 34.664,
      "train_steps_per_second": 1.083
    }
  ],
  "max_steps": 14250,
  "num_train_epochs": 10,
  "total_flos": 2.1254540922139392e+17,
  "trial_name": null,
  "trial_params": null
}