{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 22.997333333333334,
  "eval_steps": 500,
  "global_step": 4312,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.53,
      "learning_rate": 3e-05,
      "loss": 1.4597,
      "step": 100
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6618980392156862,
      "eval_loss": 1.2718806266784668,
      "eval_runtime": 7.6194,
      "eval_samples_per_second": 65.622,
      "eval_steps_per_second": 8.268,
      "step": 187
    },
    {
      "epoch": 1.0,
      "eval_exact_match": 22.4,
      "eval_f1": 30.941064425770318,
      "eval_qa_bleu": 7.097403515221558,
      "eval_qa_exact_match": 0.196,
      "eval_recite_bleu": 13.258373740102225,
      "eval_recite_exact_match": 0.0,
      "step": 187
    },
    {
      "epoch": 1.07,
      "learning_rate": 3e-05,
      "loss": 1.2905,
      "step": 200
    },
    {
      "epoch": 1.6,
      "learning_rate": 3e-05,
      "loss": 1.2157,
      "step": 300
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6670039215686274,
      "eval_loss": 1.206207036972046,
      "eval_runtime": 7.6793,
      "eval_samples_per_second": 65.11,
      "eval_steps_per_second": 8.204,
      "step": 375
    },
    {
      "epoch": 2.0,
      "eval_exact_match": 3.2,
      "eval_f1": 3.96976236704394,
      "eval_qa_bleu": 0.20484366294597936,
      "eval_qa_exact_match": 0.026,
      "eval_recite_bleu": 2.5778714769111824e-05,
      "eval_recite_exact_match": 0.0,
      "step": 375
    },
    {
      "epoch": 2.13,
      "learning_rate": 3e-05,
      "loss": 1.1852,
      "step": 400
    },
    {
      "epoch": 2.67,
      "learning_rate": 3e-05,
      "loss": 1.1861,
      "step": 500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.667513725490196,
      "eval_loss": 1.19828462600708,
      "eval_runtime": 7.1425,
      "eval_samples_per_second": 70.004,
      "eval_steps_per_second": 8.82,
      "step": 562
    },
    {
      "epoch": 3.0,
      "eval_exact_match": 4.4,
      "eval_f1": 6.333660342551587,
      "eval_qa_bleu": 0.522391008398391,
      "eval_qa_exact_match": 0.028,
      "eval_recite_bleu": 0.027569975851374308,
      "eval_recite_exact_match": 0.0,
      "step": 562
    },
    {
      "epoch": 3.2,
      "learning_rate": 3e-05,
      "loss": 1.1722,
      "step": 600
    },
    {
      "epoch": 3.73,
      "learning_rate": 3e-05,
      "loss": 1.1445,
      "step": 700
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6672,
      "eval_loss": 1.197709560394287,
      "eval_runtime": 7.7427,
      "eval_samples_per_second": 64.577,
      "eval_steps_per_second": 8.137,
      "step": 750
    },
    {
      "epoch": 4.0,
      "eval_exact_match": 7.4,
      "eval_f1": 10.950016966795278,
      "eval_qa_bleu": 0.6700083846206139,
      "eval_qa_exact_match": 0.052,
      "eval_recite_bleu": 1.259776140465531,
      "eval_recite_exact_match": 0.0,
      "step": 750
    },
    {
      "epoch": 4.27,
      "learning_rate": 3e-05,
      "loss": 1.1018,
      "step": 800
    },
    {
      "epoch": 4.8,
      "learning_rate": 3e-05,
      "loss": 1.1,
      "step": 900
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6664941176470588,
      "eval_loss": 1.2078713178634644,
      "eval_runtime": 7.7721,
      "eval_samples_per_second": 64.333,
      "eval_steps_per_second": 8.106,
      "step": 937
    },
    {
      "epoch": 5.0,
      "eval_exact_match": 6.6,
      "eval_f1": 10.12048028049545,
      "eval_qa_bleu": 0.8314625442591895,
      "eval_qa_exact_match": 0.05,
      "eval_recite_bleu": 0.8030246920772521,
      "eval_recite_exact_match": 0.0,
      "step": 937
    },
    {
      "epoch": 5.33,
      "learning_rate": 3e-05,
      "loss": 1.0666,
      "step": 1000
    },
    {
      "epoch": 5.87,
      "learning_rate": 3e-05,
      "loss": 1.0439,
      "step": 1100
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6649725490196079,
      "eval_loss": 1.22383713722229,
      "eval_runtime": 7.6639,
      "eval_samples_per_second": 65.241,
      "eval_steps_per_second": 8.22,
      "step": 1125
    },
    {
      "epoch": 6.0,
      "eval_exact_match": 9.0,
      "eval_f1": 14.19791549030679,
      "eval_qa_bleu": 2.7960558519683327,
      "eval_qa_exact_match": 0.07,
      "eval_recite_bleu": 2.082318255561628,
      "eval_recite_exact_match": 0.0,
      "step": 1125
    },
    {
      "epoch": 6.4,
      "learning_rate": 3e-05,
      "loss": 1.0029,
      "step": 1200
    },
    {
      "epoch": 6.93,
      "learning_rate": 3e-05,
      "loss": 0.9888,
      "step": 1300
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6638117647058823,
      "eval_loss": 1.2457345724105835,
      "eval_runtime": 6.7577,
      "eval_samples_per_second": 73.99,
      "eval_steps_per_second": 9.323,
      "step": 1312
    },
    {
      "epoch": 7.0,
      "eval_exact_match": 10.6,
      "eval_f1": 15.30586844512479,
      "eval_qa_bleu": 0.94556412718174,
      "eval_qa_exact_match": 0.086,
      "eval_recite_bleu": 3.1558565534273777,
      "eval_recite_exact_match": 0.0,
      "step": 1312
    },
    {
      "epoch": 7.47,
      "learning_rate": 3e-05,
      "loss": 0.9453,
      "step": 1400
    },
    {
      "epoch": 8.0,
      "learning_rate": 3e-05,
      "loss": 0.9364,
      "step": 1500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6615921568627451,
      "eval_loss": 1.2816376686096191,
      "eval_runtime": 7.5902,
      "eval_samples_per_second": 65.875,
      "eval_steps_per_second": 8.3,
      "step": 1500
    },
    {
      "epoch": 8.0,
      "eval_exact_match": 9.2,
      "eval_f1": 12.989213707726288,
      "eval_qa_bleu": 1.9797940311992752,
      "eval_qa_exact_match": 0.074,
      "eval_recite_bleu": 2.551621610315755,
      "eval_recite_exact_match": 0.0,
      "step": 1500
    },
    {
      "epoch": 8.53,
      "learning_rate": 3e-05,
      "loss": 0.8889,
      "step": 1600
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6606196078431372,
      "eval_loss": 1.3036144971847534,
      "eval_runtime": 7.4845,
      "eval_samples_per_second": 66.805,
      "eval_steps_per_second": 8.417,
      "step": 1687
    },
    {
      "epoch": 9.0,
      "eval_exact_match": 14.2,
      "eval_f1": 19.66897045942666,
      "eval_qa_bleu": 3.3530752528407004,
      "eval_qa_exact_match": 0.112,
      "eval_recite_bleu": 5.3293362575452425,
      "eval_recite_exact_match": 0.002,
      "step": 1687
    },
    {
      "epoch": 9.07,
      "learning_rate": 3e-05,
      "loss": 0.8853,
      "step": 1700
    },
    {
      "epoch": 9.6,
      "learning_rate": 3e-05,
      "loss": 0.8373,
      "step": 1800
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6587450980392157,
      "eval_loss": 1.3337457180023193,
      "eval_runtime": 6.6361,
      "eval_samples_per_second": 75.345,
      "eval_steps_per_second": 9.493,
      "step": 1875
    },
    {
      "epoch": 10.0,
      "eval_exact_match": 14.4,
      "eval_f1": 21.20886352105531,
      "eval_qa_bleu": 2.50339041197013,
      "eval_qa_exact_match": 0.106,
      "eval_recite_bleu": 6.648771889251858,
      "eval_recite_exact_match": 0.0,
      "step": 1875
    },
    {
      "epoch": 10.13,
      "learning_rate": 3e-05,
      "loss": 0.8353,
      "step": 1900
    },
    {
      "epoch": 10.67,
      "learning_rate": 3e-05,
      "loss": 0.7949,
      "step": 2000
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.6575058823529412,
      "eval_loss": 1.367775797843933,
      "eval_runtime": 7.4976,
      "eval_samples_per_second": 66.688,
      "eval_steps_per_second": 8.403,
      "step": 2062
    },
    {
      "epoch": 11.0,
      "eval_exact_match": 15.4,
      "eval_f1": 21.670913900913916,
      "eval_qa_bleu": 4.131625289761948,
      "eval_qa_exact_match": 0.104,
      "eval_recite_bleu": 8.111027123751208,
      "eval_recite_exact_match": 0.0,
      "step": 2062
    },
    {
      "epoch": 11.2,
      "learning_rate": 3e-05,
      "loss": 0.7758,
      "step": 2100
    },
    {
      "epoch": 11.73,
      "learning_rate": 3e-05,
      "loss": 0.7539,
      "step": 2200
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.6553725490196078,
      "eval_loss": 1.3984336853027344,
      "eval_runtime": 7.6892,
      "eval_samples_per_second": 65.026,
      "eval_steps_per_second": 8.193,
      "step": 2250
    },
    {
      "epoch": 12.0,
      "eval_exact_match": 16.2,
      "eval_f1": 23.074922854922868,
      "eval_qa_bleu": 10.80394366395148,
      "eval_qa_exact_match": 0.132,
      "eval_recite_bleu": 9.158207459098465,
      "eval_recite_exact_match": 0.002,
      "step": 2250
    },
    {
      "epoch": 12.27,
      "learning_rate": 3e-05,
      "loss": 0.7415,
      "step": 2300
    },
    {
      "epoch": 12.8,
      "learning_rate": 3e-05,
      "loss": 0.7133,
      "step": 2400
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.6538196078431373,
      "eval_loss": 1.4471032619476318,
      "eval_runtime": 6.7177,
      "eval_samples_per_second": 74.431,
      "eval_steps_per_second": 9.378,
      "step": 2437
    },
    {
      "epoch": 13.0,
      "eval_exact_match": 16.4,
      "eval_f1": 23.343965796845062,
      "eval_qa_bleu": 4.473837470175788,
      "eval_qa_exact_match": 0.136,
      "eval_recite_bleu": 10.58390244263022,
      "eval_recite_exact_match": 0.002,
      "step": 2437
    },
    {
      "epoch": 13.33,
      "learning_rate": 3e-05,
      "loss": 0.682,
      "step": 2500
    },
    {
      "epoch": 13.87,
      "learning_rate": 3e-05,
      "loss": 0.6704,
      "step": 2600
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.6524862745098039,
      "eval_loss": 1.4829775094985962,
      "eval_runtime": 6.6339,
      "eval_samples_per_second": 75.37,
      "eval_steps_per_second": 9.497,
      "step": 2625
    },
    {
      "epoch": 14.0,
      "eval_exact_match": 15.6,
      "eval_f1": 23.615572621393074,
      "eval_qa_bleu": 7.563435997390864,
      "eval_qa_exact_match": 0.13,
      "eval_recite_bleu": 11.120038652898813,
      "eval_recite_exact_match": 0.008,
      "step": 2625
    },
    {
      "epoch": 14.4,
      "learning_rate": 3e-05,
      "loss": 0.6412,
      "step": 2700
    },
    {
      "epoch": 14.93,
      "learning_rate": 3e-05,
      "loss": 0.6436,
      "step": 2800
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.6507529411764705,
      "eval_loss": 1.524308681488037,
      "eval_runtime": 6.6481,
      "eval_samples_per_second": 75.209,
      "eval_steps_per_second": 9.476,
      "step": 2812
    },
    {
      "epoch": 15.0,
      "eval_exact_match": 17.8,
      "eval_f1": 27.00021164021165,
      "eval_qa_bleu": 7.90349748901623,
      "eval_qa_exact_match": 0.15,
      "eval_recite_bleu": 12.277483381668418,
      "eval_recite_exact_match": 0.008,
      "step": 2812
    },
    {
      "epoch": 15.47,
      "learning_rate": 3e-05,
      "loss": 0.5969,
      "step": 2900
    },
    {
      "epoch": 16.0,
      "learning_rate": 3e-05,
      "loss": 0.6073,
      "step": 3000
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.6499764705882353,
      "eval_loss": 1.5640995502471924,
      "eval_runtime": 7.7119,
      "eval_samples_per_second": 64.835,
      "eval_steps_per_second": 8.169,
      "step": 3000
    },
    {
      "epoch": 16.0,
      "eval_exact_match": 19.2,
      "eval_f1": 27.747719298245627,
      "eval_qa_bleu": 9.793788404463436,
      "eval_qa_exact_match": 0.14,
      "eval_recite_bleu": 12.835117221544726,
      "eval_recite_exact_match": 0.01,
      "step": 3000
    },
    {
      "epoch": 16.53,
      "learning_rate": 3e-05,
      "loss": 0.5567,
      "step": 3100
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.6486588235294117,
      "eval_loss": 1.6058622598648071,
      "eval_runtime": 7.7727,
      "eval_samples_per_second": 64.328,
      "eval_steps_per_second": 8.105,
      "step": 3187
    },
    {
      "epoch": 17.0,
      "eval_exact_match": 19.6,
      "eval_f1": 26.446776949066113,
      "eval_qa_bleu": 10.294098070146385,
      "eval_qa_exact_match": 0.156,
      "eval_recite_bleu": 12.34815708077889,
      "eval_recite_exact_match": 0.004,
      "step": 3187
    },
    {
      "epoch": 17.07,
      "learning_rate": 3e-05,
      "loss": 0.5654,
      "step": 3200
    },
    {
      "epoch": 17.6,
      "learning_rate": 3e-05,
      "loss": 0.5204,
      "step": 3300
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.6472705882352942,
      "eval_loss": 1.6425700187683105,
      "eval_runtime": 7.1347,
      "eval_samples_per_second": 70.08,
      "eval_steps_per_second": 8.83,
      "step": 3375
    },
    {
      "epoch": 18.0,
      "eval_exact_match": 18.4,
      "eval_f1": 26.083665223665236,
      "eval_qa_bleu": 12.978225559731756,
      "eval_qa_exact_match": 0.142,
      "eval_recite_bleu": 12.629003802964625,
      "eval_recite_exact_match": 0.006,
      "step": 3375
    },
    {
      "epoch": 18.13,
      "learning_rate": 3e-05,
      "loss": 0.5348,
      "step": 3400
    },
    {
      "epoch": 18.67,
      "learning_rate": 3e-05,
      "loss": 0.5019,
      "step": 3500
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.6459843137254901,
      "eval_loss": 1.6983567476272583,
      "eval_runtime": 7.0372,
      "eval_samples_per_second": 71.051,
      "eval_steps_per_second": 8.952,
      "step": 3562
    },
    {
      "epoch": 19.0,
      "eval_exact_match": 19.4,
      "eval_f1": 27.658198757763987,
      "eval_qa_bleu": 10.765170656210177,
      "eval_qa_exact_match": 0.146,
      "eval_recite_bleu": 12.874816956011239,
      "eval_recite_exact_match": 0.006,
      "step": 3562
    },
    {
      "epoch": 19.2,
      "learning_rate": 3e-05,
      "loss": 0.4888,
      "step": 3600
    },
    {
      "epoch": 19.73,
      "learning_rate": 3e-05,
      "loss": 0.4676,
      "step": 3700
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.6450431372549019,
      "eval_loss": 1.74748957157135,
      "eval_runtime": 6.6386,
      "eval_samples_per_second": 75.317,
      "eval_steps_per_second": 9.49,
      "step": 3750
    },
    {
      "epoch": 20.0,
      "eval_exact_match": 17.4,
      "eval_f1": 25.5869051909052,
      "eval_qa_bleu": 7.975401244418332,
      "eval_qa_exact_match": 0.144,
      "eval_recite_bleu": 11.978547399136012,
      "eval_recite_exact_match": 0.004,
      "step": 3750
    },
    {
      "epoch": 20.27,
      "learning_rate": 3e-05,
      "loss": 0.4528,
      "step": 3800
    },
    {
      "epoch": 20.8,
      "learning_rate": 3e-05,
      "loss": 0.4408,
      "step": 3900
    },
    {
      "epoch": 21.0,
      "eval_accuracy": 0.6450274509803922,
      "eval_loss": 1.7759969234466553,
      "eval_runtime": 7.7092,
      "eval_samples_per_second": 64.857,
      "eval_steps_per_second": 8.172,
      "step": 3937
    },
    {
      "epoch": 21.0,
      "eval_exact_match": 19.0,
      "eval_f1": 27.044218949435447,
      "eval_qa_bleu": 5.961857744132503,
      "eval_qa_exact_match": 0.158,
      "eval_recite_bleu": 12.078600703350125,
      "eval_recite_exact_match": 0.004,
      "step": 3937
    },
    {
      "epoch": 21.33,
      "learning_rate": 3e-05,
      "loss": 0.4191,
      "step": 4000
    },
    {
      "epoch": 21.87,
      "learning_rate": 3e-05,
      "loss": 0.4217,
      "step": 4100
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.644078431372549,
      "eval_loss": 1.7986315488815308,
      "eval_runtime": 7.6507,
      "eval_samples_per_second": 65.353,
      "eval_steps_per_second": 8.235,
      "step": 4125
    },
    {
      "epoch": 22.0,
      "eval_exact_match": 18.2,
      "eval_f1": 25.124761904761915,
      "eval_qa_bleu": 15.392825669562168,
      "eval_qa_exact_match": 0.14,
      "eval_recite_bleu": 11.50341909728548,
      "eval_recite_exact_match": 0.006,
      "step": 4125
    },
    {
      "epoch": 22.4,
      "learning_rate": 3e-05,
      "loss": 0.393,
      "step": 4200
    },
    {
      "epoch": 22.93,
      "learning_rate": 3e-05,
      "loss": 0.3886,
      "step": 4300
    },
    {
      "epoch": 23.0,
      "eval_accuracy": 0.6439529411764706,
      "eval_loss": 1.8378911018371582,
      "eval_runtime": 7.4779,
      "eval_samples_per_second": 66.864,
      "eval_steps_per_second": 8.425,
      "step": 4312
    },
    {
      "epoch": 23.0,
      "eval_exact_match": 14.8,
      "eval_f1": 21.914444444444456,
      "eval_qa_bleu": 10.979013053975578,
      "eval_qa_exact_match": 0.128,
      "eval_recite_bleu": 11.265801595707984,
      "eval_recite_exact_match": 0.004,
      "step": 4312
    }
  ],
  "logging_steps": 100,
  "max_steps": 9350,
  "num_train_epochs": 50,
  "save_steps": 500,
  "total_flos": 9.994816220194406e+17,
  "trial_name": null,
  "trial_params": null
}