{
  "best_metric": 0.20996317267417908,
  "best_model_checkpoint": "./model_outputs/checkpoint-9750",
  "epoch": 1.9240878505041734,
  "eval_steps": 50,
  "global_step": 9750,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009866408824516053,
      "grad_norm": 0.5440589189529419,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.8768,
      "step": 50
    },
    {
      "epoch": 0.009866408824516053,
      "eval_loss": 1.8065277338027954,
      "eval_runtime": 0.199,
      "eval_samples_per_second": 25.12,
      "eval_steps_per_second": 5.024,
      "step": 50
    },
    {
      "epoch": 0.019732817649032106,
      "grad_norm": 0.5126028656959534,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.6013,
      "step": 100
    },
    {
      "epoch": 0.019732817649032106,
      "eval_loss": 1.4184890985488892,
      "eval_runtime": 0.1842,
      "eval_samples_per_second": 27.145,
      "eval_steps_per_second": 5.429,
      "step": 100
    },
    {
      "epoch": 0.02959922647354816,
      "grad_norm": 0.8953598141670227,
      "learning_rate": 2.5e-05,
      "loss": 1.3394,
      "step": 150
    },
    {
      "epoch": 0.02959922647354816,
      "eval_loss": 1.2211127281188965,
      "eval_runtime": 0.1845,
      "eval_samples_per_second": 27.095,
      "eval_steps_per_second": 5.419,
      "step": 150
    },
    {
      "epoch": 0.03946563529806421,
      "grad_norm": 1.0633267164230347,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.2225,
      "step": 200
    },
    {
      "epoch": 0.03946563529806421,
      "eval_loss": 1.1311841011047363,
      "eval_runtime": 0.1844,
      "eval_samples_per_second": 27.117,
      "eval_steps_per_second": 5.423,
      "step": 200
    },
    {
      "epoch": 0.049332044122580264,
      "grad_norm": 1.340325951576233,
      "learning_rate": 4.166666666666667e-05,
      "loss": 1.1527,
      "step": 250
    },
    {
      "epoch": 0.049332044122580264,
      "eval_loss": 1.0357060432434082,
      "eval_runtime": 0.1844,
      "eval_samples_per_second": 27.116,
      "eval_steps_per_second": 5.423,
      "step": 250
    },
    {
      "epoch": 0.05919845294709632,
      "grad_norm": 1.1488324403762817,
      "learning_rate": 5e-05,
      "loss": 1.098,
      "step": 300
    },
    {
      "epoch": 0.05919845294709632,
      "eval_loss": 0.9791008830070496,
      "eval_runtime": 0.1852,
      "eval_samples_per_second": 27.002,
      "eval_steps_per_second": 5.4,
      "step": 300
    },
    {
      "epoch": 0.06906486177161236,
      "grad_norm": 1.3122004270553589,
      "learning_rate": 5.833333333333334e-05,
      "loss": 1.0511,
      "step": 350
    },
    {
      "epoch": 0.06906486177161236,
      "eval_loss": 0.9438408017158508,
      "eval_runtime": 0.1844,
      "eval_samples_per_second": 27.119,
      "eval_steps_per_second": 5.424,
      "step": 350
    },
    {
      "epoch": 0.07893127059612842,
      "grad_norm": 1.4071924686431885,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.0327,
      "step": 400
    },
    {
      "epoch": 0.07893127059612842,
      "eval_loss": 0.9280341863632202,
      "eval_runtime": 0.1839,
      "eval_samples_per_second": 27.193,
      "eval_steps_per_second": 5.439,
      "step": 400
    },
    {
      "epoch": 0.08879767942064447,
      "grad_norm": 1.447715163230896,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.9924,
      "step": 450
    },
    {
      "epoch": 0.08879767942064447,
      "eval_loss": 0.9004782438278198,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.09,
      "eval_steps_per_second": 5.418,
      "step": 450
    },
    {
      "epoch": 0.09866408824516053,
      "grad_norm": 1.6754121780395508,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.9313,
      "step": 500
    },
    {
      "epoch": 0.09866408824516053,
      "eval_loss": 0.7985298037528992,
      "eval_runtime": 0.1849,
      "eval_samples_per_second": 27.04,
      "eval_steps_per_second": 5.408,
      "step": 500
    },
    {
      "epoch": 0.10853049706967657,
      "grad_norm": 1.7146815061569214,
      "learning_rate": 9.166666666666667e-05,
      "loss": 0.7311,
      "step": 550
    },
    {
      "epoch": 0.10853049706967657,
      "eval_loss": 0.5294134020805359,
      "eval_runtime": 0.1845,
      "eval_samples_per_second": 27.107,
      "eval_steps_per_second": 5.421,
      "step": 550
    },
    {
      "epoch": 0.11839690589419263,
      "grad_norm": 1.3303102254867554,
      "learning_rate": 0.0001,
      "loss": 0.5309,
      "step": 600
    },
    {
      "epoch": 0.11839690589419263,
      "eval_loss": 0.42509689927101135,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.075,
      "eval_steps_per_second": 5.415,
      "step": 600
    },
    {
      "epoch": 0.1282633147187087,
      "grad_norm": 0.9372690320014954,
      "learning_rate": 0.00010833333333333333,
      "loss": 0.4645,
      "step": 650
    },
    {
      "epoch": 0.1282633147187087,
      "eval_loss": 0.37647953629493713,
      "eval_runtime": 0.1851,
      "eval_samples_per_second": 27.009,
      "eval_steps_per_second": 5.402,
      "step": 650
    },
    {
      "epoch": 0.13812972354322473,
      "grad_norm": 0.9396585822105408,
      "learning_rate": 0.00011666666666666668,
      "loss": 0.4258,
      "step": 700
    },
    {
      "epoch": 0.13812972354322473,
      "eval_loss": 0.3551653325557709,
      "eval_runtime": 0.1853,
      "eval_samples_per_second": 26.982,
      "eval_steps_per_second": 5.396,
      "step": 700
    },
    {
      "epoch": 0.14799613236774078,
      "grad_norm": 0.6549014449119568,
      "learning_rate": 0.000125,
      "loss": 0.4101,
      "step": 750
    },
    {
      "epoch": 0.14799613236774078,
      "eval_loss": 0.3371706008911133,
      "eval_runtime": 0.1856,
      "eval_samples_per_second": 26.941,
      "eval_steps_per_second": 5.388,
      "step": 750
    },
    {
      "epoch": 0.15786254119225684,
      "grad_norm": 0.5201817750930786,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.3877,
      "step": 800
    },
    {
      "epoch": 0.15786254119225684,
      "eval_loss": 0.3209473192691803,
      "eval_runtime": 0.1844,
      "eval_samples_per_second": 27.12,
      "eval_steps_per_second": 5.424,
      "step": 800
    },
    {
      "epoch": 0.1677289500167729,
      "grad_norm": 0.44689691066741943,
      "learning_rate": 0.00014166666666666668,
      "loss": 0.3856,
      "step": 850
    },
    {
      "epoch": 0.1677289500167729,
      "eval_loss": 0.3198350965976715,
      "eval_runtime": 0.1841,
      "eval_samples_per_second": 27.157,
      "eval_steps_per_second": 5.431,
      "step": 850
    },
    {
      "epoch": 0.17759535884128894,
      "grad_norm": 0.4665811359882355,
      "learning_rate": 0.00015000000000000001,
      "loss": 0.3815,
      "step": 900
    },
    {
      "epoch": 0.17759535884128894,
      "eval_loss": 0.31239432096481323,
      "eval_runtime": 0.1858,
      "eval_samples_per_second": 26.91,
      "eval_steps_per_second": 5.382,
      "step": 900
    },
    {
      "epoch": 0.187461767665805,
      "grad_norm": 0.40847715735435486,
      "learning_rate": 0.00015833333333333332,
      "loss": 0.3678,
      "step": 950
    },
    {
      "epoch": 0.187461767665805,
      "eval_loss": 0.31479328870773315,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.067,
      "eval_steps_per_second": 5.413,
      "step": 950
    },
    {
      "epoch": 0.19732817649032106,
      "grad_norm": 0.4807054400444031,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.3663,
      "step": 1000
    },
    {
      "epoch": 0.19732817649032106,
      "eval_loss": 0.317302405834198,
      "eval_runtime": 0.1849,
      "eval_samples_per_second": 27.036,
      "eval_steps_per_second": 5.407,
      "step": 1000
    },
    {
      "epoch": 0.20719458531483712,
      "grad_norm": 0.40431880950927734,
      "learning_rate": 0.000175,
      "loss": 0.3618,
      "step": 1050
    },
    {
      "epoch": 0.20719458531483712,
      "eval_loss": 0.30631712079048157,
      "eval_runtime": 0.1843,
      "eval_samples_per_second": 27.129,
      "eval_steps_per_second": 5.426,
      "step": 1050
    },
    {
      "epoch": 0.21706099413935315,
      "grad_norm": 0.3184848725795746,
      "learning_rate": 0.00018333333333333334,
      "loss": 0.3585,
      "step": 1100
    },
    {
      "epoch": 0.21706099413935315,
      "eval_loss": 0.3047851026058197,
      "eval_runtime": 0.1842,
      "eval_samples_per_second": 27.139,
      "eval_steps_per_second": 5.428,
      "step": 1100
    },
    {
      "epoch": 0.2269274029638692,
      "grad_norm": 0.29751336574554443,
      "learning_rate": 0.00019166666666666667,
      "loss": 0.3545,
      "step": 1150
    },
    {
      "epoch": 0.2269274029638692,
      "eval_loss": 0.30102381110191345,
      "eval_runtime": 0.1859,
      "eval_samples_per_second": 26.901,
      "eval_steps_per_second": 5.38,
      "step": 1150
    },
    {
      "epoch": 0.23679381178838527,
      "grad_norm": 0.3572952449321747,
      "learning_rate": 0.0002,
      "loss": 0.354,
      "step": 1200
    },
    {
      "epoch": 0.23679381178838527,
      "eval_loss": 0.2969129979610443,
      "eval_runtime": 0.185,
      "eval_samples_per_second": 27.03,
      "eval_steps_per_second": 5.406,
      "step": 1200
    },
    {
      "epoch": 0.24666022061290133,
      "grad_norm": 0.397197961807251,
      "learning_rate": 0.00019998454365958754,
      "loss": 0.3505,
      "step": 1250
    },
    {
      "epoch": 0.24666022061290133,
      "eval_loss": 0.2986981272697449,
      "eval_runtime": 0.1868,
      "eval_samples_per_second": 26.768,
      "eval_steps_per_second": 5.354,
      "step": 1250
    },
    {
      "epoch": 0.2565266294374174,
      "grad_norm": 0.31810271739959717,
      "learning_rate": 0.00019993817941631932,
      "loss": 0.3469,
      "step": 1300
    },
    {
      "epoch": 0.2565266294374174,
      "eval_loss": 0.2904866635799408,
      "eval_runtime": 0.361,
      "eval_samples_per_second": 13.851,
      "eval_steps_per_second": 2.77,
      "step": 1300
    },
    {
      "epoch": 0.26639303826193345,
      "grad_norm": 0.3378036618232727,
      "learning_rate": 0.00019986092160262587,
      "loss": 0.3465,
      "step": 1350
    },
    {
      "epoch": 0.26639303826193345,
      "eval_loss": 0.29250431060791016,
      "eval_runtime": 0.1881,
      "eval_samples_per_second": 26.582,
      "eval_steps_per_second": 5.316,
      "step": 1350
    },
    {
      "epoch": 0.27625944708644945,
      "grad_norm": 0.25885072350502014,
      "learning_rate": 0.00019975279410096856,
      "loss": 0.3427,
      "step": 1400
    },
    {
      "epoch": 0.27625944708644945,
      "eval_loss": 0.2970002293586731,
      "eval_runtime": 0.1858,
      "eval_samples_per_second": 26.911,
      "eval_steps_per_second": 5.382,
      "step": 1400
    },
    {
      "epoch": 0.2861258559109655,
      "grad_norm": 0.37681177258491516,
      "learning_rate": 0.00019961383033645683,
      "loss": 0.3423,
      "step": 1450
    },
    {
      "epoch": 0.2861258559109655,
      "eval_loss": 0.28687533736228943,
      "eval_runtime": 0.1869,
      "eval_samples_per_second": 26.751,
      "eval_steps_per_second": 5.35,
      "step": 1450
    },
    {
      "epoch": 0.29599226473548157,
      "grad_norm": 0.33231380581855774,
      "learning_rate": 0.00019944407326651575,
      "loss": 0.3388,
      "step": 1500
    },
    {
      "epoch": 0.29599226473548157,
      "eval_loss": 0.28474920988082886,
      "eval_runtime": 0.1871,
      "eval_samples_per_second": 26.718,
      "eval_steps_per_second": 5.344,
      "step": 1500
    },
    {
      "epoch": 0.30585867355999763,
      "grad_norm": 0.3391518294811249,
      "learning_rate": 0.00019924357536760644,
      "loss": 0.3316,
      "step": 1550
    },
    {
      "epoch": 0.30585867355999763,
      "eval_loss": 0.27619096636772156,
      "eval_runtime": 0.2,
      "eval_samples_per_second": 25.0,
      "eval_steps_per_second": 5.0,
      "step": 1550
    },
    {
      "epoch": 0.3157250823845137,
      "grad_norm": 0.3423280715942383,
      "learning_rate": 0.0001990123986190045,
      "loss": 0.3376,
      "step": 1600
    },
    {
      "epoch": 0.3157250823845137,
      "eval_loss": 0.2841600179672241,
      "eval_runtime": 0.1876,
      "eval_samples_per_second": 26.647,
      "eval_steps_per_second": 5.329,
      "step": 1600
    },
    {
      "epoch": 0.32559149120902975,
      "grad_norm": 0.37309399247169495,
      "learning_rate": 0.00019875061448364033,
      "loss": 0.3357,
      "step": 1650
    },
    {
      "epoch": 0.32559149120902975,
      "eval_loss": 0.2811121344566345,
      "eval_runtime": 0.1875,
      "eval_samples_per_second": 26.669,
      "eval_steps_per_second": 5.334,
      "step": 1650
    },
    {
      "epoch": 0.3354579000335458,
      "grad_norm": 0.33649030327796936,
      "learning_rate": 0.00019845830388600822,
      "loss": 0.3294,
      "step": 1700
    },
    {
      "epoch": 0.3354579000335458,
      "eval_loss": 0.27847614884376526,
      "eval_runtime": 0.1869,
      "eval_samples_per_second": 26.754,
      "eval_steps_per_second": 5.351,
      "step": 1700
    },
    {
      "epoch": 0.34532430885806187,
      "grad_norm": 0.38383936882019043,
      "learning_rate": 0.00019813555718715012,
      "loss": 0.334,
      "step": 1750
    },
    {
      "epoch": 0.34532430885806187,
      "eval_loss": 0.2873836159706116,
      "eval_runtime": 0.1865,
      "eval_samples_per_second": 26.806,
      "eval_steps_per_second": 5.361,
      "step": 1750
    },
    {
      "epoch": 0.3551907176825779,
      "grad_norm": 0.3089899718761444,
      "learning_rate": 0.000197782474156723,
      "loss": 0.3275,
      "step": 1800
    },
    {
      "epoch": 0.3551907176825779,
      "eval_loss": 0.2808569669723511,
      "eval_runtime": 0.1864,
      "eval_samples_per_second": 26.824,
      "eval_steps_per_second": 5.365,
      "step": 1800
    },
    {
      "epoch": 0.36505712650709393,
      "grad_norm": 0.3028016686439514,
      "learning_rate": 0.0001973991639421571,
      "loss": 0.3303,
      "step": 1850
    },
    {
      "epoch": 0.36505712650709393,
      "eval_loss": 0.27654123306274414,
      "eval_runtime": 0.1859,
      "eval_samples_per_second": 26.894,
      "eval_steps_per_second": 5.379,
      "step": 1850
    },
    {
      "epoch": 0.37492353533161,
      "grad_norm": 0.2978042960166931,
      "learning_rate": 0.0001969857450349156,
      "loss": 0.3304,
      "step": 1900
    },
    {
      "epoch": 0.37492353533161,
      "eval_loss": 0.270946741104126,
      "eval_runtime": 0.1856,
      "eval_samples_per_second": 26.941,
      "eval_steps_per_second": 5.388,
      "step": 1900
    },
    {
      "epoch": 0.38478994415612605,
      "grad_norm": 0.3676602244377136,
      "learning_rate": 0.00019654234523386576,
      "loss": 0.3319,
      "step": 1950
    },
    {
      "epoch": 0.38478994415612605,
      "eval_loss": 0.26677393913269043,
      "eval_runtime": 0.1857,
      "eval_samples_per_second": 26.927,
      "eval_steps_per_second": 5.385,
      "step": 1950
    },
    {
      "epoch": 0.3946563529806421,
      "grad_norm": 0.243007093667984,
      "learning_rate": 0.00019606910160577286,
      "loss": 0.3253,
      "step": 2000
    },
    {
      "epoch": 0.3946563529806421,
      "eval_loss": 0.26983901858329773,
      "eval_runtime": 0.186,
      "eval_samples_per_second": 26.888,
      "eval_steps_per_second": 5.378,
      "step": 2000
    },
    {
      "epoch": 0.40452276180515817,
      "grad_norm": 0.2672829329967499,
      "learning_rate": 0.00019556616044292917,
      "loss": 0.3269,
      "step": 2050
    },
    {
      "epoch": 0.40452276180515817,
      "eval_loss": 0.2751110792160034,
      "eval_runtime": 0.187,
      "eval_samples_per_second": 26.74,
      "eval_steps_per_second": 5.348,
      "step": 2050
    },
    {
      "epoch": 0.41438917062967423,
      "grad_norm": 0.26455241441726685,
      "learning_rate": 0.00019503367721793112,
      "loss": 0.3269,
      "step": 2100
    },
    {
      "epoch": 0.41438917062967423,
      "eval_loss": 0.2611739933490753,
      "eval_runtime": 0.1862,
      "eval_samples_per_second": 26.854,
      "eval_steps_per_second": 5.371,
      "step": 2100
    },
    {
      "epoch": 0.4242555794541903,
      "grad_norm": 0.260708749294281,
      "learning_rate": 0.00019447181653561851,
      "loss": 0.3264,
      "step": 2150
    },
    {
      "epoch": 0.4242555794541903,
      "eval_loss": 0.25694364309310913,
      "eval_runtime": 0.1866,
      "eval_samples_per_second": 26.801,
      "eval_steps_per_second": 5.36,
      "step": 2150
    },
    {
      "epoch": 0.4341219882787063,
      "grad_norm": 0.3702332079410553,
      "learning_rate": 0.00019388075208219072,
      "loss": 0.3205,
      "step": 2200
    },
    {
      "epoch": 0.4341219882787063,
      "eval_loss": 0.2542697489261627,
      "eval_runtime": 0.1859,
      "eval_samples_per_second": 26.897,
      "eval_steps_per_second": 5.379,
      "step": 2200
    },
    {
      "epoch": 0.44398839710322235,
      "grad_norm": 0.2797427177429199,
      "learning_rate": 0.00019326066657151568,
      "loss": 0.3251,
      "step": 2250
    },
    {
      "epoch": 0.44398839710322235,
      "eval_loss": 0.2655983865261078,
      "eval_runtime": 0.186,
      "eval_samples_per_second": 26.886,
      "eval_steps_per_second": 5.377,
      "step": 2250
    },
    {
      "epoch": 0.4538548059277384,
      "grad_norm": 0.27329573035240173,
      "learning_rate": 0.00019261175168864823,
      "loss": 0.3203,
      "step": 2300
    },
    {
      "epoch": 0.4538548059277384,
      "eval_loss": 0.2548620402812958,
      "eval_runtime": 0.1859,
      "eval_samples_per_second": 26.897,
      "eval_steps_per_second": 5.379,
      "step": 2300
    },
    {
      "epoch": 0.4637212147522545,
      "grad_norm": 0.2299811691045761,
      "learning_rate": 0.00019193420803057483,
      "loss": 0.3171,
      "step": 2350
    },
    {
      "epoch": 0.4637212147522545,
      "eval_loss": 0.25404179096221924,
      "eval_runtime": 0.1869,
      "eval_samples_per_second": 26.755,
      "eval_steps_per_second": 5.351,
      "step": 2350
    },
    {
      "epoch": 0.47358762357677053,
      "grad_norm": 0.43620172142982483,
      "learning_rate": 0.00019122824504420402,
      "loss": 0.3192,
      "step": 2400
    },
    {
      "epoch": 0.47358762357677053,
      "eval_loss": 0.2546399235725403,
      "eval_runtime": 0.1862,
      "eval_samples_per_second": 26.85,
      "eval_steps_per_second": 5.37,
      "step": 2400
    },
    {
      "epoch": 0.4834540324012866,
      "grad_norm": 0.2562660574913025,
      "learning_rate": 0.0001904940809616205,
      "loss": 0.3204,
      "step": 2450
    },
    {
      "epoch": 0.4834540324012866,
      "eval_loss": 0.25131484866142273,
      "eval_runtime": 0.1861,
      "eval_samples_per_second": 26.861,
      "eval_steps_per_second": 5.372,
      "step": 2450
    },
    {
      "epoch": 0.49332044122580265,
      "grad_norm": 0.21843916177749634,
      "learning_rate": 0.0001897319427326239,
      "loss": 0.3211,
      "step": 2500
    },
    {
      "epoch": 0.49332044122580265,
      "eval_loss": 0.2540796399116516,
      "eval_runtime": 0.1862,
      "eval_samples_per_second": 26.851,
      "eval_steps_per_second": 5.37,
      "step": 2500
    },
    {
      "epoch": 0.5031868500503187,
      "grad_norm": 0.2513650059700012,
      "learning_rate": 0.00018894206595457228,
      "loss": 0.3172,
      "step": 2550
    },
    {
      "epoch": 0.5031868500503187,
      "eval_loss": 0.254707932472229,
      "eval_runtime": 0.1862,
      "eval_samples_per_second": 26.851,
      "eval_steps_per_second": 5.37,
      "step": 2550
    },
    {
      "epoch": 0.5130532588748348,
      "grad_norm": 0.2910959720611572,
      "learning_rate": 0.00018812469479955306,
      "loss": 0.319,
      "step": 2600
    },
    {
      "epoch": 0.5130532588748348,
      "eval_loss": 0.2549395263195038,
      "eval_runtime": 0.186,
      "eval_samples_per_second": 26.877,
      "eval_steps_per_second": 5.375,
      "step": 2600
    },
    {
      "epoch": 0.5229196676993508,
      "grad_norm": 0.36300212144851685,
      "learning_rate": 0.00018728008193890248,
      "loss": 0.3141,
      "step": 2650
    },
    {
      "epoch": 0.5229196676993508,
      "eval_loss": 0.25633907318115234,
      "eval_runtime": 0.1866,
      "eval_samples_per_second": 26.789,
      "eval_steps_per_second": 5.358,
      "step": 2650
    },
    {
      "epoch": 0.5327860765238669,
      "grad_norm": 0.27469244599342346,
      "learning_rate": 0.00018640848846509836,
      "loss": 0.3165,
      "step": 2700
    },
    {
      "epoch": 0.5327860765238669,
      "eval_loss": 0.25254935026168823,
      "eval_runtime": 0.1868,
      "eval_samples_per_second": 26.772,
      "eval_steps_per_second": 5.354,
      "step": 2700
    },
    {
      "epoch": 0.5426524853483828,
      "grad_norm": 0.24992810189723969,
      "learning_rate": 0.00018551018381104942,
      "loss": 0.3148,
      "step": 2750
    },
    {
      "epoch": 0.5426524853483828,
      "eval_loss": 0.2522678077220917,
      "eval_runtime": 0.187,
      "eval_samples_per_second": 26.739,
      "eval_steps_per_second": 5.348,
      "step": 2750
    },
    {
      "epoch": 0.5525188941728989,
      "grad_norm": 0.22494597733020782,
      "learning_rate": 0.00018458544566680613,
      "loss": 0.3178,
      "step": 2800
    },
    {
      "epoch": 0.5525188941728989,
      "eval_loss": 0.2535417973995209,
      "eval_runtime": 0.1861,
      "eval_samples_per_second": 26.866,
      "eval_steps_per_second": 5.373,
      "step": 2800
    },
    {
      "epoch": 0.562385302997415,
      "grad_norm": 0.2611429989337921,
      "learning_rate": 0.0001836345598937195,
      "loss": 0.3151,
      "step": 2850
    },
    {
      "epoch": 0.562385302997415,
      "eval_loss": 0.2576417326927185,
      "eval_runtime": 0.1858,
      "eval_samples_per_second": 26.917,
      "eval_steps_per_second": 5.383,
      "step": 2850
    },
    {
      "epoch": 0.572251711821931,
      "grad_norm": 0.2593046724796295,
      "learning_rate": 0.00018265782043607362,
      "loss": 0.3155,
      "step": 2900
    },
    {
      "epoch": 0.572251711821931,
      "eval_loss": 0.24795055389404297,
      "eval_runtime": 0.1873,
      "eval_samples_per_second": 26.699,
      "eval_steps_per_second": 5.34,
      "step": 2900
    },
    {
      "epoch": 0.5821181206464471,
      "grad_norm": 0.2805608808994293,
      "learning_rate": 0.00018165552923021946,
      "loss": 0.3084,
      "step": 2950
    },
    {
      "epoch": 0.5821181206464471,
      "eval_loss": 0.2520078122615814,
      "eval_runtime": 0.186,
      "eval_samples_per_second": 26.883,
      "eval_steps_per_second": 5.377,
      "step": 2950
    },
    {
      "epoch": 0.5919845294709631,
      "grad_norm": 0.2355504184961319,
      "learning_rate": 0.00018062799611123843,
      "loss": 0.3135,
      "step": 3000
    },
    {
      "epoch": 0.5919845294709631,
      "eval_loss": 0.2500859797000885,
      "eval_runtime": 0.1863,
      "eval_samples_per_second": 26.834,
      "eval_steps_per_second": 5.367,
      "step": 3000
    },
    {
      "epoch": 0.6018509382954792,
      "grad_norm": 0.23624199628829956,
      "learning_rate": 0.00017957553871716405,
      "loss": 0.3063,
      "step": 3050
    },
    {
      "epoch": 0.6018509382954792,
      "eval_loss": 0.24056918919086456,
      "eval_runtime": 0.1859,
      "eval_samples_per_second": 26.894,
      "eval_steps_per_second": 5.379,
      "step": 3050
    },
    {
      "epoch": 0.6117173471199953,
      "grad_norm": 0.2305937111377716,
      "learning_rate": 0.00017849848239079126,
      "loss": 0.3076,
      "step": 3100
    },
    {
      "epoch": 0.6117173471199953,
      "eval_loss": 0.2453293353319168,
      "eval_runtime": 0.1852,
      "eval_samples_per_second": 26.996,
      "eval_steps_per_second": 5.399,
      "step": 3100
    },
    {
      "epoch": 0.6215837559445113,
      "grad_norm": 0.28438571095466614,
      "learning_rate": 0.00017739716007910458,
      "loss": 0.309,
      "step": 3150
    },
    {
      "epoch": 0.6215837559445113,
      "eval_loss": 0.241606667637825,
      "eval_runtime": 0.1859,
      "eval_samples_per_second": 26.891,
      "eval_steps_per_second": 5.378,
      "step": 3150
    },
    {
      "epoch": 0.6314501647690274,
      "grad_norm": 0.31287047266960144,
      "learning_rate": 0.00017627191223035512,
      "loss": 0.3123,
      "step": 3200
    },
    {
      "epoch": 0.6314501647690274,
      "eval_loss": 0.24313075840473175,
      "eval_runtime": 0.1856,
      "eval_samples_per_second": 26.939,
      "eval_steps_per_second": 5.388,
      "step": 3200
    },
    {
      "epoch": 0.6413165735935434,
      "grad_norm": 0.2634814381599426,
      "learning_rate": 0.00017512308668881876,
      "loss": 0.3093,
      "step": 3250
    },
    {
      "epoch": 0.6413165735935434,
      "eval_loss": 0.24376794695854187,
      "eval_runtime": 0.1861,
      "eval_samples_per_second": 26.872,
      "eval_steps_per_second": 5.374,
      "step": 3250
    },
    {
      "epoch": 0.6511829824180595,
      "grad_norm": 0.22829458117485046,
      "learning_rate": 0.00017395103858726846,
      "loss": 0.3093,
      "step": 3300
    },
    {
      "epoch": 0.6511829824180595,
      "eval_loss": 0.2428242713212967,
      "eval_runtime": 0.1881,
      "eval_samples_per_second": 26.589,
      "eval_steps_per_second": 5.318,
      "step": 3300
    },
    {
      "epoch": 0.6610493912425756,
      "grad_norm": 0.26152193546295166,
      "learning_rate": 0.00017275613023719298,
      "loss": 0.3113,
      "step": 3350
    },
    {
      "epoch": 0.6610493912425756,
      "eval_loss": 0.24297213554382324,
      "eval_runtime": 0.1864,
      "eval_samples_per_second": 26.821,
      "eval_steps_per_second": 5.364,
      "step": 3350
    },
    {
      "epoch": 0.6709158000670916,
      "grad_norm": 0.21779882907867432,
      "learning_rate": 0.00017153873101679668,
      "loss": 0.3085,
      "step": 3400
    },
    {
      "epoch": 0.6709158000670916,
      "eval_loss": 0.24597088992595673,
      "eval_runtime": 0.1862,
      "eval_samples_per_second": 26.854,
      "eval_steps_per_second": 5.371,
      "step": 3400
    },
    {
      "epoch": 0.6807822088916077,
      "grad_norm": 0.21129541099071503,
      "learning_rate": 0.00017029921725681492,
      "loss": 0.3045,
      "step": 3450
    },
    {
      "epoch": 0.6807822088916077,
      "eval_loss": 0.23454919457435608,
      "eval_runtime": 0.1867,
      "eval_samples_per_second": 26.786,
      "eval_steps_per_second": 5.357,
      "step": 3450
    },
    {
      "epoch": 0.6906486177161237,
      "grad_norm": 0.19822391867637634,
      "learning_rate": 0.00016903797212418015,
      "loss": 0.3049,
      "step": 3500
    },
    {
      "epoch": 0.6906486177161237,
      "eval_loss": 0.23835961520671844,
      "eval_runtime": 0.1872,
      "eval_samples_per_second": 26.708,
      "eval_steps_per_second": 5.342,
      "step": 3500
    },
    {
      "epoch": 0.7005150265406397,
      "grad_norm": 0.2132054716348648,
      "learning_rate": 0.00016775538550357463,
      "loss": 0.3058,
      "step": 3550
    },
    {
      "epoch": 0.7005150265406397,
      "eval_loss": 0.2351282387971878,
      "eval_runtime": 0.1862,
      "eval_samples_per_second": 26.857,
      "eval_steps_per_second": 5.371,
      "step": 3550
    },
    {
      "epoch": 0.7103814353651557,
      "grad_norm": 0.23732738196849823,
      "learning_rate": 0.0001664518538769067,
      "loss": 0.3021,
      "step": 3600
    },
    {
      "epoch": 0.7103814353651557,
      "eval_loss": 0.23832817375659943,
      "eval_runtime": 0.1857,
      "eval_samples_per_second": 26.93,
      "eval_steps_per_second": 5.386,
      "step": 3600
    },
    {
      "epoch": 0.7202478441896718,
      "grad_norm": 0.3427148759365082,
      "learning_rate": 0.00016512778020074753,
      "loss": 0.3044,
      "step": 3650
    },
    {
      "epoch": 0.7202478441896718,
      "eval_loss": 0.24248237907886505,
      "eval_runtime": 0.1861,
      "eval_samples_per_second": 26.863,
      "eval_steps_per_second": 5.373,
      "step": 3650
    },
    {
      "epoch": 0.7301142530141879,
      "grad_norm": 0.33537736535072327,
      "learning_rate": 0.00016378357378176654,
      "loss": 0.3004,
      "step": 3700
    },
    {
      "epoch": 0.7301142530141879,
      "eval_loss": 0.23529915511608124,
      "eval_runtime": 0.185,
      "eval_samples_per_second": 27.026,
      "eval_steps_per_second": 5.405,
      "step": 3700
    },
    {
      "epoch": 0.7399806618387039,
      "grad_norm": 0.236049085855484,
      "learning_rate": 0.00016241965015020363,
      "loss": 0.3047,
      "step": 3750
    },
    {
      "epoch": 0.7399806618387039,
      "eval_loss": 0.23639431595802307,
      "eval_runtime": 0.1858,
      "eval_samples_per_second": 26.914,
      "eval_steps_per_second": 5.383,
      "step": 3750
    },
    {
      "epoch": 0.74984707066322,
      "grad_norm": 0.23827500641345978,
      "learning_rate": 0.0001610364309314178,
      "loss": 0.2991,
      "step": 3800
    },
    {
      "epoch": 0.74984707066322,
      "eval_loss": 0.23711517453193665,
      "eval_runtime": 0.1845,
      "eval_samples_per_second": 27.098,
      "eval_steps_per_second": 5.42,
      "step": 3800
    },
    {
      "epoch": 0.759713479487736,
      "grad_norm": 0.24710151553153992,
      "learning_rate": 0.00015963434371555116,
      "loss": 0.2993,
      "step": 3850
    },
    {
      "epoch": 0.759713479487736,
      "eval_loss": 0.2302863895893097,
      "eval_runtime": 0.1857,
      "eval_samples_per_second": 26.929,
      "eval_steps_per_second": 5.386,
      "step": 3850
    },
    {
      "epoch": 0.7695798883122521,
      "grad_norm": 0.18992306292057037,
      "learning_rate": 0.00015821382192534968,
      "loss": 0.3022,
      "step": 3900
    },
    {
      "epoch": 0.7695798883122521,
      "eval_loss": 0.23323240876197815,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.071,
      "eval_steps_per_second": 5.414,
      "step": 3900
    },
    {
      "epoch": 0.7794462971367682,
      "grad_norm": 0.2224888652563095,
      "learning_rate": 0.00015677530468218045,
      "loss": 0.3028,
      "step": 3950
    },
    {
      "epoch": 0.7794462971367682,
      "eval_loss": 0.2351154088973999,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.091,
      "eval_steps_per_second": 5.418,
      "step": 3950
    },
    {
      "epoch": 0.7893127059612842,
      "grad_norm": 0.2259168177843094,
      "learning_rate": 0.0001553192366702874,
      "loss": 0.3019,
      "step": 4000
    },
    {
      "epoch": 0.7893127059612842,
      "eval_loss": 0.227354496717453,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.09,
      "eval_steps_per_second": 5.418,
      "step": 4000
    },
    {
      "epoch": 0.7991791147858003,
      "grad_norm": 0.291294127702713,
      "learning_rate": 0.0001538460679993277,
      "loss": 0.2989,
      "step": 4050
    },
    {
      "epoch": 0.7991791147858003,
      "eval_loss": 0.24032792448997498,
      "eval_runtime": 0.185,
      "eval_samples_per_second": 27.022,
      "eval_steps_per_second": 5.404,
      "step": 4050
    },
    {
      "epoch": 0.8090455236103163,
      "grad_norm": 0.23510493338108063,
      "learning_rate": 0.00015235625406523058,
      "loss": 0.3004,
      "step": 4100
    },
    {
      "epoch": 0.8090455236103163,
      "eval_loss": 0.2415897101163864,
      "eval_runtime": 0.1848,
      "eval_samples_per_second": 27.05,
      "eval_steps_per_second": 5.41,
      "step": 4100
    },
    {
      "epoch": 0.8189119324348324,
      "grad_norm": 0.21880796551704407,
      "learning_rate": 0.00015085025540942238,
      "loss": 0.3,
      "step": 4150
    },
    {
      "epoch": 0.8189119324348324,
      "eval_loss": 0.24045029282569885,
      "eval_runtime": 0.1849,
      "eval_samples_per_second": 27.045,
      "eval_steps_per_second": 5.409,
      "step": 4150
    },
    {
      "epoch": 0.8287783412593485,
      "grad_norm": 0.24291107058525085,
      "learning_rate": 0.0001493285375764608,
      "loss": 0.2976,
      "step": 4200
    },
    {
      "epoch": 0.8287783412593485,
      "eval_loss": 0.24324026703834534,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.067,
      "eval_steps_per_second": 5.413,
      "step": 4200
    },
    {
      "epoch": 0.8386447500838645,
      "grad_norm": 0.2263229787349701,
      "learning_rate": 0.0001477915709701226,
      "loss": 0.298,
      "step": 4250
    },
    {
      "epoch": 0.8386447500838645,
      "eval_loss": 0.2447391003370285,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.068,
      "eval_steps_per_second": 5.414,
      "step": 4250
    },
    {
      "epoch": 0.8485111589083806,
      "grad_norm": 0.20994020998477936,
      "learning_rate": 0.00014623983070798918,
      "loss": 0.297,
      "step": 4300
    },
    {
      "epoch": 0.8485111589083806,
      "eval_loss": 0.24018962681293488,
      "eval_runtime": 0.1843,
      "eval_samples_per_second": 27.127,
      "eval_steps_per_second": 5.425,
      "step": 4300
    },
    {
      "epoch": 0.8583775677328965,
      "grad_norm": 0.19051186740398407,
      "learning_rate": 0.000144673796474575,
      "loss": 0.2929,
      "step": 4350
    },
    {
      "epoch": 0.8583775677328965,
      "eval_loss": 0.23791976273059845,
      "eval_runtime": 0.1849,
      "eval_samples_per_second": 27.037,
      "eval_steps_per_second": 5.407,
      "step": 4350
    },
    {
      "epoch": 0.8682439765574126,
      "grad_norm": 0.23745004832744598,
      "learning_rate": 0.00014309395237304426,
      "loss": 0.2989,
      "step": 4400
    },
    {
      "epoch": 0.8682439765574126,
      "eval_loss": 0.23674488067626953,
      "eval_runtime": 0.1858,
      "eval_samples_per_second": 26.904,
      "eval_steps_per_second": 5.381,
      "step": 4400
    },
    {
      "epoch": 0.8781103853819286,
      "grad_norm": 0.19539549946784973,
      "learning_rate": 0.0001415007867755616,
      "loss": 0.2972,
      "step": 4450
    },
    {
      "epoch": 0.8781103853819286,
      "eval_loss": 0.23721548914909363,
      "eval_runtime": 0.1843,
      "eval_samples_per_second": 27.128,
      "eval_steps_per_second": 5.426,
      "step": 4450
    },
    {
      "epoch": 0.8879767942064447,
      "grad_norm": 0.26372581720352173,
      "learning_rate": 0.00013989479217232315,
      "loss": 0.2989,
      "step": 4500
    },
    {
      "epoch": 0.8879767942064447,
      "eval_loss": 0.2338230162858963,
      "eval_runtime": 0.1851,
      "eval_samples_per_second": 27.014,
      "eval_steps_per_second": 5.403,
      "step": 4500
    },
    {
      "epoch": 0.8978432030309608,
      "grad_norm": 0.2618955373764038,
      "learning_rate": 0.00013827646501931472,
      "loss": 0.2896,
      "step": 4550
    },
    {
      "epoch": 0.8978432030309608,
      "eval_loss": 0.23268970847129822,
      "eval_runtime": 0.1842,
      "eval_samples_per_second": 27.148,
      "eval_steps_per_second": 5.43,
      "step": 4550
    },
    {
      "epoch": 0.9077096118554768,
      "grad_norm": 0.2664230167865753,
      "learning_rate": 0.00013664630558484379,
      "loss": 0.2889,
      "step": 4600
    },
    {
      "epoch": 0.9077096118554768,
      "eval_loss": 0.22873499989509583,
      "eval_runtime": 0.1845,
      "eval_samples_per_second": 27.094,
      "eval_steps_per_second": 5.419,
      "step": 4600
    },
    {
      "epoch": 0.9175760206799929,
      "grad_norm": 0.21087364852428436,
      "learning_rate": 0.00013500481779489326,
      "loss": 0.2958,
      "step": 4650
    },
    {
      "epoch": 0.9175760206799929,
      "eval_loss": 0.23380012810230255,
      "eval_runtime": 0.1848,
      "eval_samples_per_second": 27.061,
      "eval_steps_per_second": 5.412,
      "step": 4650
    },
    {
      "epoch": 0.927442429504509,
      "grad_norm": 0.21377532184123993,
      "learning_rate": 0.00013335250907734448,
      "loss": 0.296,
      "step": 4700
    },
    {
      "epoch": 0.927442429504509,
      "eval_loss": 0.23471400141716003,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.075,
      "eval_steps_per_second": 5.415,
      "step": 4700
    },
    {
      "epoch": 0.937308838329025,
      "grad_norm": 0.2032233327627182,
      "learning_rate": 0.0001316898902051175,
      "loss": 0.2927,
      "step": 4750
    },
    {
      "epoch": 0.937308838329025,
      "eval_loss": 0.23120686411857605,
      "eval_runtime": 0.185,
      "eval_samples_per_second": 27.026,
      "eval_steps_per_second": 5.405,
      "step": 4750
    },
    {
      "epoch": 0.9471752471535411,
      "grad_norm": 0.22421005368232727,
      "learning_rate": 0.00013001747513827764,
      "loss": 0.2956,
      "step": 4800
    },
    {
      "epoch": 0.9471752471535411,
      "eval_loss": 0.23076245188713074,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.079,
      "eval_steps_per_second": 5.416,
      "step": 4800
    },
    {
      "epoch": 0.9570416559780571,
      "grad_norm": 0.2466699182987213,
      "learning_rate": 0.0001283357808651566,
      "loss": 0.2907,
      "step": 4850
    },
    {
      "epoch": 0.9570416559780571,
      "eval_loss": 0.22873035073280334,
      "eval_runtime": 0.1843,
      "eval_samples_per_second": 27.128,
      "eval_steps_per_second": 5.426,
      "step": 4850
    },
    {
      "epoch": 0.9669080648025732,
      "grad_norm": 0.20787620544433594,
      "learning_rate": 0.00012664532724253745,
      "loss": 0.2954,
      "step": 4900
    },
    {
      "epoch": 0.9669080648025732,
      "eval_loss": 0.22716188430786133,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.076,
      "eval_steps_per_second": 5.415,
      "step": 4900
    },
    {
      "epoch": 0.9767744736270892,
      "grad_norm": 0.22663253545761108,
      "learning_rate": 0.00012494663683495304,
      "loss": 0.2897,
      "step": 4950
    },
    {
      "epoch": 0.9767744736270892,
      "eval_loss": 0.2276424914598465,
      "eval_runtime": 0.1848,
      "eval_samples_per_second": 27.05,
      "eval_steps_per_second": 5.41,
      "step": 4950
    },
    {
      "epoch": 0.9866408824516053,
      "grad_norm": 0.20295187830924988,
      "learning_rate": 0.00012324023475314725,
      "loss": 0.2958,
      "step": 5000
    },
    {
      "epoch": 0.9866408824516053,
      "eval_loss": 0.23117908835411072,
      "eval_runtime": 0.185,
      "eval_samples_per_second": 27.024,
      "eval_steps_per_second": 5.405,
      "step": 5000
    },
    {
      "epoch": 0.9965072912761214,
      "grad_norm": 0.2138393521308899,
      "learning_rate": 0.0001215266484917493,
      "loss": 0.2876,
      "step": 5050
    },
    {
      "epoch": 0.9965072912761214,
      "eval_loss": 0.22664275765419006,
      "eval_runtime": 0.1852,
      "eval_samples_per_second": 27.002,
      "eval_steps_per_second": 5.4,
      "step": 5050
    },
    {
      "epoch": 1.0063737001006374,
      "grad_norm": 0.1731952726840973,
      "learning_rate": 0.00011980640776621077,
      "loss": 0.2843,
      "step": 5100
    },
    {
      "epoch": 1.0063737001006374,
      "eval_loss": 0.22769823670387268,
      "eval_runtime": 0.185,
      "eval_samples_per_second": 27.027,
      "eval_steps_per_second": 5.405,
      "step": 5100
    },
    {
      "epoch": 1.0162401089251534,
      "grad_norm": 0.21639849245548248,
      "learning_rate": 0.0001180800443490566,
      "loss": 0.2744,
      "step": 5150
    },
    {
      "epoch": 1.0162401089251534,
      "eval_loss": 0.23013372719287872,
      "eval_runtime": 0.1844,
      "eval_samples_per_second": 27.117,
      "eval_steps_per_second": 5.423,
      "step": 5150
    },
    {
      "epoch": 1.0261065177496695,
      "grad_norm": 0.2546159625053406,
      "learning_rate": 0.0001163480919054998,
      "loss": 0.2745,
      "step": 5200
    },
    {
      "epoch": 1.0261065177496695,
      "eval_loss": 0.22475233674049377,
      "eval_runtime": 0.1858,
      "eval_samples_per_second": 26.911,
      "eval_steps_per_second": 5.382,
      "step": 5200
    },
    {
      "epoch": 1.0359729265741855,
      "grad_norm": 0.21476522088050842,
      "learning_rate": 0.00011461108582847126,
      "loss": 0.2734,
      "step": 5250
    },
    {
      "epoch": 1.0359729265741855,
      "eval_loss": 0.22389909625053406,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.07,
      "eval_steps_per_second": 5.414,
      "step": 5250
    },
    {
      "epoch": 1.0458393353987017,
      "grad_norm": 0.21301063895225525,
      "learning_rate": 0.00011286956307311555,
      "loss": 0.2703,
      "step": 5300
    },
    {
      "epoch": 1.0458393353987017,
      "eval_loss": 0.2267945110797882,
      "eval_runtime": 0.1843,
      "eval_samples_per_second": 27.135,
      "eval_steps_per_second": 5.427,
      "step": 5300
    },
    {
      "epoch": 1.0557057442232176,
      "grad_norm": 0.221356600522995,
      "learning_rate": 0.00011112406199080346,
      "loss": 0.2717,
      "step": 5350
    },
    {
      "epoch": 1.0557057442232176,
      "eval_loss": 0.2249325066804886,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.078,
      "eval_steps_per_second": 5.416,
      "step": 5350
    },
    {
      "epoch": 1.0655721530477338,
      "grad_norm": 0.22443543374538422,
      "learning_rate": 0.00010937512216271338,
      "loss": 0.2705,
      "step": 5400
    },
    {
      "epoch": 1.0655721530477338,
      "eval_loss": 0.2256511151790619,
      "eval_runtime": 0.1845,
      "eval_samples_per_second": 27.097,
      "eval_steps_per_second": 5.419,
      "step": 5400
    },
    {
      "epoch": 1.0754385618722497,
      "grad_norm": 0.20267891883850098,
      "learning_rate": 0.00010762328423303218,
      "loss": 0.2735,
      "step": 5450
    },
    {
      "epoch": 1.0754385618722497,
      "eval_loss": 0.22567923367023468,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.086,
      "eval_steps_per_second": 5.417,
      "step": 5450
    },
    {
      "epoch": 1.085304970696766,
      "grad_norm": 0.21884013712406158,
      "learning_rate": 0.00010586908974182767,
      "loss": 0.2756,
      "step": 5500
    },
    {
      "epoch": 1.085304970696766,
      "eval_loss": 0.22433491051197052,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.077,
      "eval_steps_per_second": 5.415,
      "step": 5500
    },
    {
      "epoch": 1.0951713795212819,
      "grad_norm": 0.21301575005054474,
      "learning_rate": 0.00010411308095764393,
      "loss": 0.2714,
      "step": 5550
    },
    {
      "epoch": 1.0951713795212819,
      "eval_loss": 0.22806711494922638,
      "eval_runtime": 0.1844,
      "eval_samples_per_second": 27.113,
      "eval_steps_per_second": 5.423,
      "step": 5550
    },
    {
      "epoch": 1.1050377883457978,
      "grad_norm": 0.3255505859851837,
      "learning_rate": 0.0001023558007098717,
      "loss": 0.2715,
      "step": 5600
    },
    {
      "epoch": 1.1050377883457978,
      "eval_loss": 0.22649523615837097,
      "eval_runtime": 0.1848,
      "eval_samples_per_second": 27.059,
      "eval_steps_per_second": 5.412,
      "step": 5600
    },
    {
      "epoch": 1.114904197170314,
      "grad_norm": 0.30551499128341675,
      "learning_rate": 0.00010059779222094512,
      "loss": 0.2737,
      "step": 5650
    },
    {
      "epoch": 1.114904197170314,
      "eval_loss": 0.2232808619737625,
      "eval_runtime": 0.1871,
      "eval_samples_per_second": 26.725,
      "eval_steps_per_second": 5.345,
      "step": 5650
    },
    {
      "epoch": 1.12477060599483,
      "grad_norm": 0.23381146788597107,
      "learning_rate": 9.88395989384173e-05,
      "loss": 0.2709,
      "step": 5700
    },
    {
      "epoch": 1.12477060599483,
      "eval_loss": 0.22571022808551788,
      "eval_runtime": 0.1849,
      "eval_samples_per_second": 27.039,
      "eval_steps_per_second": 5.408,
      "step": 5700
    },
    {
      "epoch": 1.134637014819346,
      "grad_norm": 0.20602265000343323,
      "learning_rate": 9.708176436696595e-05,
      "loss": 0.2721,
      "step": 5750
    },
    {
      "epoch": 1.134637014819346,
      "eval_loss": 0.2223341017961502,
      "eval_runtime": 0.1843,
      "eval_samples_per_second": 27.136,
      "eval_steps_per_second": 5.427,
      "step": 5750
    },
    {
      "epoch": 1.144503423643862,
      "grad_norm": 0.210631862282753,
      "learning_rate": 9.532483190038153e-05,
      "loss": 0.2679,
      "step": 5800
    },
    {
      "epoch": 1.144503423643862,
      "eval_loss": 0.2227615863084793,
      "eval_runtime": 0.185,
      "eval_samples_per_second": 27.02,
      "eval_steps_per_second": 5.404,
      "step": 5800
    },
    {
      "epoch": 1.1543698324683782,
      "grad_norm": 0.2317020148038864,
      "learning_rate": 9.356934465358979e-05,
      "loss": 0.272,
      "step": 5850
    },
    {
      "epoch": 1.1543698324683782,
      "eval_loss": 0.2202722728252411,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.087,
      "eval_steps_per_second": 5.417,
      "step": 5850
    },
    {
      "epoch": 1.1642362412928942,
      "grad_norm": 0.2153208702802658,
      "learning_rate": 9.181584529476025e-05,
      "loss": 0.2705,
      "step": 5900
    },
    {
      "epoch": 1.1642362412928942,
      "eval_loss": 0.2222643345594406,
      "eval_runtime": 0.1845,
      "eval_samples_per_second": 27.105,
      "eval_steps_per_second": 5.421,
      "step": 5900
    },
    {
      "epoch": 1.1741026501174103,
      "grad_norm": 0.26965272426605225,
      "learning_rate": 9.006487587755295e-05,
      "loss": 0.2699,
      "step": 5950
    },
    {
      "epoch": 1.1741026501174103,
      "eval_loss": 0.218417689204216,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.091,
      "eval_steps_per_second": 5.418,
      "step": 5950
    },
    {
      "epoch": 1.1839690589419263,
      "grad_norm": 0.21898072957992554,
      "learning_rate": 8.831697767355519e-05,
      "loss": 0.2698,
      "step": 6000
    },
    {
      "epoch": 1.1839690589419263,
      "eval_loss": 0.22147616744041443,
      "eval_runtime": 0.1855,
      "eval_samples_per_second": 26.955,
      "eval_steps_per_second": 5.391,
      "step": 6000
    },
    {
      "epoch": 1.1938354677664424,
      "grad_norm": 0.19538186490535736,
      "learning_rate": 8.65726910049599e-05,
      "loss": 0.2683,
      "step": 6050
    },
    {
      "epoch": 1.1938354677664424,
      "eval_loss": 0.21991392970085144,
      "eval_runtime": 0.1851,
      "eval_samples_per_second": 27.012,
      "eval_steps_per_second": 5.402,
      "step": 6050
    },
    {
      "epoch": 1.2037018765909584,
      "grad_norm": 0.21497651934623718,
      "learning_rate": 8.483255507753762e-05,
      "loss": 0.2696,
      "step": 6100
    },
    {
      "epoch": 1.2037018765909584,
      "eval_loss": 0.22209680080413818,
      "eval_runtime": 0.1849,
      "eval_samples_per_second": 27.037,
      "eval_steps_per_second": 5.407,
      "step": 6100
    },
    {
      "epoch": 1.2135682854154746,
      "grad_norm": 0.2190115749835968,
      "learning_rate": 8.309710781395356e-05,
      "loss": 0.2676,
      "step": 6150
    },
    {
      "epoch": 1.2135682854154746,
      "eval_loss": 0.22349147498607635,
      "eval_runtime": 0.1845,
      "eval_samples_per_second": 27.1,
      "eval_steps_per_second": 5.42,
      "step": 6150
    },
    {
      "epoch": 1.2234346942399905,
      "grad_norm": 0.18856999278068542,
      "learning_rate": 8.136688568748113e-05,
      "loss": 0.2642,
      "step": 6200
    },
    {
      "epoch": 1.2234346942399905,
      "eval_loss": 0.2215404212474823,
      "eval_runtime": 0.1844,
      "eval_samples_per_second": 27.115,
      "eval_steps_per_second": 5.423,
      "step": 6200
    },
    {
      "epoch": 1.2333011030645067,
      "grad_norm": 0.2113649994134903,
      "learning_rate": 7.964242355616385e-05,
      "loss": 0.2702,
      "step": 6250
    },
    {
      "epoch": 1.2333011030645067,
      "eval_loss": 0.22224430739879608,
      "eval_runtime": 0.1851,
      "eval_samples_per_second": 27.018,
      "eval_steps_per_second": 5.404,
      "step": 6250
    },
    {
      "epoch": 1.2433056416125658,
      "grad_norm": 0.23301884531974792,
      "learning_rate": 7.792425449747635e-05,
      "loss": 0.2701,
      "step": 6300
    },
    {
      "epoch": 1.2433056416125658,
      "eval_loss": 0.22047898173332214,
      "eval_runtime": 0.1979,
      "eval_samples_per_second": 25.262,
      "eval_steps_per_second": 5.052,
      "step": 6300
    },
    {
      "epoch": 1.2531720504370818,
      "grad_norm": 0.211432546377182,
      "learning_rate": 7.621290964353581e-05,
      "loss": 0.2696,
      "step": 6350
    },
    {
      "epoch": 1.2531720504370818,
      "eval_loss": 0.22401615977287292,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.079,
      "eval_steps_per_second": 5.416,
      "step": 6350
    },
    {
      "epoch": 1.263038459261598,
      "grad_norm": 0.2142704576253891,
      "learning_rate": 7.450891801691468e-05,
      "loss": 0.2671,
      "step": 6400
    },
    {
      "epoch": 1.263038459261598,
      "eval_loss": 0.2236458957195282,
      "eval_runtime": 0.1853,
      "eval_samples_per_second": 26.977,
      "eval_steps_per_second": 5.395,
      "step": 6400
    },
    {
      "epoch": 1.2729048680861141,
      "grad_norm": 0.1931128203868866,
      "learning_rate": 7.281280636710576e-05,
      "loss": 0.269,
      "step": 6450
    },
    {
      "epoch": 1.2729048680861141,
      "eval_loss": 0.22148045897483826,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.085,
      "eval_steps_per_second": 5.417,
      "step": 6450
    },
    {
      "epoch": 1.28277127691063,
      "grad_norm": 0.2120596170425415,
      "learning_rate": 7.112509900768989e-05,
      "loss": 0.2692,
      "step": 6500
    },
    {
      "epoch": 1.28277127691063,
      "eval_loss": 0.2202589511871338,
      "eval_runtime": 0.1852,
      "eval_samples_per_second": 27.003,
      "eval_steps_per_second": 5.401,
      "step": 6500
    },
    {
      "epoch": 1.292637685735146,
      "grad_norm": 0.20692607760429382,
      "learning_rate": 6.944631765425629e-05,
      "loss": 0.2647,
      "step": 6550
    },
    {
      "epoch": 1.292637685735146,
      "eval_loss": 0.2197343409061432,
      "eval_runtime": 0.185,
      "eval_samples_per_second": 27.031,
      "eval_steps_per_second": 5.406,
      "step": 6550
    },
    {
      "epoch": 1.3025040945596622,
      "grad_norm": 0.2301376610994339,
      "learning_rate": 6.777698126312647e-05,
      "loss": 0.2647,
      "step": 6600
    },
    {
      "epoch": 1.3025040945596622,
      "eval_loss": 0.2176646739244461,
      "eval_runtime": 0.1857,
      "eval_samples_per_second": 26.923,
      "eval_steps_per_second": 5.385,
      "step": 6600
    },
    {
      "epoch": 1.3123705033841782,
      "grad_norm": 0.2805488705635071,
      "learning_rate": 6.611760587093094e-05,
      "loss": 0.2635,
      "step": 6650
    },
    {
      "epoch": 1.3123705033841782,
      "eval_loss": 0.21856221556663513,
      "eval_runtime": 0.185,
      "eval_samples_per_second": 27.033,
      "eval_steps_per_second": 5.407,
      "step": 6650
    },
    {
      "epoch": 1.3222369122086943,
      "grad_norm": 0.2176360934972763,
      "learning_rate": 6.446870443508839e-05,
      "loss": 0.2628,
      "step": 6700
    },
    {
      "epoch": 1.3222369122086943,
      "eval_loss": 0.2163912057876587,
      "eval_runtime": 0.1848,
      "eval_samples_per_second": 27.054,
      "eval_steps_per_second": 5.411,
      "step": 6700
    },
    {
      "epoch": 1.3321033210332103,
      "grad_norm": 0.20765335857868195,
      "learning_rate": 6.283078667523681e-05,
      "loss": 0.2678,
      "step": 6750
    },
    {
      "epoch": 1.3321033210332103,
      "eval_loss": 0.2164166420698166,
      "eval_runtime": 0.1851,
      "eval_samples_per_second": 27.017,
      "eval_steps_per_second": 5.403,
      "step": 6750
    },
    {
      "epoch": 1.3419697298577264,
      "grad_norm": 0.2131153792142868,
      "learning_rate": 6.120435891566542e-05,
      "loss": 0.266,
      "step": 6800
    },
    {
      "epoch": 1.3419697298577264,
      "eval_loss": 0.21605464816093445,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.077,
      "eval_steps_per_second": 5.415,
      "step": 6800
    },
    {
      "epoch": 1.3518361386822424,
      "grad_norm": 0.20908519625663757,
      "learning_rate": 5.9589923928796434e-05,
      "loss": 0.2646,
      "step": 6850
    },
    {
      "epoch": 1.3518361386822424,
      "eval_loss": 0.21923890709877014,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.068,
      "eval_steps_per_second": 5.414,
      "step": 6850
    },
    {
      "epoch": 1.3617025475067586,
      "grad_norm": 0.20536689460277557,
      "learning_rate": 5.7987980779764463e-05,
      "loss": 0.2631,
      "step": 6900
    },
    {
      "epoch": 1.3617025475067586,
      "eval_loss": 0.2201545685529709,
      "eval_runtime": 0.1848,
      "eval_samples_per_second": 27.055,
      "eval_steps_per_second": 5.411,
      "step": 6900
    },
    {
      "epoch": 1.3715689563312745,
      "grad_norm": 0.21380028128623962,
      "learning_rate": 5.639902467214212e-05,
      "loss": 0.2651,
      "step": 6950
    },
    {
      "epoch": 1.3715689563312745,
      "eval_loss": 0.2177874594926834,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.069,
      "eval_steps_per_second": 5.414,
      "step": 6950
    },
    {
      "epoch": 1.3814353651557907,
      "grad_norm": 0.2062164545059204,
      "learning_rate": 5.482354679485948e-05,
      "loss": 0.2715,
      "step": 7000
    },
    {
      "epoch": 1.3814353651557907,
      "eval_loss": 0.21779461205005646,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.086,
      "eval_steps_per_second": 5.417,
      "step": 7000
    },
    {
      "epoch": 1.3913017739803066,
      "grad_norm": 0.22111141681671143,
      "learning_rate": 5.326203417036413e-05,
      "loss": 0.2601,
      "step": 7050
    },
    {
      "epoch": 1.3913017739803066,
      "eval_loss": 0.2183108627796173,
      "eval_runtime": 0.1845,
      "eval_samples_per_second": 27.097,
      "eval_steps_per_second": 5.419,
      "step": 7050
    },
    {
      "epoch": 1.4011681828048226,
      "grad_norm": 0.2018529623746872,
      "learning_rate": 5.17149695040698e-05,
      "loss": 0.2658,
      "step": 7100
    },
    {
      "epoch": 1.4011681828048226,
      "eval_loss": 0.2196209877729416,
      "eval_runtime": 0.1848,
      "eval_samples_per_second": 27.049,
      "eval_steps_per_second": 5.41,
      "step": 7100
    },
    {
      "epoch": 1.4110345916293388,
      "grad_norm": 0.2138175666332245,
      "learning_rate": 5.01828310351389e-05,
      "loss": 0.2647,
      "step": 7150
    },
    {
      "epoch": 1.4110345916293388,
      "eval_loss": 0.21882668137550354,
      "eval_runtime": 0.1844,
      "eval_samples_per_second": 27.118,
      "eval_steps_per_second": 5.424,
      "step": 7150
    },
    {
      "epoch": 1.420901000453855,
      "grad_norm": 0.21395103633403778,
      "learning_rate": 4.866609238864609e-05,
      "loss": 0.2614,
      "step": 7200
    },
    {
      "epoch": 1.420901000453855,
      "eval_loss": 0.21599645912647247,
      "eval_runtime": 0.1865,
      "eval_samples_per_second": 26.812,
      "eval_steps_per_second": 5.362,
      "step": 7200
    },
    {
      "epoch": 1.4307674092783709,
      "grad_norm": 0.19431644678115845,
      "learning_rate": 4.7165222429168156e-05,
      "loss": 0.2631,
      "step": 7250
    },
    {
      "epoch": 1.4307674092783709,
      "eval_loss": 0.21445360779762268,
      "eval_runtime": 0.1853,
      "eval_samples_per_second": 26.984,
      "eval_steps_per_second": 5.397,
      "step": 7250
    },
    {
      "epoch": 1.4406338181028868,
      "grad_norm": 0.2003539353609085,
      "learning_rate": 4.568068511584529e-05,
      "loss": 0.2619,
      "step": 7300
    },
    {
      "epoch": 1.4406338181028868,
      "eval_loss": 0.21582575142383575,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.092,
      "eval_steps_per_second": 5.418,
      "step": 7300
    },
    {
      "epoch": 1.450500226927403,
      "grad_norm": 0.2108386605978012,
      "learning_rate": 4.421293935895882e-05,
      "loss": 0.2618,
      "step": 7350
    },
    {
      "epoch": 1.450500226927403,
      "eval_loss": 0.21553893387317657,
      "eval_runtime": 0.1856,
      "eval_samples_per_second": 26.942,
      "eval_steps_per_second": 5.388,
      "step": 7350
    },
    {
      "epoch": 1.460366635751919,
      "grad_norm": 0.2119232714176178,
      "learning_rate": 4.2762438878069955e-05,
      "loss": 0.2667,
      "step": 7400
    },
    {
      "epoch": 1.460366635751919,
      "eval_loss": 0.21502795815467834,
      "eval_runtime": 0.1851,
      "eval_samples_per_second": 27.012,
      "eval_steps_per_second": 5.402,
      "step": 7400
    },
    {
      "epoch": 1.470233044576435,
      "grad_norm": 0.2125096172094345,
      "learning_rate": 4.13296320617627e-05,
      "loss": 0.2625,
      "step": 7450
    },
    {
      "epoch": 1.470233044576435,
      "eval_loss": 0.2130458652973175,
      "eval_runtime": 0.1857,
      "eval_samples_per_second": 26.926,
      "eval_steps_per_second": 5.385,
      "step": 7450
    },
    {
      "epoch": 1.480099453400951,
      "grad_norm": 0.20715288817882538,
      "learning_rate": 3.991496182903498e-05,
      "loss": 0.2607,
      "step": 7500
    },
    {
      "epoch": 1.480099453400951,
      "eval_loss": 0.2125515192747116,
      "eval_runtime": 0.1855,
      "eval_samples_per_second": 26.953,
      "eval_steps_per_second": 5.391,
      "step": 7500
    },
    {
      "epoch": 1.4899658622254672,
      "grad_norm": 0.19759340584278107,
      "learning_rate": 3.851886549238062e-05,
      "loss": 0.2596,
      "step": 7550
    },
    {
      "epoch": 1.4899658622254672,
      "eval_loss": 0.2126692235469818,
      "eval_runtime": 0.1848,
      "eval_samples_per_second": 27.05,
      "eval_steps_per_second": 5.41,
      "step": 7550
    },
    {
      "epoch": 1.4998322710499832,
      "grad_norm": 0.1995980143547058,
      "learning_rate": 3.714177462260412e-05,
      "loss": 0.263,
      "step": 7600
    },
    {
      "epoch": 1.4998322710499832,
      "eval_loss": 0.2116161286830902,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.091,
      "eval_steps_per_second": 5.418,
      "step": 7600
    },
    {
      "epoch": 1.5096986798744991,
      "grad_norm": 0.23178283870220184,
      "learning_rate": 3.578411491541079e-05,
      "loss": 0.2623,
      "step": 7650
    },
    {
      "epoch": 1.5096986798744991,
      "eval_loss": 0.21194568276405334,
      "eval_runtime": 0.1847,
      "eval_samples_per_second": 27.075,
      "eval_steps_per_second": 5.415,
      "step": 7650
    },
    {
      "epoch": 1.5195650886990153,
      "grad_norm": 0.2182031124830246,
      "learning_rate": 3.444630605981256e-05,
      "loss": 0.2558,
      "step": 7700
    },
    {
      "epoch": 1.5195650886990153,
      "eval_loss": 0.21379950642585754,
      "eval_runtime": 0.1846,
      "eval_samples_per_second": 27.079,
      "eval_steps_per_second": 5.416,
      "step": 7700
    },
    {
      "epoch": 1.5294314975235315,
      "grad_norm": 0.2095683068037033,
      "learning_rate": 3.312876160839099e-05,
      "loss": 0.261,
      "step": 7750
    },
    {
      "epoch": 1.5294314975235315,
      "eval_loss": 0.21340413391590118,
      "eval_runtime": 0.1852,
      "eval_samples_per_second": 26.992,
      "eval_steps_per_second": 5.398,
      "step": 7750
    },
    {
      "epoch": 1.5392979063480474,
      "grad_norm": 0.2217652052640915,
      "learning_rate": 3.183188884945714e-05,
      "loss": 0.2615,
      "step": 7800
    },
    {
      "epoch": 1.5392979063480474,
      "eval_loss": 0.21244795620441437,
      "eval_runtime": 0.1841,
      "eval_samples_per_second": 27.163,
      "eval_steps_per_second": 5.433,
      "step": 7800
    },
    {
      "epoch": 1.5491643151725634,
      "grad_norm": 0.21375681459903717,
      "learning_rate": 3.055608868114761e-05,
      "loss": 0.2571,
      "step": 7850
    },
    {
      "epoch": 1.5491643151725634,
      "eval_loss": 0.21376018226146698,
      "eval_runtime": 0.1845,
      "eval_samples_per_second": 27.101,
      "eval_steps_per_second": 5.42,
      "step": 7850
    },
    {
      "epoch": 1.5590307239970795,
      "grad_norm": 0.22278694808483124,
      "learning_rate": 2.930175548749645e-05,
      "loss": 0.2614,
      "step": 7900
    },
    {
      "epoch": 1.5590307239970795,
      "eval_loss": 0.21383312344551086,
      "eval_runtime": 0.1851,
      "eval_samples_per_second": 27.013,
      "eval_steps_per_second": 5.403,
      "step": 7900
    },
    {
      "epoch": 1.5688971328215957,
      "grad_norm": 0.2054503709077835,
      "learning_rate": 2.806927701652029e-05,
      "loss": 0.2628,
      "step": 7950
    },
    {
      "epoch": 1.5688971328215957,
      "eval_loss": 0.21245412528514862,
      "eval_runtime": 0.1849,
      "eval_samples_per_second": 27.044,
      "eval_steps_per_second": 5.409,
      "step": 7950
    },
    {
      "epoch": 1.5787635416461117,
      "grad_norm": 0.20330142974853516,
      "learning_rate": 2.6859034260355042e-05,
      "loss": 0.2587,
      "step": 8000
    },
    {
      "epoch": 1.5787635416461117,
      "eval_loss": 0.21231678128242493,
      "eval_runtime": 0.1841,
      "eval_samples_per_second": 27.158,
      "eval_steps_per_second": 5.432,
      "step": 8000
    },
    {
      "epoch": 1.5886299504706276,
      "grad_norm": 0.20486824214458466,
      "learning_rate": 2.567140133748118e-05,
      "loss": 0.2584,
      "step": 8050
    },
    {
      "epoch": 1.5886299504706276,
      "eval_loss": 0.2128223478794098,
      "eval_runtime": 0.2002,
      "eval_samples_per_second": 24.971,
      "eval_steps_per_second": 4.994,
      "step": 8050
    },
    {
      "epoch": 1.5984963592951438,
      "grad_norm": 0.21578329801559448,
      "learning_rate": 2.4506745377073535e-05,
      "loss": 0.2581,
      "step": 8100
    },
    {
      "epoch": 1.5984963592951438,
      "eval_loss": 0.2118312120437622,
      "eval_runtime": 0.1883,
      "eval_samples_per_second": 26.558,
      "eval_steps_per_second": 5.312,
      "step": 8100
    },
    {
      "epoch": 1.60836276811966,
      "grad_norm": 0.21013592183589935,
      "learning_rate": 2.3365426405511802e-05,
      "loss": 0.2613,
      "step": 8150
    },
    {
      "epoch": 1.60836276811966,
      "eval_loss": 0.21296799182891846,
      "eval_runtime": 0.1872,
      "eval_samples_per_second": 26.709,
      "eval_steps_per_second": 5.342,
      "step": 8150
    },
    {
      "epoch": 1.618229176944176,
      "grad_norm": 0.22792735695838928,
      "learning_rate": 2.224779723508692e-05,
      "loss": 0.2592,
      "step": 8200
    },
    {
      "epoch": 1.618229176944176,
      "eval_loss": 0.2124718874692917,
      "eval_runtime": 0.1884,
      "eval_samples_per_second": 26.537,
      "eval_steps_per_second": 5.307,
      "step": 8200
    },
    {
      "epoch": 1.6280955857686918,
      "grad_norm": 0.2071530818939209,
      "learning_rate": 2.1154203354937074e-05,
      "loss": 0.2589,
      "step": 8250
    },
    {
      "epoch": 1.6280955857686918,
      "eval_loss": 0.21143141388893127,
      "eval_runtime": 0.1867,
      "eval_samples_per_second": 26.788,
      "eval_steps_per_second": 5.358,
      "step": 8250
    },
    {
      "epoch": 1.637961994593208,
      "grad_norm": 0.2163563370704651,
      "learning_rate": 2.0084982824248034e-05,
      "loss": 0.2591,
      "step": 8300
    },
    {
      "epoch": 1.637961994593208,
      "eval_loss": 0.2110225409269333,
      "eval_runtime": 0.188,
      "eval_samples_per_second": 26.59,
      "eval_steps_per_second": 5.318,
      "step": 8300
    },
    {
      "epoch": 1.647828403417724,
      "grad_norm": 0.1936330944299698,
      "learning_rate": 1.9040466167749727e-05,
      "loss": 0.2574,
      "step": 8350
    },
    {
      "epoch": 1.647828403417724,
      "eval_loss": 0.21078595519065857,
      "eval_runtime": 0.1871,
      "eval_samples_per_second": 26.723,
      "eval_steps_per_second": 5.345,
      "step": 8350
    },
    {
      "epoch": 1.65769481224224,
      "grad_norm": 0.20487622916698456,
      "learning_rate": 1.802097627354231e-05,
      "loss": 0.2602,
      "step": 8400
    },
    {
      "epoch": 1.65769481224224,
      "eval_loss": 0.21095280349254608,
      "eval_runtime": 0.1863,
      "eval_samples_per_second": 26.838,
      "eval_steps_per_second": 5.368,
      "step": 8400
    },
    {
      "epoch": 1.667561221066756,
      "grad_norm": 0.20645365118980408,
      "learning_rate": 1.7026828293282892e-05,
      "loss": 0.2607,
      "step": 8450
    },
    {
      "epoch": 1.667561221066756,
      "eval_loss": 0.2102348506450653,
      "eval_runtime": 0.187,
      "eval_samples_per_second": 26.732,
      "eval_steps_per_second": 5.346,
      "step": 8450
    },
    {
      "epoch": 1.6774276298912723,
      "grad_norm": 0.21055789291858673,
      "learning_rate": 1.605832954476346e-05,
      "loss": 0.2567,
      "step": 8500
    },
    {
      "epoch": 1.6774276298912723,
      "eval_loss": 0.2107050120830536,
      "eval_runtime": 0.1878,
      "eval_samples_per_second": 26.618,
      "eval_steps_per_second": 5.324,
      "step": 8500
    },
    {
      "epoch": 1.6872940387157882,
      "grad_norm": 0.1864519864320755,
      "learning_rate": 1.5115779416911014e-05,
      "loss": 0.2568,
      "step": 8550
    },
    {
      "epoch": 1.6872940387157882,
      "eval_loss": 0.21016505360603333,
      "eval_runtime": 0.1867,
      "eval_samples_per_second": 26.788,
      "eval_steps_per_second": 5.358,
      "step": 8550
    },
    {
      "epoch": 1.6971604475403042,
      "grad_norm": 0.2051512748003006,
      "learning_rate": 1.4199469277238143e-05,
      "loss": 0.2614,
      "step": 8600
    },
    {
      "epoch": 1.6971604475403042,
      "eval_loss": 0.21043357253074646,
      "eval_runtime": 0.1874,
      "eval_samples_per_second": 26.678,
      "eval_steps_per_second": 5.336,
      "step": 8600
    },
    {
      "epoch": 1.7070268563648203,
      "grad_norm": 0.21571016311645508,
      "learning_rate": 1.330968238177368e-05,
      "loss": 0.2563,
      "step": 8650
    },
    {
      "epoch": 1.7070268563648203,
      "eval_loss": 0.21016483008861542,
      "eval_runtime": 0.187,
      "eval_samples_per_second": 26.739,
      "eval_steps_per_second": 5.348,
      "step": 8650
    },
    {
      "epoch": 1.7168932651893365,
      "grad_norm": 0.2853296399116516,
      "learning_rate": 1.2446693787500697e-05,
      "loss": 0.2589,
      "step": 8700
    },
    {
      "epoch": 1.7168932651893365,
      "eval_loss": 0.21085257828235626,
      "eval_runtime": 0.1856,
      "eval_samples_per_second": 26.945,
      "eval_steps_per_second": 5.389,
      "step": 8700
    },
    {
      "epoch": 1.7267596740138524,
      "grad_norm": 0.19415363669395447,
      "learning_rate": 1.1610770267328852e-05,
      "loss": 0.2613,
      "step": 8750
    },
    {
      "epoch": 1.7267596740138524,
      "eval_loss": 0.21094462275505066,
      "eval_runtime": 0.1866,
      "eval_samples_per_second": 26.798,
      "eval_steps_per_second": 5.36,
      "step": 8750
    },
    {
      "epoch": 1.7366260828383684,
      "grad_norm": 0.20800048112869263,
      "learning_rate": 1.0802170227627873e-05,
      "loss": 0.2583,
      "step": 8800
    },
    {
      "epoch": 1.7366260828383684,
      "eval_loss": 0.21065060794353485,
      "eval_runtime": 0.187,
      "eval_samples_per_second": 26.739,
      "eval_steps_per_second": 5.348,
      "step": 8800
    },
    {
      "epoch": 1.7464924916628846,
      "grad_norm": 0.21770550310611725,
      "learning_rate": 1.0021143628347196e-05,
      "loss": 0.2634,
      "step": 8850
    },
    {
      "epoch": 1.7464924916628846,
      "eval_loss": 0.2107730358839035,
      "eval_runtime": 0.1872,
      "eval_samples_per_second": 26.713,
      "eval_steps_per_second": 5.343,
      "step": 8850
    },
    {
      "epoch": 1.7563589004874007,
      "grad_norm": 0.20711322128772736,
      "learning_rate": 9.26793190574664e-06,
      "loss": 0.2586,
      "step": 8900
    },
    {
      "epoch": 1.7563589004874007,
      "eval_loss": 0.21060581505298615,
      "eval_runtime": 0.1873,
      "eval_samples_per_second": 26.696,
      "eval_steps_per_second": 5.339,
      "step": 8900
    },
    {
      "epoch": 1.7662253093119167,
      "grad_norm": 0.1906907856464386,
      "learning_rate": 8.542767897761917e-06,
      "loss": 0.2579,
      "step": 8950
    },
    {
      "epoch": 1.7662253093119167,
      "eval_loss": 0.21017535030841827,
      "eval_runtime": 0.1865,
      "eval_samples_per_second": 26.808,
      "eval_steps_per_second": 5.362,
      "step": 8950
    },
    {
      "epoch": 1.7760917181364326,
      "grad_norm": 0.23783724009990692,
      "learning_rate": 7.845875772028289e-06,
      "loss": 0.257,
      "step": 9000
    },
    {
      "epoch": 1.7760917181364326,
      "eval_loss": 0.21030497550964355,
      "eval_runtime": 0.1866,
      "eval_samples_per_second": 26.796,
      "eval_steps_per_second": 5.359,
      "step": 9000
    },
    {
      "epoch": 1.7859581269609488,
      "grad_norm": 0.20180200040340424,
      "learning_rate": 7.177470956584253e-06,
      "loss": 0.2577,
      "step": 9050
    },
    {
      "epoch": 1.7859581269609488,
      "eval_loss": 0.21059055626392365,
      "eval_runtime": 3427.3667,
      "eval_samples_per_second": 0.001,
      "eval_steps_per_second": 0.0,
      "step": 9050
    },
    {
      "epoch": 1.795824535785465,
      "grad_norm": 0.19345524907112122,
      "learning_rate": 6.537760073277066e-06,
      "loss": 0.2569,
      "step": 9100
    },
    {
      "epoch": 1.795824535785465,
      "eval_loss": 0.21125975251197815,
      "eval_runtime": 0.187,
      "eval_samples_per_second": 26.737,
      "eval_steps_per_second": 5.347,
      "step": 9100
    },
    {
      "epoch": 1.8056909446099807,
      "grad_norm": 0.20590749382972717,
      "learning_rate": 5.926940873890263e-06,
      "loss": 0.2575,
      "step": 9150
    },
    {
      "epoch": 1.8056909446099807,
      "eval_loss": 0.21117892861366272,
      "eval_runtime": 0.1868,
      "eval_samples_per_second": 26.762,
      "eval_steps_per_second": 5.352,
      "step": 9150
    },
    {
      "epoch": 1.8155573534344969,
      "grad_norm": 0.22785334289073944,
      "learning_rate": 5.345202179013353e-06,
      "loss": 0.2593,
      "step": 9200
    },
    {
      "epoch": 1.8155573534344969,
      "eval_loss": 0.2109127938747406,
      "eval_runtime": 0.1867,
      "eval_samples_per_second": 26.785,
      "eval_steps_per_second": 5.357,
      "step": 9200
    },
    {
      "epoch": 1.825423762259013,
      "grad_norm": 0.20013581216335297,
      "learning_rate": 4.792723819672351e-06,
      "loss": 0.2564,
      "step": 9250
    },
    {
      "epoch": 1.825423762259013,
      "eval_loss": 0.21093535423278809,
      "eval_runtime": 0.1869,
      "eval_samples_per_second": 26.753,
      "eval_steps_per_second": 5.351,
      "step": 9250
    },
    {
      "epoch": 1.835290171083529,
      "grad_norm": 0.18782545626163483,
      "learning_rate": 4.269676581739079e-06,
      "loss": 0.2592,
      "step": 9300
    },
    {
      "epoch": 1.835290171083529,
      "eval_loss": 0.21116788685321808,
      "eval_runtime": 0.187,
      "eval_samples_per_second": 26.734,
      "eval_steps_per_second": 5.347,
      "step": 9300
    },
    {
      "epoch": 1.845156579908045,
      "grad_norm": 0.284829318523407,
      "learning_rate": 3.776222153136788e-06,
      "loss": 0.2616,
      "step": 9350
    },
    {
      "epoch": 1.845156579908045,
      "eval_loss": 0.2108537256717682,
      "eval_runtime": 0.187,
      "eval_samples_per_second": 26.739,
      "eval_steps_per_second": 5.348,
      "step": 9350
    },
    {
      "epoch": 1.855022988732561,
      "grad_norm": 0.21778038144111633,
      "learning_rate": 3.3125130738579922e-06,
      "loss": 0.2585,
      "step": 9400
    },
    {
      "epoch": 1.855022988732561,
      "eval_loss": 0.21062783896923065,
      "eval_runtime": 0.1869,
      "eval_samples_per_second": 26.748,
      "eval_steps_per_second": 5.35,
      "step": 9400
    },
    {
      "epoch": 1.8648893975570773,
      "grad_norm": 0.22620098292827606,
      "learning_rate": 2.878692688810314e-06,
      "loss": 0.2578,
      "step": 9450
    },
    {
      "epoch": 1.8648893975570773,
      "eval_loss": 0.21043607592582703,
      "eval_runtime": 0.187,
      "eval_samples_per_second": 26.734,
      "eval_steps_per_second": 5.347,
      "step": 9450
    },
    {
      "epoch": 1.8747558063815932,
      "grad_norm": 0.2136729210615158,
      "learning_rate": 2.4748951035047596e-06,
      "loss": 0.2605,
      "step": 9500
    },
    {
      "epoch": 1.8747558063815932,
      "eval_loss": 0.21055133640766144,
      "eval_runtime": 0.1877,
      "eval_samples_per_second": 26.643,
      "eval_steps_per_second": 5.329,
      "step": 9500
    },
    {
      "epoch": 1.8846222152061092,
      "grad_norm": 0.19579312205314636,
      "learning_rate": 2.101245142600039e-06,
      "loss": 0.2592,
      "step": 9550
    },
    {
      "epoch": 1.8846222152061092,
      "eval_loss": 0.21009139716625214,
      "eval_runtime": 0.1864,
      "eval_samples_per_second": 26.826,
      "eval_steps_per_second": 5.365,
      "step": 9550
    },
    {
      "epoch": 1.8944886240306253,
      "grad_norm": 0.2088346630334854,
      "learning_rate": 1.7578583113159962e-06,
      "loss": 0.2575,
      "step": 9600
    },
    {
      "epoch": 1.8944886240306253,
      "eval_loss": 0.21053215861320496,
      "eval_runtime": 0.1864,
      "eval_samples_per_second": 26.826,
      "eval_steps_per_second": 5.365,
      "step": 9600
    },
    {
      "epoch": 1.9043550328551415,
      "grad_norm": 0.23137839138507843,
      "learning_rate": 1.4448407597277392e-06,
      "loss": 0.2606,
      "step": 9650
    },
    {
      "epoch": 1.9043550328551415,
      "eval_loss": 0.21052584052085876,
      "eval_runtime": 0.1872,
      "eval_samples_per_second": 26.706,
      "eval_steps_per_second": 5.341,
      "step": 9650
    },
    {
      "epoch": 1.9142214416796575,
      "grad_norm": 0.21073339879512787,
      "learning_rate": 1.1622892499519421e-06,
      "loss": 0.2558,
      "step": 9700
    },
    {
      "epoch": 1.9142214416796575,
      "eval_loss": 0.21001854538917542,
      "eval_runtime": 0.1866,
      "eval_samples_per_second": 26.791,
      "eval_steps_per_second": 5.358,
      "step": 9700
    },
    {
      "epoch": 1.9240878505041734,
      "grad_norm": 0.20351086556911469,
      "learning_rate": 9.102911262349856e-07,
      "loss": 0.2534,
      "step": 9750
    },
    {
      "epoch": 1.9240878505041734,
      "eval_loss": 0.20996317267417908,
      "eval_runtime": 0.182,
      "eval_samples_per_second": 27.467,
      "eval_steps_per_second": 5.493,
      "step": 9750
    }
  ],
  "logging_steps": 50,
  "max_steps": 10134,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 250,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.0792506223580856e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}