{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.225,
  "eval_steps": 500,
  "global_step": 1800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00125,
      "grad_norm": 0.07407524436712265,
      "learning_rate": 2.25e-06,
      "loss": 2.2311376571655273,
      "step": 10
    },
    {
      "epoch": 0.0025,
      "grad_norm": 0.07508661597967148,
      "learning_rate": 4.75e-06,
      "loss": 2.254044532775879,
      "step": 20
    },
    {
      "epoch": 0.00375,
      "grad_norm": 0.08043646067380905,
      "learning_rate": 7.25e-06,
      "loss": 2.2791187286376955,
      "step": 30
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.07152171432971954,
      "learning_rate": 9.750000000000002e-06,
      "loss": 2.29660701751709,
      "step": 40
    },
    {
      "epoch": 0.00625,
      "grad_norm": 0.07239466905593872,
      "learning_rate": 1.225e-05,
      "loss": 2.2157821655273438,
      "step": 50
    },
    {
      "epoch": 0.0075,
      "grad_norm": 0.07063177227973938,
      "learning_rate": 1.475e-05,
      "loss": 2.2415660858154296,
      "step": 60
    },
    {
      "epoch": 0.00875,
      "grad_norm": 0.07137205451726913,
      "learning_rate": 1.725e-05,
      "loss": 2.188554573059082,
      "step": 70
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.07028814405202866,
      "learning_rate": 1.9750000000000002e-05,
      "loss": 2.177708053588867,
      "step": 80
    },
    {
      "epoch": 0.01125,
      "grad_norm": 0.07327590137720108,
      "learning_rate": 2.2250000000000002e-05,
      "loss": 2.22082405090332,
      "step": 90
    },
    {
      "epoch": 0.0125,
      "grad_norm": 0.06442452222108841,
      "learning_rate": 2.4750000000000002e-05,
      "loss": 2.2364925384521483,
      "step": 100
    },
    {
      "epoch": 0.01375,
      "grad_norm": 0.06526491045951843,
      "learning_rate": 2.725e-05,
      "loss": 2.2664451599121094,
      "step": 110
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.06679920852184296,
      "learning_rate": 2.975e-05,
      "loss": 2.188519287109375,
      "step": 120
    },
    {
      "epoch": 0.01625,
      "grad_norm": 0.07065364718437195,
      "learning_rate": 3.2250000000000005e-05,
      "loss": 2.270803451538086,
      "step": 130
    },
    {
      "epoch": 0.0175,
      "grad_norm": 0.07348085194826126,
      "learning_rate": 3.475e-05,
      "loss": 2.213222122192383,
      "step": 140
    },
    {
      "epoch": 0.01875,
      "grad_norm": 0.0695674866437912,
      "learning_rate": 3.7250000000000004e-05,
      "loss": 2.208852767944336,
      "step": 150
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.07005158066749573,
      "learning_rate": 3.9750000000000004e-05,
      "loss": 2.161180305480957,
      "step": 160
    },
    {
      "epoch": 0.02125,
      "grad_norm": 0.06879838556051254,
      "learning_rate": 4.2250000000000004e-05,
      "loss": 2.1557788848876953,
      "step": 170
    },
    {
      "epoch": 0.0225,
      "grad_norm": 0.07037137448787689,
      "learning_rate": 4.4750000000000004e-05,
      "loss": 2.2376060485839844,
      "step": 180
    },
    {
      "epoch": 0.02375,
      "grad_norm": 0.06982705742120743,
      "learning_rate": 4.7249999999999997e-05,
      "loss": 2.217011260986328,
      "step": 190
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.06674907356500626,
      "learning_rate": 4.975e-05,
      "loss": 2.256929397583008,
      "step": 200
    },
    {
      "epoch": 0.02625,
      "grad_norm": 0.07002667337656021,
      "learning_rate": 4.999983575010662e-05,
      "loss": 2.179422378540039,
      "step": 210
    },
    {
      "epoch": 0.0275,
      "grad_norm": 0.0717761367559433,
      "learning_rate": 4.9999267975468225e-05,
      "loss": 2.2249755859375,
      "step": 220
    },
    {
      "epoch": 0.02875,
      "grad_norm": 0.07136455923318863,
      "learning_rate": 4.9998294657516734e-05,
      "loss": 2.222713088989258,
      "step": 230
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.07328379899263382,
      "learning_rate": 4.999691581204152e-05,
      "loss": 2.1884521484375,
      "step": 240
    },
    {
      "epoch": 0.03125,
      "grad_norm": 0.07197773456573486,
      "learning_rate": 4.999513146141048e-05,
      "loss": 2.2087806701660155,
      "step": 250
    },
    {
      "epoch": 0.0325,
      "grad_norm": 0.07391443103551865,
      "learning_rate": 4.999294163456975e-05,
      "loss": 2.2307754516601563,
      "step": 260
    },
    {
      "epoch": 0.03375,
      "grad_norm": 0.07325068861246109,
      "learning_rate": 4.9990346367043114e-05,
      "loss": 2.198223114013672,
      "step": 270
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.07103461772203445,
      "learning_rate": 4.9987345700931586e-05,
      "loss": 2.228559112548828,
      "step": 280
    },
    {
      "epoch": 0.03625,
      "grad_norm": 0.07458475232124329,
      "learning_rate": 4.998393968491256e-05,
      "loss": 2.1933670043945312,
      "step": 290
    },
    {
      "epoch": 0.0375,
      "grad_norm": 0.06882892549037933,
      "learning_rate": 4.9980128374239156e-05,
      "loss": 2.2323123931884767,
      "step": 300
    },
    {
      "epoch": 0.03875,
      "grad_norm": 0.07062669098377228,
      "learning_rate": 4.997591183073923e-05,
      "loss": 2.2588329315185547,
      "step": 310
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.07071756571531296,
      "learning_rate": 4.997129012281443e-05,
      "loss": 2.11534423828125,
      "step": 320
    },
    {
      "epoch": 0.04125,
      "grad_norm": 0.06557808071374893,
      "learning_rate": 4.996626332543907e-05,
      "loss": 2.1684688568115233,
      "step": 330
    },
    {
      "epoch": 0.0425,
      "grad_norm": 0.0680045560002327,
      "learning_rate": 4.9960831520158904e-05,
      "loss": 2.1954887390136717,
      "step": 340
    },
    {
      "epoch": 0.04375,
      "grad_norm": 0.07579149305820465,
      "learning_rate": 4.9954994795089786e-05,
      "loss": 2.120090103149414,
      "step": 350
    },
    {
      "epoch": 0.045,
      "grad_norm": 0.07108975201845169,
      "learning_rate": 4.994875324491629e-05,
      "loss": 2.1955142974853517,
      "step": 360
    },
    {
      "epoch": 0.04625,
      "grad_norm": 0.07410898804664612,
      "learning_rate": 4.994210697089014e-05,
      "loss": 2.2114063262939454,
      "step": 370
    },
    {
      "epoch": 0.0475,
      "grad_norm": 0.06744780391454697,
      "learning_rate": 4.993505608082857e-05,
      "loss": 2.1475910186767577,
      "step": 380
    },
    {
      "epoch": 0.04875,
      "grad_norm": 0.06901328265666962,
      "learning_rate": 4.992760068911258e-05,
      "loss": 2.2015354156494142,
      "step": 390
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.0716853216290474,
      "learning_rate": 4.991974091668509e-05,
      "loss": 2.2013004302978514,
      "step": 400
    },
    {
      "epoch": 0.05125,
      "grad_norm": 0.07426904886960983,
      "learning_rate": 4.991147689104895e-05,
      "loss": 2.170287322998047,
      "step": 410
    },
    {
      "epoch": 0.0525,
      "grad_norm": 0.06732422858476639,
      "learning_rate": 4.990280874626491e-05,
      "loss": 2.1896934509277344,
      "step": 420
    },
    {
      "epoch": 0.05375,
      "grad_norm": 0.06560606509447098,
      "learning_rate": 4.98937366229494e-05,
      "loss": 2.2104034423828125,
      "step": 430
    },
    {
      "epoch": 0.055,
      "grad_norm": 0.06944538652896881,
      "learning_rate": 4.98842606682723e-05,
      "loss": 2.126712989807129,
      "step": 440
    },
    {
      "epoch": 0.05625,
      "grad_norm": 0.06867734342813492,
      "learning_rate": 4.98743810359545e-05,
      "loss": 2.234315299987793,
      "step": 450
    },
    {
      "epoch": 0.0575,
      "grad_norm": 0.07116754353046417,
      "learning_rate": 4.986409788626546e-05,
      "loss": 2.184938430786133,
      "step": 460
    },
    {
      "epoch": 0.05875,
      "grad_norm": 0.06651467084884644,
      "learning_rate": 4.985341138602056e-05,
      "loss": 2.152230453491211,
      "step": 470
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.06930788606405258,
      "learning_rate": 4.984232170857842e-05,
      "loss": 2.181766891479492,
      "step": 480
    },
    {
      "epoch": 0.06125,
      "grad_norm": 0.07047919183969498,
      "learning_rate": 4.9830829033838096e-05,
      "loss": 2.187181091308594,
      "step": 490
    },
    {
      "epoch": 0.0625,
      "grad_norm": 0.07572918385267258,
      "learning_rate": 4.981893354823614e-05,
      "loss": 2.1816110610961914,
      "step": 500
    },
    {
      "epoch": 0.06375,
      "grad_norm": 0.07347270846366882,
      "learning_rate": 4.9806635444743595e-05,
      "loss": 2.2242454528808593,
      "step": 510
    },
    {
      "epoch": 0.065,
      "grad_norm": 0.06983183324337006,
      "learning_rate": 4.979393492286284e-05,
      "loss": 2.1648427963256838,
      "step": 520
    },
    {
      "epoch": 0.06625,
      "grad_norm": 0.06904461979866028,
      "learning_rate": 4.9780832188624375e-05,
      "loss": 2.1993919372558595,
      "step": 530
    },
    {
      "epoch": 0.0675,
      "grad_norm": 0.0692647248506546,
      "learning_rate": 4.976732745458348e-05,
      "loss": 2.1890403747558596,
      "step": 540
    },
    {
      "epoch": 0.06875,
      "grad_norm": 0.06891040503978729,
      "learning_rate": 4.975342093981675e-05,
      "loss": 2.1854305267333984,
      "step": 550
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.0714147537946701,
      "learning_rate": 4.973911286991856e-05,
      "loss": 2.219981002807617,
      "step": 560
    },
    {
      "epoch": 0.07125,
      "grad_norm": 0.07621099054813385,
      "learning_rate": 4.9724403476997384e-05,
      "loss": 2.182698440551758,
      "step": 570
    },
    {
      "epoch": 0.0725,
      "grad_norm": 0.07603556662797928,
      "learning_rate": 4.970929299967204e-05,
      "loss": 2.1696598052978517,
      "step": 580
    },
    {
      "epoch": 0.07375,
      "grad_norm": 0.07282973825931549,
      "learning_rate": 4.969378168306784e-05,
      "loss": 2.1716108322143555,
      "step": 590
    },
    {
      "epoch": 0.075,
      "grad_norm": 0.06459035724401474,
      "learning_rate": 4.967786977881257e-05,
      "loss": 2.1681615829467775,
      "step": 600
    },
    {
      "epoch": 0.07625,
      "grad_norm": 0.07177098840475082,
      "learning_rate": 4.966155754503245e-05,
      "loss": 2.175799560546875,
      "step": 610
    },
    {
      "epoch": 0.0775,
      "grad_norm": 0.06652142852544785,
      "learning_rate": 4.964484524634792e-05,
      "loss": 2.1907752990722655,
      "step": 620
    },
    {
      "epoch": 0.07875,
      "grad_norm": 0.0727977305650711,
      "learning_rate": 4.962773315386935e-05,
      "loss": 2.1232975006103514,
      "step": 630
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.07146522402763367,
      "learning_rate": 4.961022154519267e-05,
      "loss": 2.1747344970703124,
      "step": 640
    },
    {
      "epoch": 0.08125,
      "grad_norm": 0.06721191853284836,
      "learning_rate": 4.959231070439482e-05,
      "loss": 2.2015512466430662,
      "step": 650
    },
    {
      "epoch": 0.0825,
      "grad_norm": 0.06989260762929916,
      "learning_rate": 4.957400092202917e-05,
      "loss": 2.187305450439453,
      "step": 660
    },
    {
      "epoch": 0.08375,
      "grad_norm": 0.06762924790382385,
      "learning_rate": 4.955529249512082e-05,
      "loss": 2.1369449615478517,
      "step": 670
    },
    {
      "epoch": 0.085,
      "grad_norm": 0.06524205207824707,
      "learning_rate": 4.953618572716175e-05,
      "loss": 2.162923049926758,
      "step": 680
    },
    {
      "epoch": 0.08625,
      "grad_norm": 0.0713450163602829,
      "learning_rate": 4.951668092810593e-05,
      "loss": 2.2041025161743164,
      "step": 690
    },
    {
      "epoch": 0.0875,
      "grad_norm": 0.0673929825425148,
      "learning_rate": 4.949677841436423e-05,
      "loss": 2.124150848388672,
      "step": 700
    },
    {
      "epoch": 0.08875,
      "grad_norm": 0.06595680117607117,
      "learning_rate": 4.947647850879938e-05,
      "loss": 2.1866949081420897,
      "step": 710
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.06935486942529678,
      "learning_rate": 4.945578154072065e-05,
      "loss": 2.168051528930664,
      "step": 720
    },
    {
      "epoch": 0.09125,
      "grad_norm": 0.07811375707387924,
      "learning_rate": 4.943468784587856e-05,
      "loss": 2.2397434234619142,
      "step": 730
    },
    {
      "epoch": 0.0925,
      "grad_norm": 0.06744036078453064,
      "learning_rate": 4.9413197766459394e-05,
      "loss": 2.206024169921875,
      "step": 740
    },
    {
      "epoch": 0.09375,
      "grad_norm": 0.06797294318675995,
      "learning_rate": 4.939131165107969e-05,
      "loss": 2.16088981628418,
      "step": 750
    },
    {
      "epoch": 0.095,
      "grad_norm": 0.06975241005420685,
      "learning_rate": 4.936902985478055e-05,
      "loss": 2.2246139526367186,
      "step": 760
    },
    {
      "epoch": 0.09625,
      "grad_norm": 0.0649324357509613,
      "learning_rate": 4.9346352739021895e-05,
      "loss": 2.1035717010498045,
      "step": 770
    },
    {
      "epoch": 0.0975,
      "grad_norm": 0.0682821199297905,
      "learning_rate": 4.93232806716766e-05,
      "loss": 2.1523895263671875,
      "step": 780
    },
    {
      "epoch": 0.09875,
      "grad_norm": 0.06782221794128418,
      "learning_rate": 4.9299814027024536e-05,
      "loss": 2.199502182006836,
      "step": 790
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.06508056074380875,
      "learning_rate": 4.927595318574649e-05,
      "loss": 2.1637832641601564,
      "step": 800
    },
    {
      "epoch": 0.10125,
      "grad_norm": 0.06774035096168518,
      "learning_rate": 4.9251698534917965e-05,
      "loss": 2.227591323852539,
      "step": 810
    },
    {
      "epoch": 0.1025,
      "grad_norm": 0.06584357470273972,
      "learning_rate": 4.9227050468002954e-05,
      "loss": 2.204555130004883,
      "step": 820
    },
    {
      "epoch": 0.10375,
      "grad_norm": 0.06837441772222519,
      "learning_rate": 4.92020093848475e-05,
      "loss": 2.1817317962646485,
      "step": 830
    },
    {
      "epoch": 0.105,
      "grad_norm": 0.06870482861995697,
      "learning_rate": 4.9176575691673265e-05,
      "loss": 2.1889968872070313,
      "step": 840
    },
    {
      "epoch": 0.10625,
      "grad_norm": 0.06784000247716904,
      "learning_rate": 4.9150749801070884e-05,
      "loss": 2.1590961456298827,
      "step": 850
    },
    {
      "epoch": 0.1075,
      "grad_norm": 0.06860610842704773,
      "learning_rate": 4.912453213199332e-05,
      "loss": 2.1836381912231446,
      "step": 860
    },
    {
      "epoch": 0.10875,
      "grad_norm": 0.06371141225099564,
      "learning_rate": 4.909792310974904e-05,
      "loss": 2.136852264404297,
      "step": 870
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.06636381894350052,
      "learning_rate": 4.9070923165995135e-05,
      "loss": 2.123456573486328,
      "step": 880
    },
    {
      "epoch": 0.11125,
      "grad_norm": 0.06704560667276382,
      "learning_rate": 4.9043532738730284e-05,
      "loss": 2.210599899291992,
      "step": 890
    },
    {
      "epoch": 0.1125,
      "grad_norm": 0.06560536473989487,
      "learning_rate": 4.901575227228769e-05,
      "loss": 2.1686206817626954,
      "step": 900
    },
    {
      "epoch": 0.11375,
      "grad_norm": 0.06711073964834213,
      "learning_rate": 4.898758221732783e-05,
      "loss": 2.1291219711303713,
      "step": 910
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.07313625514507294,
      "learning_rate": 4.89590230308312e-05,
      "loss": 2.1620079040527345,
      "step": 920
    },
    {
      "epoch": 0.11625,
      "grad_norm": 0.07415210455656052,
      "learning_rate": 4.8930075176090844e-05,
      "loss": 2.2015281677246095,
      "step": 930
    },
    {
      "epoch": 0.1175,
      "grad_norm": 0.06644522398710251,
      "learning_rate": 4.890073912270486e-05,
      "loss": 2.154953384399414,
      "step": 940
    },
    {
      "epoch": 0.11875,
      "grad_norm": 0.0653102844953537,
      "learning_rate": 4.887101534656882e-05,
      "loss": 2.186296081542969,
      "step": 950
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.06608347594738007,
      "learning_rate": 4.8840904329867955e-05,
      "loss": 2.1376876831054688,
      "step": 960
    },
    {
      "epoch": 0.12125,
      "grad_norm": 0.06143832579255104,
      "learning_rate": 4.881040656106944e-05,
      "loss": 2.1134971618652343,
      "step": 970
    },
    {
      "epoch": 0.1225,
      "grad_norm": 0.06736340373754501,
      "learning_rate": 4.8779522534914414e-05,
      "loss": 2.1506475448608398,
      "step": 980
    },
    {
      "epoch": 0.12375,
      "grad_norm": 0.06936616450548172,
      "learning_rate": 4.874825275240996e-05,
      "loss": 2.125407028198242,
      "step": 990
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.07099438458681107,
      "learning_rate": 4.871659772082097e-05,
      "loss": 2.2163812637329103,
      "step": 1000
    },
    {
      "epoch": 0.12625,
      "grad_norm": 0.06724845618009567,
      "learning_rate": 4.868455795366195e-05,
      "loss": 2.1106163024902345,
      "step": 1010
    },
    {
      "epoch": 0.1275,
      "grad_norm": 0.0672367662191391,
      "learning_rate": 4.8652133970688636e-05,
      "loss": 2.071784019470215,
      "step": 1020
    },
    {
      "epoch": 0.12875,
      "grad_norm": 0.06996046751737595,
      "learning_rate": 4.861932629788962e-05,
      "loss": 2.198385238647461,
      "step": 1030
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.06979460269212723,
      "learning_rate": 4.858613546747777e-05,
      "loss": 2.111052894592285,
      "step": 1040
    },
    {
      "epoch": 0.13125,
      "grad_norm": 0.06372442096471786,
      "learning_rate": 4.8552562017881634e-05,
      "loss": 2.133512496948242,
      "step": 1050
    },
    {
      "epoch": 0.1325,
      "grad_norm": 0.06758279353380203,
      "learning_rate": 4.851860649373666e-05,
      "loss": 2.1648277282714843,
      "step": 1060
    },
    {
      "epoch": 0.13375,
      "grad_norm": 0.07092203199863434,
      "learning_rate": 4.848426944587644e-05,
      "loss": 2.1862049102783203,
      "step": 1070
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.06371688097715378,
      "learning_rate": 4.844955143132366e-05,
      "loss": 2.197888946533203,
      "step": 1080
    },
    {
      "epoch": 0.13625,
      "grad_norm": 0.06636802852153778,
      "learning_rate": 4.841445301328117e-05,
      "loss": 2.172176742553711,
      "step": 1090
    },
    {
      "epoch": 0.1375,
      "grad_norm": 0.06878487765789032,
      "learning_rate": 4.8378974761122794e-05,
      "loss": 2.2152605056762695,
      "step": 1100
    },
    {
      "epoch": 0.13875,
      "grad_norm": 0.06840436160564423,
      "learning_rate": 4.834311725038409e-05,
      "loss": 2.2093944549560547,
      "step": 1110
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.06905412673950195,
      "learning_rate": 4.830688106275305e-05,
      "loss": 2.1413171768188475,
      "step": 1120
    },
    {
      "epoch": 0.14125,
      "grad_norm": 0.06591488420963287,
      "learning_rate": 4.827026678606063e-05,
      "loss": 2.2281288146972655,
      "step": 1130
    },
    {
      "epoch": 0.1425,
      "grad_norm": 0.06814192980527878,
      "learning_rate": 4.823327501427121e-05,
      "loss": 2.241679000854492,
      "step": 1140
    },
    {
      "epoch": 0.14375,
      "grad_norm": 0.06960905343294144,
      "learning_rate": 4.8195906347473e-05,
      "loss": 2.1834861755371096,
      "step": 1150
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.07025773823261261,
      "learning_rate": 4.8158161391868276e-05,
      "loss": 2.2492515563964846,
      "step": 1160
    },
    {
      "epoch": 0.14625,
      "grad_norm": 0.06904685497283936,
      "learning_rate": 4.812004075976352e-05,
      "loss": 2.1875804901123046,
      "step": 1170
    },
    {
      "epoch": 0.1475,
      "grad_norm": 0.06693335622549057,
      "learning_rate": 4.808154506955955e-05,
      "loss": 2.23748779296875,
      "step": 1180
    },
    {
      "epoch": 0.14875,
      "grad_norm": 0.06956436485052109,
      "learning_rate": 4.804267494574145e-05,
      "loss": 2.2373470306396483,
      "step": 1190
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.07944408059120178,
      "learning_rate": 4.800343101886842e-05,
      "loss": 2.2595752716064452,
      "step": 1200
    },
    {
      "epoch": 0.15125,
      "grad_norm": 0.06570149958133698,
      "learning_rate": 4.7963813925563586e-05,
      "loss": 2.2755199432373048,
      "step": 1210
    },
    {
      "epoch": 0.1525,
      "grad_norm": 0.06871885061264038,
      "learning_rate": 4.792382430850366e-05,
      "loss": 2.2775409698486326,
      "step": 1220
    },
    {
      "epoch": 0.15375,
      "grad_norm": 0.06904297322034836,
      "learning_rate": 4.788346281640852e-05,
      "loss": 2.2760135650634767,
      "step": 1230
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.06859485059976578,
      "learning_rate": 4.784273010403065e-05,
      "loss": 2.259684753417969,
      "step": 1240
    },
    {
      "epoch": 0.15625,
      "grad_norm": 0.06464747339487076,
      "learning_rate": 4.780162683214457e-05,
      "loss": 2.308213806152344,
      "step": 1250
    },
    {
      "epoch": 0.1575,
      "grad_norm": 0.06627296656370163,
      "learning_rate": 4.77601536675361e-05,
      "loss": 2.3249114990234374,
      "step": 1260
    },
    {
      "epoch": 0.15875,
      "grad_norm": 0.06741499155759811,
      "learning_rate": 4.77183112829915e-05,
      "loss": 2.350922393798828,
      "step": 1270
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.06964308768510818,
      "learning_rate": 4.7676100357286624e-05,
      "loss": 2.244986915588379,
      "step": 1280
    },
    {
      "epoch": 0.16125,
      "grad_norm": 0.0634695515036583,
      "learning_rate": 4.7633521575175874e-05,
      "loss": 2.2805355072021483,
      "step": 1290
    },
    {
      "epoch": 0.1625,
      "grad_norm": 0.06915149837732315,
      "learning_rate": 4.759057562738109e-05,
      "loss": 2.3919076919555664,
      "step": 1300
    },
    {
      "epoch": 0.16375,
      "grad_norm": 0.06478839367628098,
      "learning_rate": 4.754726321058034e-05,
      "loss": 2.259754180908203,
      "step": 1310
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.06912291049957275,
      "learning_rate": 4.7503585027396646e-05,
      "loss": 2.2800014495849608,
      "step": 1320
    },
    {
      "epoch": 0.16625,
      "grad_norm": 0.06265991181135178,
      "learning_rate": 4.745954178638656e-05,
      "loss": 2.291925811767578,
      "step": 1330
    },
    {
      "epoch": 0.1675,
      "grad_norm": 0.06898439675569534,
      "learning_rate": 4.741513420202867e-05,
      "loss": 2.360172080993652,
      "step": 1340
    },
    {
      "epoch": 0.16875,
      "grad_norm": 0.06620923429727554,
      "learning_rate": 4.7370362994712036e-05,
      "loss": 2.368211364746094,
      "step": 1350
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.07054638862609863,
      "learning_rate": 4.732522889072447e-05,
      "loss": 2.3170215606689455,
      "step": 1360
    },
    {
      "epoch": 0.17125,
      "grad_norm": 0.07074211537837982,
      "learning_rate": 4.7279732622240766e-05,
      "loss": 2.276025581359863,
      "step": 1370
    },
    {
      "epoch": 0.1725,
      "grad_norm": 0.06905434280633926,
      "learning_rate": 4.723387492731084e-05,
      "loss": 2.2982337951660154,
      "step": 1380
    },
    {
      "epoch": 0.17375,
      "grad_norm": 0.06408198177814484,
      "learning_rate": 4.718765654984773e-05,
      "loss": 2.3407756805419924,
      "step": 1390
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.06612099707126617,
      "learning_rate": 4.7141078239615566e-05,
      "loss": 2.2696388244628904,
      "step": 1400
    },
    {
      "epoch": 0.17625,
      "grad_norm": 0.06839219480752945,
      "learning_rate": 4.709414075221734e-05,
      "loss": 2.227347755432129,
      "step": 1410
    },
    {
      "epoch": 0.1775,
      "grad_norm": 0.06748591363430023,
      "learning_rate": 4.7046844849082725e-05,
      "loss": 2.310345458984375,
      "step": 1420
    },
    {
      "epoch": 0.17875,
      "grad_norm": 0.06979987025260925,
      "learning_rate": 4.699919129745567e-05,
      "loss": 2.3062816619873048,
      "step": 1430
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.06655360758304596,
      "learning_rate": 4.695118087038198e-05,
      "loss": 2.30033016204834,
      "step": 1440
    },
    {
      "epoch": 0.18125,
      "grad_norm": 0.0673886388540268,
      "learning_rate": 4.690281434669677e-05,
      "loss": 2.3160564422607424,
      "step": 1450
    },
    {
      "epoch": 0.1825,
      "grad_norm": 0.0680936723947525,
      "learning_rate": 4.685409251101183e-05,
      "loss": 2.378013801574707,
      "step": 1460
    },
    {
      "epoch": 0.18375,
      "grad_norm": 0.07122587412595749,
      "learning_rate": 4.680501615370288e-05,
      "loss": 2.2907699584960937,
      "step": 1470
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.07071765512228012,
      "learning_rate": 4.675558607089677e-05,
      "loss": 2.3167457580566406,
      "step": 1480
    },
    {
      "epoch": 0.18625,
      "grad_norm": 0.07135665416717529,
      "learning_rate": 4.6705803064458575e-05,
      "loss": 2.36193962097168,
      "step": 1490
    },
    {
      "epoch": 0.1875,
      "grad_norm": 0.07396227866411209,
      "learning_rate": 4.665566794197854e-05,
      "loss": 2.3736957550048827,
      "step": 1500
    },
    {
      "epoch": 0.18875,
      "grad_norm": 0.06626343727111816,
      "learning_rate": 4.6605181516759047e-05,
      "loss": 2.3442031860351564,
      "step": 1510
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.06523022055625916,
      "learning_rate": 4.6554344607801335e-05,
      "loss": 2.3258100509643556,
      "step": 1520
    },
    {
      "epoch": 0.19125,
      "grad_norm": 0.06743918359279633,
      "learning_rate": 4.6503158039792324e-05,
      "loss": 2.286594009399414,
      "step": 1530
    },
    {
      "epoch": 0.1925,
      "grad_norm": 0.06837300956249237,
      "learning_rate": 4.645162264309112e-05,
      "loss": 2.320156288146973,
      "step": 1540
    },
    {
      "epoch": 0.19375,
      "grad_norm": 0.06472747027873993,
      "learning_rate": 4.639973925371562e-05,
      "loss": 2.39737548828125,
      "step": 1550
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.06730551272630692,
      "learning_rate": 4.634750871332896e-05,
      "loss": 2.3250579833984375,
      "step": 1560
    },
    {
      "epoch": 0.19625,
      "grad_norm": 0.06307782232761383,
      "learning_rate": 4.6294931869225774e-05,
      "loss": 2.300210189819336,
      "step": 1570
    },
    {
      "epoch": 0.1975,
      "grad_norm": 0.06609486788511276,
      "learning_rate": 4.6242009574318554e-05,
      "loss": 2.2991945266723635,
      "step": 1580
    },
    {
      "epoch": 0.19875,
      "grad_norm": 0.0668780580163002,
      "learning_rate": 4.6188742687123754e-05,
      "loss": 2.2811546325683594,
      "step": 1590
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.07088489830493927,
      "learning_rate": 4.6135132071747864e-05,
      "loss": 2.32708740234375,
      "step": 1600
    },
    {
      "epoch": 0.20125,
      "grad_norm": 0.06599114835262299,
      "learning_rate": 4.608117859787342e-05,
      "loss": 2.26273250579834,
      "step": 1610
    },
    {
      "epoch": 0.2025,
      "grad_norm": 0.06767764687538147,
      "learning_rate": 4.602688314074487e-05,
      "loss": 2.3097578048706056,
      "step": 1620
    },
    {
      "epoch": 0.20375,
      "grad_norm": 0.06805308163166046,
      "learning_rate": 4.597224658115438e-05,
      "loss": 2.322467231750488,
      "step": 1630
    },
    {
      "epoch": 0.205,
      "grad_norm": 0.06872619688510895,
      "learning_rate": 4.5917269805427567e-05,
      "loss": 2.303860092163086,
      "step": 1640
    },
    {
      "epoch": 0.20625,
      "grad_norm": 0.06846166402101517,
      "learning_rate": 4.5861953705409086e-05,
      "loss": 2.2648983001708984,
      "step": 1650
    },
    {
      "epoch": 0.2075,
      "grad_norm": 0.06966902315616608,
      "learning_rate": 4.58062991784482e-05,
      "loss": 2.3156864166259767,
      "step": 1660
    },
    {
      "epoch": 0.20875,
      "grad_norm": 0.06632312387228012,
      "learning_rate": 4.575030712738419e-05,
      "loss": 2.302859878540039,
      "step": 1670
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.07102814316749573,
      "learning_rate": 4.5693978460531725e-05,
      "loss": 2.3624195098876952,
      "step": 1680
    },
    {
      "epoch": 0.21125,
      "grad_norm": 0.06857810169458389,
      "learning_rate": 4.563731409166615e-05,
      "loss": 2.360994338989258,
      "step": 1690
    },
    {
      "epoch": 0.2125,
      "grad_norm": 0.06877714395523071,
      "learning_rate": 4.558031494000858e-05,
      "loss": 2.322928237915039,
      "step": 1700
    },
    {
      "epoch": 0.21375,
      "grad_norm": 0.07134252786636353,
      "learning_rate": 4.5522981930211114e-05,
      "loss": 2.2678380966186524,
      "step": 1710
    },
    {
      "epoch": 0.215,
      "grad_norm": 0.07036204636096954,
      "learning_rate": 4.546531599234173e-05,
      "loss": 2.314472770690918,
      "step": 1720
    },
    {
      "epoch": 0.21625,
      "grad_norm": 0.06794637441635132,
      "learning_rate": 4.540731806186922e-05,
      "loss": 2.385559844970703,
      "step": 1730
    },
    {
      "epoch": 0.2175,
      "grad_norm": 0.06780719757080078,
      "learning_rate": 4.5348989079648065e-05,
      "loss": 2.261928939819336,
      "step": 1740
    },
    {
      "epoch": 0.21875,
      "grad_norm": 0.06747139990329742,
      "learning_rate": 4.5290329991903115e-05,
      "loss": 2.295556831359863,
      "step": 1750
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.06255707144737244,
      "learning_rate": 4.5231341750214256e-05,
      "loss": 2.338383102416992,
      "step": 1760
    },
    {
      "epoch": 0.22125,
      "grad_norm": 0.06699945777654648,
      "learning_rate": 4.5172025311501004e-05,
      "loss": 2.2839550018310546,
      "step": 1770
    },
    {
      "epoch": 0.2225,
      "grad_norm": 0.06443452835083008,
      "learning_rate": 4.511238163800692e-05,
      "loss": 2.3070682525634765,
      "step": 1780
    },
    {
      "epoch": 0.22375,
      "grad_norm": 0.06801515817642212,
      "learning_rate": 4.505241169728407e-05,
      "loss": 2.313974952697754,
      "step": 1790
    },
    {
      "epoch": 0.225,
      "grad_norm": 0.06966620683670044,
      "learning_rate": 4.499211646217727e-05,
      "loss": 2.2580539703369142,
      "step": 1800
    }
  ],
  "logging_steps": 10,
  "max_steps": 8000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1075411496665088e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}