{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8239607794668974,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0032958431178675895,
      "grad_norm": 74.5,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.7989,
      "step": 10
    },
    {
      "epoch": 0.006591686235735179,
      "grad_norm": 212.0,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.7703,
      "step": 20
    },
    {
      "epoch": 0.009887529353602768,
      "grad_norm": 482.0,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.821,
      "step": 30
    },
    {
      "epoch": 0.013183372471470358,
      "grad_norm": 396.0,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.8979,
      "step": 40
    },
    {
      "epoch": 0.01647921558933795,
      "grad_norm": 488.0,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.17,
      "step": 50
    },
    {
      "epoch": 0.019775058707205535,
      "grad_norm": 608.0,
      "learning_rate": 4.800000000000001e-06,
      "loss": 1.4005,
      "step": 60
    },
    {
      "epoch": 0.023070901825073126,
      "grad_norm": 452.0,
      "learning_rate": 5.600000000000001e-06,
      "loss": 1.7127,
      "step": 70
    },
    {
      "epoch": 0.026366744942940716,
      "grad_norm": 442.0,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 1.4108,
      "step": 80
    },
    {
      "epoch": 0.029662588060808306,
      "grad_norm": 111.0,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 2.3574,
      "step": 90
    },
    {
      "epoch": 0.0329584311786759,
      "grad_norm": 109.5,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.8115,
      "step": 100
    },
    {
      "epoch": 0.03625427429654349,
      "grad_norm": 41.25,
      "learning_rate": 8.8e-06,
      "loss": 1.5844,
      "step": 110
    },
    {
      "epoch": 0.03955011741441107,
      "grad_norm": 50.0,
      "learning_rate": 9.600000000000001e-06,
      "loss": 1.5766,
      "step": 120
    },
    {
      "epoch": 0.04284596053227866,
      "grad_norm": 155.0,
      "learning_rate": 1.04e-05,
      "loss": 5.5685,
      "step": 130
    },
    {
      "epoch": 0.04614180365014625,
      "grad_norm": 102.0,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 1.4693,
      "step": 140
    },
    {
      "epoch": 0.04943764676801384,
      "grad_norm": 256.0,
      "learning_rate": 1.2e-05,
      "loss": 1.4063,
      "step": 150
    },
    {
      "epoch": 0.05273348988588143,
      "grad_norm": 81.0,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 1.2139,
      "step": 160
    },
    {
      "epoch": 0.05602933300374902,
      "grad_norm": 61.75,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 1.2504,
      "step": 170
    },
    {
      "epoch": 0.05932517612161661,
      "grad_norm": 12.0625,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 1.338,
      "step": 180
    },
    {
      "epoch": 0.0626210192394842,
      "grad_norm": 36.5,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 1.2007,
      "step": 190
    },
    {
      "epoch": 0.0659168623573518,
      "grad_norm": 35.0,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.2368,
      "step": 200
    },
    {
      "epoch": 0.06921270547521938,
      "grad_norm": 17.125,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 2.1113,
      "step": 210
    },
    {
      "epoch": 0.07250854859308697,
      "grad_norm": 29.125,
      "learning_rate": 1.76e-05,
      "loss": 1.2671,
      "step": 220
    },
    {
      "epoch": 0.07580439171095456,
      "grad_norm": 43.0,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 1.2318,
      "step": 230
    },
    {
      "epoch": 0.07910023482882214,
      "grad_norm": 26.625,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 1.0837,
      "step": 240
    },
    {
      "epoch": 0.08239607794668974,
      "grad_norm": 26.5,
      "learning_rate": 2e-05,
      "loss": 1.2709,
      "step": 250
    },
    {
      "epoch": 0.08569192106455732,
      "grad_norm": 23.875,
      "learning_rate": 1.9999025240093045e-05,
      "loss": 1.0127,
      "step": 260
    },
    {
      "epoch": 0.08898776418242492,
      "grad_norm": 22.875,
      "learning_rate": 1.9996101150403543e-05,
      "loss": 1.0939,
      "step": 270
    },
    {
      "epoch": 0.0922836073002925,
      "grad_norm": 15.1875,
      "learning_rate": 1.9991228300988586e-05,
      "loss": 1.1039,
      "step": 280
    },
    {
      "epoch": 0.0955794504181601,
      "grad_norm": 44.25,
      "learning_rate": 1.9984407641819812e-05,
      "loss": 1.1438,
      "step": 290
    },
    {
      "epoch": 0.09887529353602768,
      "grad_norm": 27.25,
      "learning_rate": 1.9975640502598243e-05,
      "loss": 1.1164,
      "step": 300
    },
    {
      "epoch": 0.10217113665389528,
      "grad_norm": 31.25,
      "learning_rate": 1.9964928592495046e-05,
      "loss": 1.051,
      "step": 310
    },
    {
      "epoch": 0.10546697977176286,
      "grad_norm": 15.625,
      "learning_rate": 1.9952273999818312e-05,
      "loss": 0.9002,
      "step": 320
    },
    {
      "epoch": 0.10876282288963045,
      "grad_norm": 14.75,
      "learning_rate": 1.9937679191605964e-05,
      "loss": 0.9201,
      "step": 330
    },
    {
      "epoch": 0.11205866600749805,
      "grad_norm": 41.0,
      "learning_rate": 1.9921147013144782e-05,
      "loss": 1.0844,
      "step": 340
    },
    {
      "epoch": 0.11535450912536563,
      "grad_norm": 17.875,
      "learning_rate": 1.9902680687415704e-05,
      "loss": 0.946,
      "step": 350
    },
    {
      "epoch": 0.11865035224323323,
      "grad_norm": 13.1875,
      "learning_rate": 1.988228381446553e-05,
      "loss": 0.9666,
      "step": 360
    },
    {
      "epoch": 0.12194619536110081,
      "grad_norm": 15.4375,
      "learning_rate": 1.985996037070505e-05,
      "loss": 0.9573,
      "step": 370
    },
    {
      "epoch": 0.1252420384789684,
      "grad_norm": 19.125,
      "learning_rate": 1.983571470813386e-05,
      "loss": 1.0149,
      "step": 380
    },
    {
      "epoch": 0.128537881596836,
      "grad_norm": 17.125,
      "learning_rate": 1.9809551553491918e-05,
      "loss": 1.0354,
      "step": 390
    },
    {
      "epoch": 0.1318337247147036,
      "grad_norm": 21.75,
      "learning_rate": 1.9781476007338058e-05,
      "loss": 1.0042,
      "step": 400
    },
    {
      "epoch": 0.13512956783257116,
      "grad_norm": 12.4375,
      "learning_rate": 1.9751493543055634e-05,
      "loss": 0.8969,
      "step": 410
    },
    {
      "epoch": 0.13842541095043875,
      "grad_norm": 14.75,
      "learning_rate": 1.9719610005785466e-05,
      "loss": 0.9187,
      "step": 420
    },
    {
      "epoch": 0.14172125406830635,
      "grad_norm": 16.875,
      "learning_rate": 1.9685831611286312e-05,
      "loss": 0.918,
      "step": 430
    },
    {
      "epoch": 0.14501709718617395,
      "grad_norm": 28.0,
      "learning_rate": 1.9650164944723116e-05,
      "loss": 0.8747,
      "step": 440
    },
    {
      "epoch": 0.14831294030404152,
      "grad_norm": 15.375,
      "learning_rate": 1.961261695938319e-05,
      "loss": 0.9015,
      "step": 450
    },
    {
      "epoch": 0.15160878342190912,
      "grad_norm": 26.75,
      "learning_rate": 1.9573194975320672e-05,
      "loss": 0.9056,
      "step": 460
    },
    {
      "epoch": 0.1549046265397767,
      "grad_norm": 28.75,
      "learning_rate": 1.9531906677929472e-05,
      "loss": 0.859,
      "step": 470
    },
    {
      "epoch": 0.15820046965764428,
      "grad_norm": 12.5,
      "learning_rate": 1.9488760116444966e-05,
      "loss": 0.8454,
      "step": 480
    },
    {
      "epoch": 0.16149631277551188,
      "grad_norm": 19.75,
      "learning_rate": 1.944376370237481e-05,
      "loss": 0.9058,
      "step": 490
    },
    {
      "epoch": 0.16479215589337948,
      "grad_norm": 18.0,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.8204,
      "step": 500
    },
    {
      "epoch": 0.16808799901124707,
      "grad_norm": 8.4375,
      "learning_rate": 1.9348256763960146e-05,
      "loss": 0.8926,
      "step": 510
    },
    {
      "epoch": 0.17138384212911464,
      "grad_norm": 12.0625,
      "learning_rate": 1.9297764858882516e-05,
      "loss": 0.8243,
      "step": 520
    },
    {
      "epoch": 0.17467968524698224,
      "grad_norm": 16.125,
      "learning_rate": 1.9245460336123136e-05,
      "loss": 0.8746,
      "step": 530
    },
    {
      "epoch": 0.17797552836484984,
      "grad_norm": 20.25,
      "learning_rate": 1.9191353392552346e-05,
      "loss": 0.878,
      "step": 540
    },
    {
      "epoch": 0.18127137148271744,
      "grad_norm": 17.75,
      "learning_rate": 1.913545457642601e-05,
      "loss": 0.7852,
      "step": 550
    },
    {
      "epoch": 0.184567214600585,
      "grad_norm": 7.625,
      "learning_rate": 1.907777478532909e-05,
      "loss": 0.7749,
      "step": 560
    },
    {
      "epoch": 0.1878630577184526,
      "grad_norm": 18.5,
      "learning_rate": 1.901832526405114e-05,
      "loss": 0.9067,
      "step": 570
    },
    {
      "epoch": 0.1911589008363202,
      "grad_norm": 15.5625,
      "learning_rate": 1.895711760239413e-05,
      "loss": 0.8603,
      "step": 580
    },
    {
      "epoch": 0.19445474395418777,
      "grad_norm": 16.375,
      "learning_rate": 1.889416373291298e-05,
      "loss": 0.8449,
      "step": 590
    },
    {
      "epoch": 0.19775058707205537,
      "grad_norm": 12.4375,
      "learning_rate": 1.8829475928589272e-05,
      "loss": 0.8277,
      "step": 600
    },
    {
      "epoch": 0.20104643018992296,
      "grad_norm": 16.0,
      "learning_rate": 1.8763066800438638e-05,
      "loss": 0.8477,
      "step": 610
    },
    {
      "epoch": 0.20434227330779056,
      "grad_norm": 13.6875,
      "learning_rate": 1.869494929505219e-05,
      "loss": 0.7843,
      "step": 620
    },
    {
      "epoch": 0.20763811642565813,
      "grad_norm": 8.0,
      "learning_rate": 1.8625136692072577e-05,
      "loss": 0.7957,
      "step": 630
    },
    {
      "epoch": 0.21093395954352573,
      "grad_norm": 11.8125,
      "learning_rate": 1.855364260160507e-05,
      "loss": 0.8254,
      "step": 640
    },
    {
      "epoch": 0.21422980266139333,
      "grad_norm": 8.6875,
      "learning_rate": 1.848048096156426e-05,
      "loss": 0.8332,
      "step": 650
    },
    {
      "epoch": 0.2175256457792609,
      "grad_norm": 9.1875,
      "learning_rate": 1.8405666034956842e-05,
      "loss": 0.8352,
      "step": 660
    },
    {
      "epoch": 0.2208214888971285,
      "grad_norm": 20.125,
      "learning_rate": 1.8329212407100996e-05,
      "loss": 0.8683,
      "step": 670
    },
    {
      "epoch": 0.2241173320149961,
      "grad_norm": 14.25,
      "learning_rate": 1.8251134982782952e-05,
      "loss": 0.7979,
      "step": 680
    },
    {
      "epoch": 0.2274131751328637,
      "grad_norm": 17.625,
      "learning_rate": 1.8171448983351284e-05,
      "loss": 0.8485,
      "step": 690
    },
    {
      "epoch": 0.23070901825073126,
      "grad_norm": 15.0625,
      "learning_rate": 1.8090169943749477e-05,
      "loss": 0.7674,
      "step": 700
    },
    {
      "epoch": 0.23400486136859885,
      "grad_norm": 4.96875,
      "learning_rate": 1.8007313709487334e-05,
      "loss": 0.7827,
      "step": 710
    },
    {
      "epoch": 0.23730070448646645,
      "grad_norm": 13.8125,
      "learning_rate": 1.792289643355191e-05,
      "loss": 0.7801,
      "step": 720
    },
    {
      "epoch": 0.24059654760433402,
      "grad_norm": 8.6875,
      "learning_rate": 1.78369345732584e-05,
      "loss": 0.8158,
      "step": 730
    },
    {
      "epoch": 0.24389239072220162,
      "grad_norm": 20.875,
      "learning_rate": 1.7749444887041797e-05,
      "loss": 0.777,
      "step": 740
    },
    {
      "epoch": 0.24718823384006922,
      "grad_norm": 11.0,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.7895,
      "step": 750
    },
    {
      "epoch": 0.2504840769579368,
      "grad_norm": 12.5,
      "learning_rate": 1.7569950556517566e-05,
      "loss": 0.7931,
      "step": 760
    },
    {
      "epoch": 0.2537799200758044,
      "grad_norm": 13.6875,
      "learning_rate": 1.747798090498532e-05,
      "loss": 0.8357,
      "step": 770
    },
    {
      "epoch": 0.257075763193672,
      "grad_norm": 12.0625,
      "learning_rate": 1.7384553406258842e-05,
      "loss": 0.7632,
      "step": 780
    },
    {
      "epoch": 0.26037160631153955,
      "grad_norm": 8.625,
      "learning_rate": 1.7289686274214116e-05,
      "loss": 0.7747,
      "step": 790
    },
    {
      "epoch": 0.2636674494294072,
      "grad_norm": 18.75,
      "learning_rate": 1.7193398003386514e-05,
      "loss": 0.7421,
      "step": 800
    },
    {
      "epoch": 0.26696329254727474,
      "grad_norm": 8.625,
      "learning_rate": 1.709570736536521e-05,
      "loss": 0.7616,
      "step": 810
    },
    {
      "epoch": 0.2702591356651423,
      "grad_norm": 14.0,
      "learning_rate": 1.6996633405133656e-05,
      "loss": 0.752,
      "step": 820
    },
    {
      "epoch": 0.27355497878300994,
      "grad_norm": 14.5625,
      "learning_rate": 1.68961954373567e-05,
      "loss": 0.7501,
      "step": 830
    },
    {
      "epoch": 0.2768508219008775,
      "grad_norm": 19.125,
      "learning_rate": 1.6794413042615168e-05,
      "loss": 0.7874,
      "step": 840
    },
    {
      "epoch": 0.28014666501874513,
      "grad_norm": 13.6875,
      "learning_rate": 1.6691306063588583e-05,
      "loss": 0.7667,
      "step": 850
    },
    {
      "epoch": 0.2834425081366127,
      "grad_norm": 11.9375,
      "learning_rate": 1.6586894601186804e-05,
      "loss": 0.7087,
      "step": 860
    },
    {
      "epoch": 0.2867383512544803,
      "grad_norm": 8.5,
      "learning_rate": 1.6481199010631312e-05,
      "loss": 0.7752,
      "step": 870
    },
    {
      "epoch": 0.2900341943723479,
      "grad_norm": 7.59375,
      "learning_rate": 1.63742398974869e-05,
      "loss": 0.7764,
      "step": 880
    },
    {
      "epoch": 0.29333003749021547,
      "grad_norm": 8.875,
      "learning_rate": 1.6266038113644605e-05,
      "loss": 0.7502,
      "step": 890
    },
    {
      "epoch": 0.29662588060808304,
      "grad_norm": 9.1875,
      "learning_rate": 1.6156614753256583e-05,
      "loss": 0.7126,
      "step": 900
    },
    {
      "epoch": 0.29992172372595066,
      "grad_norm": 14.8125,
      "learning_rate": 1.6045991148623752e-05,
      "loss": 0.778,
      "step": 910
    },
    {
      "epoch": 0.30321756684381823,
      "grad_norm": 10.375,
      "learning_rate": 1.5934188866037017e-05,
      "loss": 0.7234,
      "step": 920
    },
    {
      "epoch": 0.3065134099616858,
      "grad_norm": 11.1875,
      "learning_rate": 1.5821229701572897e-05,
      "loss": 0.8043,
      "step": 930
    },
    {
      "epoch": 0.3098092530795534,
      "grad_norm": 12.9375,
      "learning_rate": 1.570713567684432e-05,
      "loss": 0.7271,
      "step": 940
    },
    {
      "epoch": 0.313105096197421,
      "grad_norm": 4.21875,
      "learning_rate": 1.5591929034707468e-05,
      "loss": 0.6921,
      "step": 950
    },
    {
      "epoch": 0.31640093931528857,
      "grad_norm": 8.625,
      "learning_rate": 1.5475632234925505e-05,
      "loss": 0.78,
      "step": 960
    },
    {
      "epoch": 0.3196967824331562,
      "grad_norm": 7.34375,
      "learning_rate": 1.5358267949789968e-05,
      "loss": 0.7616,
      "step": 970
    },
    {
      "epoch": 0.32299262555102376,
      "grad_norm": 7.8125,
      "learning_rate": 1.5239859059700794e-05,
      "loss": 0.7442,
      "step": 980
    },
    {
      "epoch": 0.3262884686688914,
      "grad_norm": 12.5625,
      "learning_rate": 1.5120428648705716e-05,
      "loss": 0.7209,
      "step": 990
    },
    {
      "epoch": 0.32958431178675895,
      "grad_norm": 11.3125,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7091,
      "step": 1000
    },
    {
      "epoch": 0.3328801549046265,
      "grad_norm": 5.03125,
      "learning_rate": 1.4878596591387329e-05,
      "loss": 0.7082,
      "step": 1010
    },
    {
      "epoch": 0.33617599802249415,
      "grad_norm": 4.03125,
      "learning_rate": 1.4756242090702756e-05,
      "loss": 0.6973,
      "step": 1020
    },
    {
      "epoch": 0.3394718411403617,
      "grad_norm": 13.625,
      "learning_rate": 1.463296035119862e-05,
      "loss": 0.7182,
      "step": 1030
    },
    {
      "epoch": 0.3427676842582293,
      "grad_norm": 15.6875,
      "learning_rate": 1.4508775406894308e-05,
      "loss": 0.6958,
      "step": 1040
    },
    {
      "epoch": 0.3460635273760969,
      "grad_norm": 16.75,
      "learning_rate": 1.4383711467890776e-05,
      "loss": 0.6861,
      "step": 1050
    },
    {
      "epoch": 0.3493593704939645,
      "grad_norm": 5.0,
      "learning_rate": 1.4257792915650728e-05,
      "loss": 0.675,
      "step": 1060
    },
    {
      "epoch": 0.35265521361183205,
      "grad_norm": 10.125,
      "learning_rate": 1.413104429824542e-05,
      "loss": 0.7937,
      "step": 1070
    },
    {
      "epoch": 0.3559510567296997,
      "grad_norm": 4.28125,
      "learning_rate": 1.4003490325568953e-05,
      "loss": 0.7161,
      "step": 1080
    },
    {
      "epoch": 0.35924689984756725,
      "grad_norm": 7.03125,
      "learning_rate": 1.3875155864521031e-05,
      "loss": 0.713,
      "step": 1090
    },
    {
      "epoch": 0.36254274296543487,
      "grad_norm": 19.0,
      "learning_rate": 1.3746065934159123e-05,
      "loss": 0.6866,
      "step": 1100
    },
    {
      "epoch": 0.36583858608330244,
      "grad_norm": 11.5625,
      "learning_rate": 1.3616245700820922e-05,
      "loss": 0.7129,
      "step": 1110
    },
    {
      "epoch": 0.36913442920117,
      "grad_norm": 4.03125,
      "learning_rate": 1.3485720473218153e-05,
      "loss": 0.7217,
      "step": 1120
    },
    {
      "epoch": 0.37243027231903764,
      "grad_norm": 7.78125,
      "learning_rate": 1.3354515697502552e-05,
      "loss": 0.7008,
      "step": 1130
    },
    {
      "epoch": 0.3757261154369052,
      "grad_norm": 9.25,
      "learning_rate": 1.3222656952305113e-05,
      "loss": 0.6741,
      "step": 1140
    },
    {
      "epoch": 0.3790219585547728,
      "grad_norm": 7.3125,
      "learning_rate": 1.3090169943749475e-05,
      "loss": 0.7249,
      "step": 1150
    },
    {
      "epoch": 0.3823178016726404,
      "grad_norm": 4.53125,
      "learning_rate": 1.2957080500440469e-05,
      "loss": 0.6616,
      "step": 1160
    },
    {
      "epoch": 0.38561364479050797,
      "grad_norm": 3.5625,
      "learning_rate": 1.2823414568428767e-05,
      "loss": 0.6848,
      "step": 1170
    },
    {
      "epoch": 0.38890948790837554,
      "grad_norm": 9.875,
      "learning_rate": 1.2689198206152657e-05,
      "loss": 0.6417,
      "step": 1180
    },
    {
      "epoch": 0.39220533102624316,
      "grad_norm": 8.75,
      "learning_rate": 1.2554457579357906e-05,
      "loss": 0.6812,
      "step": 1190
    },
    {
      "epoch": 0.39550117414411073,
      "grad_norm": 6.3125,
      "learning_rate": 1.2419218955996677e-05,
      "loss": 0.6304,
      "step": 1200
    },
    {
      "epoch": 0.3987970172619783,
      "grad_norm": 5.28125,
      "learning_rate": 1.2283508701106559e-05,
      "loss": 0.6765,
      "step": 1210
    },
    {
      "epoch": 0.40209286037984593,
      "grad_norm": 6.4375,
      "learning_rate": 1.2147353271670634e-05,
      "loss": 0.6552,
      "step": 1220
    },
    {
      "epoch": 0.4053887034977135,
      "grad_norm": 9.375,
      "learning_rate": 1.2010779211459649e-05,
      "loss": 0.6989,
      "step": 1230
    },
    {
      "epoch": 0.4086845466155811,
      "grad_norm": 4.875,
      "learning_rate": 1.187381314585725e-05,
      "loss": 0.6349,
      "step": 1240
    },
    {
      "epoch": 0.4119803897334487,
      "grad_norm": 13.625,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.6537,
      "step": 1250
    },
    {
      "epoch": 0.41527623285131626,
      "grad_norm": 3.609375,
      "learning_rate": 1.159881187691835e-05,
      "loss": 0.6216,
      "step": 1260
    },
    {
      "epoch": 0.4185720759691839,
      "grad_norm": 9.9375,
      "learning_rate": 1.1460830285624119e-05,
      "loss": 0.6572,
      "step": 1270
    },
    {
      "epoch": 0.42186791908705146,
      "grad_norm": 6.84375,
      "learning_rate": 1.1322563902571227e-05,
      "loss": 0.6511,
      "step": 1280
    },
    {
      "epoch": 0.425163762204919,
      "grad_norm": 7.09375,
      "learning_rate": 1.1184039683065014e-05,
      "loss": 0.6671,
      "step": 1290
    },
    {
      "epoch": 0.42845960532278665,
      "grad_norm": 4.875,
      "learning_rate": 1.1045284632676535e-05,
      "loss": 0.6279,
      "step": 1300
    },
    {
      "epoch": 0.4317554484406542,
      "grad_norm": 8.6875,
      "learning_rate": 1.0906325801977804e-05,
      "loss": 0.6442,
      "step": 1310
    },
    {
      "epoch": 0.4350512915585218,
      "grad_norm": 22.75,
      "learning_rate": 1.0767190281268187e-05,
      "loss": 0.6786,
      "step": 1320
    },
    {
      "epoch": 0.4383471346763894,
      "grad_norm": 5.625,
      "learning_rate": 1.0627905195293135e-05,
      "loss": 0.6482,
      "step": 1330
    },
    {
      "epoch": 0.441642977794257,
      "grad_norm": 10.5,
      "learning_rate": 1.0488497697956134e-05,
      "loss": 0.6611,
      "step": 1340
    },
    {
      "epoch": 0.44493882091212456,
      "grad_norm": 3.109375,
      "learning_rate": 1.0348994967025012e-05,
      "loss": 0.6093,
      "step": 1350
    },
    {
      "epoch": 0.4482346640299922,
      "grad_norm": 4.21875,
      "learning_rate": 1.0209424198833571e-05,
      "loss": 0.7027,
      "step": 1360
    },
    {
      "epoch": 0.45153050714785975,
      "grad_norm": 3.546875,
      "learning_rate": 1.0069812602979617e-05,
      "loss": 0.665,
      "step": 1370
    },
    {
      "epoch": 0.4548263502657274,
      "grad_norm": 8.0,
      "learning_rate": 9.930187397020385e-06,
      "loss": 0.6277,
      "step": 1380
    },
    {
      "epoch": 0.45812219338359494,
      "grad_norm": 12.9375,
      "learning_rate": 9.790575801166432e-06,
      "loss": 0.6452,
      "step": 1390
    },
    {
      "epoch": 0.4614180365014625,
      "grad_norm": 8.75,
      "learning_rate": 9.651005032974994e-06,
      "loss": 0.6493,
      "step": 1400
    },
    {
      "epoch": 0.46471387961933014,
      "grad_norm": 5.53125,
      "learning_rate": 9.511502302043867e-06,
      "loss": 0.6633,
      "step": 1410
    },
    {
      "epoch": 0.4680097227371977,
      "grad_norm": 5.4375,
      "learning_rate": 9.372094804706867e-06,
      "loss": 0.6568,
      "step": 1420
    },
    {
      "epoch": 0.4713055658550653,
      "grad_norm": 8.1875,
      "learning_rate": 9.232809718731815e-06,
      "loss": 0.6294,
      "step": 1430
    },
    {
      "epoch": 0.4746014089729329,
      "grad_norm": 6.3125,
      "learning_rate": 9.093674198022201e-06,
      "loss": 0.6133,
      "step": 1440
    },
    {
      "epoch": 0.4778972520908005,
      "grad_norm": 4.8125,
      "learning_rate": 8.954715367323468e-06,
      "loss": 0.6392,
      "step": 1450
    },
    {
      "epoch": 0.48119309520866804,
      "grad_norm": 3.34375,
      "learning_rate": 8.815960316934991e-06,
      "loss": 0.6103,
      "step": 1460
    },
    {
      "epoch": 0.48448893832653567,
      "grad_norm": 3.375,
      "learning_rate": 8.677436097428775e-06,
      "loss": 0.6069,
      "step": 1470
    },
    {
      "epoch": 0.48778478144440324,
      "grad_norm": 11.4375,
      "learning_rate": 8.539169714375885e-06,
      "loss": 0.7006,
      "step": 1480
    },
    {
      "epoch": 0.49108062456227086,
      "grad_norm": 6.34375,
      "learning_rate": 8.401188123081653e-06,
      "loss": 0.6163,
      "step": 1490
    },
    {
      "epoch": 0.49437646768013843,
      "grad_norm": 7.78125,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.6063,
      "step": 1500
    },
    {
      "epoch": 0.497672310798006,
      "grad_norm": 6.28125,
      "learning_rate": 8.126186854142752e-06,
      "loss": 0.6397,
      "step": 1510
    },
    {
      "epoch": 0.5009681539158736,
      "grad_norm": 5.03125,
      "learning_rate": 7.989220788540356e-06,
      "loss": 0.6607,
      "step": 1520
    },
    {
      "epoch": 0.5042639970337411,
      "grad_norm": 2.90625,
      "learning_rate": 7.852646728329368e-06,
      "loss": 0.592,
      "step": 1530
    },
    {
      "epoch": 0.5075598401516088,
      "grad_norm": 7.09375,
      "learning_rate": 7.716491298893443e-06,
      "loss": 0.6278,
      "step": 1540
    },
    {
      "epoch": 0.5108556832694764,
      "grad_norm": 4.375,
      "learning_rate": 7.580781044003324e-06,
      "loss": 0.6428,
      "step": 1550
    },
    {
      "epoch": 0.514151526387344,
      "grad_norm": 5.5625,
      "learning_rate": 7.445542420642097e-06,
      "loss": 0.6257,
      "step": 1560
    },
    {
      "epoch": 0.5174473695052115,
      "grad_norm": 4.34375,
      "learning_rate": 7.310801793847344e-06,
      "loss": 0.6454,
      "step": 1570
    },
    {
      "epoch": 0.5207432126230791,
      "grad_norm": 5.5625,
      "learning_rate": 7.176585431571235e-06,
      "loss": 0.6416,
      "step": 1580
    },
    {
      "epoch": 0.5240390557409468,
      "grad_norm": 3.828125,
      "learning_rate": 7.042919499559538e-06,
      "loss": 0.6382,
      "step": 1590
    },
    {
      "epoch": 0.5273348988588143,
      "grad_norm": 8.4375,
      "learning_rate": 6.909830056250527e-06,
      "loss": 0.6197,
      "step": 1600
    },
    {
      "epoch": 0.5306307419766819,
      "grad_norm": 6.5,
      "learning_rate": 6.777343047694891e-06,
      "loss": 0.6428,
      "step": 1610
    },
    {
      "epoch": 0.5339265850945495,
      "grad_norm": 3.40625,
      "learning_rate": 6.645484302497452e-06,
      "loss": 0.5975,
      "step": 1620
    },
    {
      "epoch": 0.5372224282124171,
      "grad_norm": 3.96875,
      "learning_rate": 6.5142795267818505e-06,
      "loss": 0.6372,
      "step": 1630
    },
    {
      "epoch": 0.5405182713302846,
      "grad_norm": 4.96875,
      "learning_rate": 6.383754299179079e-06,
      "loss": 0.634,
      "step": 1640
    },
    {
      "epoch": 0.5438141144481523,
      "grad_norm": 10.625,
      "learning_rate": 6.25393406584088e-06,
      "loss": 0.6324,
      "step": 1650
    },
    {
      "epoch": 0.5471099575660199,
      "grad_norm": 3.734375,
      "learning_rate": 6.124844135478971e-06,
      "loss": 0.6464,
      "step": 1660
    },
    {
      "epoch": 0.5504058006838874,
      "grad_norm": 3.265625,
      "learning_rate": 5.996509674431053e-06,
      "loss": 0.6476,
      "step": 1670
    },
    {
      "epoch": 0.553701643801755,
      "grad_norm": 5.125,
      "learning_rate": 5.868955701754584e-06,
      "loss": 0.6262,
      "step": 1680
    },
    {
      "epoch": 0.5569974869196226,
      "grad_norm": 6.21875,
      "learning_rate": 5.742207084349274e-06,
      "loss": 0.5765,
      "step": 1690
    },
    {
      "epoch": 0.5602933300374903,
      "grad_norm": 2.4375,
      "learning_rate": 5.616288532109225e-06,
      "loss": 0.6095,
      "step": 1700
    },
    {
      "epoch": 0.5635891731553578,
      "grad_norm": 4.96875,
      "learning_rate": 5.491224593105695e-06,
      "loss": 0.6332,
      "step": 1710
    },
    {
      "epoch": 0.5668850162732254,
      "grad_norm": 3.765625,
      "learning_rate": 5.367039648801386e-06,
      "loss": 0.6095,
      "step": 1720
    },
    {
      "epoch": 0.570180859391093,
      "grad_norm": 4.4375,
      "learning_rate": 5.243757909297247e-06,
      "loss": 0.5548,
      "step": 1730
    },
    {
      "epoch": 0.5734767025089605,
      "grad_norm": 5.71875,
      "learning_rate": 5.121403408612672e-06,
      "loss": 0.5679,
      "step": 1740
    },
    {
      "epoch": 0.5767725456268281,
      "grad_norm": 2.96875,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.5992,
      "step": 1750
    },
    {
      "epoch": 0.5800683887446958,
      "grad_norm": 6.21875,
      "learning_rate": 4.879571351294287e-06,
      "loss": 0.6256,
      "step": 1760
    },
    {
      "epoch": 0.5833642318625634,
      "grad_norm": 3.953125,
      "learning_rate": 4.76014094029921e-06,
      "loss": 0.6058,
      "step": 1770
    },
    {
      "epoch": 0.5866600749804309,
      "grad_norm": 13.6875,
      "learning_rate": 4.641732050210032e-06,
      "loss": 0.6687,
      "step": 1780
    },
    {
      "epoch": 0.5899559180982985,
      "grad_norm": 2.9375,
      "learning_rate": 4.524367765074499e-06,
      "loss": 0.6087,
      "step": 1790
    },
    {
      "epoch": 0.5932517612161661,
      "grad_norm": 3.921875,
      "learning_rate": 4.408070965292534e-06,
      "loss": 0.6073,
      "step": 1800
    },
    {
      "epoch": 0.5965476043340338,
      "grad_norm": 3.59375,
      "learning_rate": 4.292864323155684e-06,
      "loss": 0.583,
      "step": 1810
    },
    {
      "epoch": 0.5998434474519013,
      "grad_norm": 4.875,
      "learning_rate": 4.178770298427107e-06,
      "loss": 0.5835,
      "step": 1820
    },
    {
      "epoch": 0.6031392905697689,
      "grad_norm": 4.375,
      "learning_rate": 4.065811133962987e-06,
      "loss": 0.5982,
      "step": 1830
    },
    {
      "epoch": 0.6064351336876365,
      "grad_norm": 3.640625,
      "learning_rate": 3.954008851376252e-06,
      "loss": 0.5824,
      "step": 1840
    },
    {
      "epoch": 0.609730976805504,
      "grad_norm": 4.75,
      "learning_rate": 3.8433852467434175e-06,
      "loss": 0.5917,
      "step": 1850
    },
    {
      "epoch": 0.6130268199233716,
      "grad_norm": 6.0625,
      "learning_rate": 3.7339618863553983e-06,
      "loss": 0.5987,
      "step": 1860
    },
    {
      "epoch": 0.6163226630412393,
      "grad_norm": 3.140625,
      "learning_rate": 3.625760102513103e-06,
      "loss": 0.6388,
      "step": 1870
    },
    {
      "epoch": 0.6196185061591069,
      "grad_norm": 3.59375,
      "learning_rate": 3.5188009893686916e-06,
      "loss": 0.5854,
      "step": 1880
    },
    {
      "epoch": 0.6229143492769744,
      "grad_norm": 3.21875,
      "learning_rate": 3.4131053988131947e-06,
      "loss": 0.6122,
      "step": 1890
    },
    {
      "epoch": 0.626210192394842,
      "grad_norm": 5.03125,
      "learning_rate": 3.308693936411421e-06,
      "loss": 0.5812,
      "step": 1900
    },
    {
      "epoch": 0.6295060355127096,
      "grad_norm": 3.59375,
      "learning_rate": 3.2055869573848374e-06,
      "loss": 0.5677,
      "step": 1910
    },
    {
      "epoch": 0.6328018786305771,
      "grad_norm": 3.390625,
      "learning_rate": 3.103804562643302e-06,
      "loss": 0.6102,
      "step": 1920
    },
    {
      "epoch": 0.6360977217484448,
      "grad_norm": 8.125,
      "learning_rate": 3.003366594866345e-06,
      "loss": 0.6108,
      "step": 1930
    },
    {
      "epoch": 0.6393935648663124,
      "grad_norm": 7.96875,
      "learning_rate": 2.9042926346347932e-06,
      "loss": 0.5646,
      "step": 1940
    },
    {
      "epoch": 0.64268940798418,
      "grad_norm": 3.5,
      "learning_rate": 2.8066019966134907e-06,
      "loss": 0.549,
      "step": 1950
    },
    {
      "epoch": 0.6459852511020475,
      "grad_norm": 3.015625,
      "learning_rate": 2.7103137257858867e-06,
      "loss": 0.6216,
      "step": 1960
    },
    {
      "epoch": 0.6492810942199151,
      "grad_norm": 5.625,
      "learning_rate": 2.615446593741161e-06,
      "loss": 0.6014,
      "step": 1970
    },
    {
      "epoch": 0.6525769373377828,
      "grad_norm": 5.09375,
      "learning_rate": 2.522019095014683e-06,
      "loss": 0.546,
      "step": 1980
    },
    {
      "epoch": 0.6558727804556503,
      "grad_norm": 2.390625,
      "learning_rate": 2.4300494434824373e-06,
      "loss": 0.5791,
      "step": 1990
    },
    {
      "epoch": 0.6591686235735179,
      "grad_norm": 6.03125,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.6427,
      "step": 2000
    },
    {
      "epoch": 0.6624644666913855,
      "grad_norm": 3.421875,
      "learning_rate": 2.2505551129582047e-06,
      "loss": 0.5651,
      "step": 2010
    },
    {
      "epoch": 0.665760309809253,
      "grad_norm": 4.71875,
      "learning_rate": 2.163065426741603e-06,
      "loss": 0.5815,
      "step": 2020
    },
    {
      "epoch": 0.6690561529271206,
      "grad_norm": 5.28125,
      "learning_rate": 2.0771035664480944e-06,
      "loss": 0.6387,
      "step": 2030
    },
    {
      "epoch": 0.6723519960449883,
      "grad_norm": 6.40625,
      "learning_rate": 1.9926862905126663e-06,
      "loss": 0.6163,
      "step": 2040
    },
    {
      "epoch": 0.6756478391628559,
      "grad_norm": 4.5,
      "learning_rate": 1.9098300562505266e-06,
      "loss": 0.5507,
      "step": 2050
    },
    {
      "epoch": 0.6789436822807234,
      "grad_norm": 2.796875,
      "learning_rate": 1.8285510166487154e-06,
      "loss": 0.572,
      "step": 2060
    },
    {
      "epoch": 0.682239525398591,
      "grad_norm": 3.15625,
      "learning_rate": 1.7488650172170496e-06,
      "loss": 0.6,
      "step": 2070
    },
    {
      "epoch": 0.6855353685164586,
      "grad_norm": 2.40625,
      "learning_rate": 1.6707875928990059e-06,
      "loss": 0.5752,
      "step": 2080
    },
    {
      "epoch": 0.6888312116343263,
      "grad_norm": 4.375,
      "learning_rate": 1.5943339650431578e-06,
      "loss": 0.623,
      "step": 2090
    },
    {
      "epoch": 0.6921270547521938,
      "grad_norm": 3.609375,
      "learning_rate": 1.5195190384357405e-06,
      "loss": 0.5989,
      "step": 2100
    },
    {
      "epoch": 0.6954228978700614,
      "grad_norm": 2.609375,
      "learning_rate": 1.446357398394934e-06,
      "loss": 0.5616,
      "step": 2110
    },
    {
      "epoch": 0.698718740987929,
      "grad_norm": 6.5625,
      "learning_rate": 1.3748633079274254e-06,
      "loss": 0.6017,
      "step": 2120
    },
    {
      "epoch": 0.7020145841057965,
      "grad_norm": 3.328125,
      "learning_rate": 1.30505070494781e-06,
      "loss": 0.5848,
      "step": 2130
    },
    {
      "epoch": 0.7053104272236641,
      "grad_norm": 4.84375,
      "learning_rate": 1.2369331995613664e-06,
      "loss": 0.6331,
      "step": 2140
    },
    {
      "epoch": 0.7086062703415318,
      "grad_norm": 6.125,
      "learning_rate": 1.1705240714107301e-06,
      "loss": 0.5609,
      "step": 2150
    },
    {
      "epoch": 0.7119021134593994,
      "grad_norm": 4.34375,
      "learning_rate": 1.1058362670870248e-06,
      "loss": 0.5933,
      "step": 2160
    },
    {
      "epoch": 0.7151979565772669,
      "grad_norm": 3.421875,
      "learning_rate": 1.042882397605871e-06,
      "loss": 0.5997,
      "step": 2170
    },
    {
      "epoch": 0.7184937996951345,
      "grad_norm": 5.21875,
      "learning_rate": 9.816747359488632e-07,
      "loss": 0.568,
      "step": 2180
    },
    {
      "epoch": 0.7217896428130021,
      "grad_norm": 3.359375,
      "learning_rate": 9.222252146709143e-07,
      "loss": 0.5857,
      "step": 2190
    },
    {
      "epoch": 0.7250854859308697,
      "grad_norm": 3.875,
      "learning_rate": 8.645454235739903e-07,
      "loss": 0.6007,
      "step": 2200
    },
    {
      "epoch": 0.7283813290487373,
      "grad_norm": 6.09375,
      "learning_rate": 8.086466074476562e-07,
      "loss": 0.6058,
      "step": 2210
    },
    {
      "epoch": 0.7316771721666049,
      "grad_norm": 3.90625,
      "learning_rate": 7.545396638768698e-07,
      "loss": 0.5955,
      "step": 2220
    },
    {
      "epoch": 0.7349730152844725,
      "grad_norm": 3.640625,
      "learning_rate": 7.022351411174866e-07,
      "loss": 0.6081,
      "step": 2230
    },
    {
      "epoch": 0.73826885840234,
      "grad_norm": 4.375,
      "learning_rate": 6.517432360398556e-07,
      "loss": 0.5757,
      "step": 2240
    },
    {
      "epoch": 0.7415647015202076,
      "grad_norm": 4.03125,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.5924,
      "step": 2250
    },
    {
      "epoch": 0.7448605446380753,
      "grad_norm": 4.09375,
      "learning_rate": 5.562362976251901e-07,
      "loss": 0.6272,
      "step": 2260
    },
    {
      "epoch": 0.7481563877559428,
      "grad_norm": 5.15625,
      "learning_rate": 5.112398835550348e-07,
      "loss": 0.5836,
      "step": 2270
    },
    {
      "epoch": 0.7514522308738104,
      "grad_norm": 3.609375,
      "learning_rate": 4.6809332207053083e-07,
      "loss": 0.5314,
      "step": 2280
    },
    {
      "epoch": 0.754748073991678,
      "grad_norm": 4.75,
      "learning_rate": 4.268050246793276e-07,
      "loss": 0.6496,
      "step": 2290
    },
    {
      "epoch": 0.7580439171095456,
      "grad_norm": 5.25,
      "learning_rate": 3.8738304061681107e-07,
      "loss": 0.6164,
      "step": 2300
    },
    {
      "epoch": 0.7613397602274131,
      "grad_norm": 3.453125,
      "learning_rate": 3.498350552768859e-07,
      "loss": 0.5891,
      "step": 2310
    },
    {
      "epoch": 0.7646356033452808,
      "grad_norm": 4.4375,
      "learning_rate": 3.1416838871368925e-07,
      "loss": 0.5948,
      "step": 2320
    },
    {
      "epoch": 0.7679314464631484,
      "grad_norm": 3.671875,
      "learning_rate": 2.8038999421453827e-07,
      "loss": 0.5806,
      "step": 2330
    },
    {
      "epoch": 0.7712272895810159,
      "grad_norm": 3.390625,
      "learning_rate": 2.4850645694436736e-07,
      "loss": 0.5701,
      "step": 2340
    },
    {
      "epoch": 0.7745231326988835,
      "grad_norm": 2.84375,
      "learning_rate": 2.1852399266194312e-07,
      "loss": 0.5605,
      "step": 2350
    },
    {
      "epoch": 0.7778189758167511,
      "grad_norm": 2.296875,
      "learning_rate": 1.9044844650808468e-07,
      "loss": 0.569,
      "step": 2360
    },
    {
      "epoch": 0.7811148189346188,
      "grad_norm": 6.125,
      "learning_rate": 1.6428529186614195e-07,
      "loss": 0.5978,
      "step": 2370
    },
    {
      "epoch": 0.7844106620524863,
      "grad_norm": 2.734375,
      "learning_rate": 1.400396292949513e-07,
      "loss": 0.5669,
      "step": 2380
    },
    {
      "epoch": 0.7877065051703539,
      "grad_norm": 3.4375,
      "learning_rate": 1.1771618553447217e-07,
      "loss": 0.5828,
      "step": 2390
    },
    {
      "epoch": 0.7910023482882215,
      "grad_norm": 3.015625,
      "learning_rate": 9.731931258429638e-08,
      "loss": 0.5724,
      "step": 2400
    },
    {
      "epoch": 0.794298191406089,
      "grad_norm": 3.734375,
      "learning_rate": 7.885298685522235e-08,
      "loss": 0.5522,
      "step": 2410
    },
    {
      "epoch": 0.7975940345239566,
      "grad_norm": 2.296875,
      "learning_rate": 6.232080839403631e-08,
      "loss": 0.5494,
      "step": 2420
    },
    {
      "epoch": 0.8008898776418243,
      "grad_norm": 4.3125,
      "learning_rate": 4.772600018168816e-08,
      "loss": 0.6146,
      "step": 2430
    },
    {
      "epoch": 0.8041857207596919,
      "grad_norm": 3.578125,
      "learning_rate": 3.50714075049563e-08,
      "loss": 0.5484,
      "step": 2440
    },
    {
      "epoch": 0.8074815638775594,
      "grad_norm": 5.28125,
      "learning_rate": 2.4359497401758026e-08,
      "loss": 0.5862,
      "step": 2450
    },
    {
      "epoch": 0.810777406995427,
      "grad_norm": 5.09375,
      "learning_rate": 1.5592358180189782e-08,
      "loss": 0.6255,
      "step": 2460
    },
    {
      "epoch": 0.8140732501132946,
      "grad_norm": 3.015625,
      "learning_rate": 8.771699011416169e-09,
      "loss": 0.5877,
      "step": 2470
    },
    {
      "epoch": 0.8173690932311622,
      "grad_norm": 2.75,
      "learning_rate": 3.898849596456477e-09,
      "loss": 0.5634,
      "step": 2480
    },
    {
      "epoch": 0.8206649363490298,
      "grad_norm": 2.6875,
      "learning_rate": 9.74759906957612e-10,
      "loss": 0.5582,
      "step": 2490
    },
    {
      "epoch": 0.8239607794668974,
      "grad_norm": 3.65625,
      "learning_rate": 0.0,
      "loss": 0.5774,
      "step": 2500
    }
  ],
  "logging_steps": 10,
  "max_steps": 2500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.5664155394048e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}