{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.872885979268958,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00436442989634479,
      "grad_norm": 74.5,
      "learning_rate": 8.000000000000001e-07,
      "loss": 1.044,
      "step": 10
    },
    {
      "epoch": 0.00872885979268958,
      "grad_norm": 166.0,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 1.0152,
      "step": 20
    },
    {
      "epoch": 0.01309328968903437,
      "grad_norm": 249.0,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 1.0905,
      "step": 30
    },
    {
      "epoch": 0.01745771958537916,
      "grad_norm": 210.0,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 1.1291,
      "step": 40
    },
    {
      "epoch": 0.02182214948172395,
      "grad_norm": 151.0,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.384,
      "step": 50
    },
    {
      "epoch": 0.02618657937806874,
      "grad_norm": 516.0,
      "learning_rate": 4.800000000000001e-06,
      "loss": 1.5235,
      "step": 60
    },
    {
      "epoch": 0.03055100927441353,
      "grad_norm": 684.0,
      "learning_rate": 5.600000000000001e-06,
      "loss": 1.888,
      "step": 70
    },
    {
      "epoch": 0.03491543917075832,
      "grad_norm": 504.0,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 1.9856,
      "step": 80
    },
    {
      "epoch": 0.03927986906710311,
      "grad_norm": 175.0,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 1.628,
      "step": 90
    },
    {
      "epoch": 0.0436442989634479,
      "grad_norm": 964.0,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.8794,
      "step": 100
    },
    {
      "epoch": 0.04800872885979269,
      "grad_norm": 151.0,
      "learning_rate": 8.8e-06,
      "loss": 1.4958,
      "step": 110
    },
    {
      "epoch": 0.05237315875613748,
      "grad_norm": 45.25,
      "learning_rate": 9.600000000000001e-06,
      "loss": 1.741,
      "step": 120
    },
    {
      "epoch": 0.05673758865248227,
      "grad_norm": 50.5,
      "learning_rate": 1.04e-05,
      "loss": 1.5346,
      "step": 130
    },
    {
      "epoch": 0.06110201854882706,
      "grad_norm": 116.5,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 1.498,
      "step": 140
    },
    {
      "epoch": 0.06546644844517185,
      "grad_norm": 35.0,
      "learning_rate": 1.2e-05,
      "loss": 1.3559,
      "step": 150
    },
    {
      "epoch": 0.06983087834151663,
      "grad_norm": 57.5,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 1.3762,
      "step": 160
    },
    {
      "epoch": 0.07419530823786143,
      "grad_norm": 40.25,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 1.5166,
      "step": 170
    },
    {
      "epoch": 0.07855973813420622,
      "grad_norm": 22.5,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 1.3441,
      "step": 180
    },
    {
      "epoch": 0.08292416803055101,
      "grad_norm": 14.875,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 1.6288,
      "step": 190
    },
    {
      "epoch": 0.0872885979268958,
      "grad_norm": 203.0,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 3.1069,
      "step": 200
    },
    {
      "epoch": 0.09165302782324058,
      "grad_norm": 57.5,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 1.5319,
      "step": 210
    },
    {
      "epoch": 0.09601745771958538,
      "grad_norm": 26.0,
      "learning_rate": 1.76e-05,
      "loss": 1.5801,
      "step": 220
    },
    {
      "epoch": 0.10038188761593017,
      "grad_norm": 24.125,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 1.3185,
      "step": 230
    },
    {
      "epoch": 0.10474631751227496,
      "grad_norm": 29.875,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 1.3118,
      "step": 240
    },
    {
      "epoch": 0.10911074740861974,
      "grad_norm": 18.625,
      "learning_rate": 2e-05,
      "loss": 1.1643,
      "step": 250
    },
    {
      "epoch": 0.11347517730496454,
      "grad_norm": 25.5,
      "learning_rate": 1.9999025240093045e-05,
      "loss": 1.3403,
      "step": 260
    },
    {
      "epoch": 0.11783960720130933,
      "grad_norm": 20.75,
      "learning_rate": 1.9996101150403543e-05,
      "loss": 1.2914,
      "step": 270
    },
    {
      "epoch": 0.12220403709765412,
      "grad_norm": 17.625,
      "learning_rate": 1.9991228300988586e-05,
      "loss": 1.2328,
      "step": 280
    },
    {
      "epoch": 0.1265684669939989,
      "grad_norm": 37.25,
      "learning_rate": 1.9984407641819812e-05,
      "loss": 1.2335,
      "step": 290
    },
    {
      "epoch": 0.1309328968903437,
      "grad_norm": 18.75,
      "learning_rate": 1.9975640502598243e-05,
      "loss": 1.1872,
      "step": 300
    },
    {
      "epoch": 0.1352973267866885,
      "grad_norm": 22.625,
      "learning_rate": 1.9964928592495046e-05,
      "loss": 1.1274,
      "step": 310
    },
    {
      "epoch": 0.13966175668303327,
      "grad_norm": 15.8125,
      "learning_rate": 1.9952273999818312e-05,
      "loss": 1.235,
      "step": 320
    },
    {
      "epoch": 0.14402618657937807,
      "grad_norm": 16.5,
      "learning_rate": 1.9937679191605964e-05,
      "loss": 1.1159,
      "step": 330
    },
    {
      "epoch": 0.14839061647572285,
      "grad_norm": 17.5,
      "learning_rate": 1.9921147013144782e-05,
      "loss": 1.0252,
      "step": 340
    },
    {
      "epoch": 0.15275504637206766,
      "grad_norm": 6.78125,
      "learning_rate": 1.9902680687415704e-05,
      "loss": 1.0231,
      "step": 350
    },
    {
      "epoch": 0.15711947626841244,
      "grad_norm": 20.375,
      "learning_rate": 1.988228381446553e-05,
      "loss": 1.086,
      "step": 360
    },
    {
      "epoch": 0.16148390616475722,
      "grad_norm": 29.375,
      "learning_rate": 1.985996037070505e-05,
      "loss": 1.0803,
      "step": 370
    },
    {
      "epoch": 0.16584833606110203,
      "grad_norm": 10.9375,
      "learning_rate": 1.983571470813386e-05,
      "loss": 0.987,
      "step": 380
    },
    {
      "epoch": 0.1702127659574468,
      "grad_norm": 12.625,
      "learning_rate": 1.9809551553491918e-05,
      "loss": 1.2151,
      "step": 390
    },
    {
      "epoch": 0.1745771958537916,
      "grad_norm": 14.125,
      "learning_rate": 1.9781476007338058e-05,
      "loss": 0.9436,
      "step": 400
    },
    {
      "epoch": 0.1789416257501364,
      "grad_norm": 56.0,
      "learning_rate": 1.9751493543055634e-05,
      "loss": 1.0427,
      "step": 410
    },
    {
      "epoch": 0.18330605564648117,
      "grad_norm": 17.5,
      "learning_rate": 1.9719610005785466e-05,
      "loss": 1.0463,
      "step": 420
    },
    {
      "epoch": 0.18767048554282598,
      "grad_norm": 14.4375,
      "learning_rate": 1.9685831611286312e-05,
      "loss": 1.0375,
      "step": 430
    },
    {
      "epoch": 0.19203491543917076,
      "grad_norm": 14.3125,
      "learning_rate": 1.9650164944723116e-05,
      "loss": 1.0024,
      "step": 440
    },
    {
      "epoch": 0.19639934533551553,
      "grad_norm": 10.75,
      "learning_rate": 1.961261695938319e-05,
      "loss": 1.0858,
      "step": 450
    },
    {
      "epoch": 0.20076377523186034,
      "grad_norm": 8.1875,
      "learning_rate": 1.9573194975320672e-05,
      "loss": 0.9583,
      "step": 460
    },
    {
      "epoch": 0.20512820512820512,
      "grad_norm": 16.75,
      "learning_rate": 1.9531906677929472e-05,
      "loss": 1.048,
      "step": 470
    },
    {
      "epoch": 0.20949263502454993,
      "grad_norm": 24.25,
      "learning_rate": 1.9488760116444966e-05,
      "loss": 0.9226,
      "step": 480
    },
    {
      "epoch": 0.2138570649208947,
      "grad_norm": 12.125,
      "learning_rate": 1.944376370237481e-05,
      "loss": 1.0239,
      "step": 490
    },
    {
      "epoch": 0.2182214948172395,
      "grad_norm": 15.9375,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.9481,
      "step": 500
    },
    {
      "epoch": 0.2225859247135843,
      "grad_norm": 8.0,
      "learning_rate": 1.9348256763960146e-05,
      "loss": 0.9285,
      "step": 510
    },
    {
      "epoch": 0.22695035460992907,
      "grad_norm": 16.625,
      "learning_rate": 1.9297764858882516e-05,
      "loss": 0.9811,
      "step": 520
    },
    {
      "epoch": 0.23131478450627388,
      "grad_norm": 7.34375,
      "learning_rate": 1.9245460336123136e-05,
      "loss": 0.8843,
      "step": 530
    },
    {
      "epoch": 0.23567921440261866,
      "grad_norm": 12.5,
      "learning_rate": 1.9191353392552346e-05,
      "loss": 0.869,
      "step": 540
    },
    {
      "epoch": 0.24004364429896344,
      "grad_norm": 10.25,
      "learning_rate": 1.913545457642601e-05,
      "loss": 0.9146,
      "step": 550
    },
    {
      "epoch": 0.24440807419530824,
      "grad_norm": 20.75,
      "learning_rate": 1.907777478532909e-05,
      "loss": 0.904,
      "step": 560
    },
    {
      "epoch": 0.24877250409165302,
      "grad_norm": 13.9375,
      "learning_rate": 1.901832526405114e-05,
      "loss": 0.8561,
      "step": 570
    },
    {
      "epoch": 0.2531369339879978,
      "grad_norm": 9.25,
      "learning_rate": 1.895711760239413e-05,
      "loss": 0.9274,
      "step": 580
    },
    {
      "epoch": 0.2575013638843426,
      "grad_norm": 9.5625,
      "learning_rate": 1.889416373291298e-05,
      "loss": 0.9191,
      "step": 590
    },
    {
      "epoch": 0.2618657937806874,
      "grad_norm": 6.75,
      "learning_rate": 1.8829475928589272e-05,
      "loss": 0.9027,
      "step": 600
    },
    {
      "epoch": 0.2662302236770322,
      "grad_norm": 9.5,
      "learning_rate": 1.8763066800438638e-05,
      "loss": 0.9427,
      "step": 610
    },
    {
      "epoch": 0.270594653573377,
      "grad_norm": 13.125,
      "learning_rate": 1.869494929505219e-05,
      "loss": 0.8664,
      "step": 620
    },
    {
      "epoch": 0.27495908346972175,
      "grad_norm": 19.375,
      "learning_rate": 1.8625136692072577e-05,
      "loss": 0.9021,
      "step": 630
    },
    {
      "epoch": 0.27932351336606653,
      "grad_norm": 14.5,
      "learning_rate": 1.855364260160507e-05,
      "loss": 0.9041,
      "step": 640
    },
    {
      "epoch": 0.28368794326241137,
      "grad_norm": 8.75,
      "learning_rate": 1.848048096156426e-05,
      "loss": 0.8482,
      "step": 650
    },
    {
      "epoch": 0.28805237315875615,
      "grad_norm": 18.5,
      "learning_rate": 1.8405666034956842e-05,
      "loss": 0.9267,
      "step": 660
    },
    {
      "epoch": 0.2924168030551009,
      "grad_norm": 5.875,
      "learning_rate": 1.8329212407100996e-05,
      "loss": 0.8678,
      "step": 670
    },
    {
      "epoch": 0.2967812329514457,
      "grad_norm": 12.625,
      "learning_rate": 1.8251134982782952e-05,
      "loss": 0.8268,
      "step": 680
    },
    {
      "epoch": 0.3011456628477905,
      "grad_norm": 12.875,
      "learning_rate": 1.8171448983351284e-05,
      "loss": 0.8265,
      "step": 690
    },
    {
      "epoch": 0.3055100927441353,
      "grad_norm": 8.6875,
      "learning_rate": 1.8090169943749477e-05,
      "loss": 0.8773,
      "step": 700
    },
    {
      "epoch": 0.3098745226404801,
      "grad_norm": 16.75,
      "learning_rate": 1.8007313709487334e-05,
      "loss": 0.8472,
      "step": 710
    },
    {
      "epoch": 0.3142389525368249,
      "grad_norm": 7.53125,
      "learning_rate": 1.792289643355191e-05,
      "loss": 0.8534,
      "step": 720
    },
    {
      "epoch": 0.31860338243316966,
      "grad_norm": 15.125,
      "learning_rate": 1.78369345732584e-05,
      "loss": 0.8069,
      "step": 730
    },
    {
      "epoch": 0.32296781232951444,
      "grad_norm": 11.5,
      "learning_rate": 1.7749444887041797e-05,
      "loss": 0.8533,
      "step": 740
    },
    {
      "epoch": 0.32733224222585927,
      "grad_norm": 14.75,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.8662,
      "step": 750
    },
    {
      "epoch": 0.33169667212220405,
      "grad_norm": 15.75,
      "learning_rate": 1.7569950556517566e-05,
      "loss": 0.8397,
      "step": 760
    },
    {
      "epoch": 0.33606110201854883,
      "grad_norm": 19.875,
      "learning_rate": 1.747798090498532e-05,
      "loss": 0.7535,
      "step": 770
    },
    {
      "epoch": 0.3404255319148936,
      "grad_norm": 11.75,
      "learning_rate": 1.7384553406258842e-05,
      "loss": 0.8573,
      "step": 780
    },
    {
      "epoch": 0.3447899618112384,
      "grad_norm": 6.9375,
      "learning_rate": 1.7289686274214116e-05,
      "loss": 0.8559,
      "step": 790
    },
    {
      "epoch": 0.3491543917075832,
      "grad_norm": 10.9375,
      "learning_rate": 1.7193398003386514e-05,
      "loss": 0.8459,
      "step": 800
    },
    {
      "epoch": 0.353518821603928,
      "grad_norm": 12.0625,
      "learning_rate": 1.709570736536521e-05,
      "loss": 0.8352,
      "step": 810
    },
    {
      "epoch": 0.3578832515002728,
      "grad_norm": 4.59375,
      "learning_rate": 1.6996633405133656e-05,
      "loss": 0.8146,
      "step": 820
    },
    {
      "epoch": 0.36224768139661756,
      "grad_norm": 15.0,
      "learning_rate": 1.68961954373567e-05,
      "loss": 0.78,
      "step": 830
    },
    {
      "epoch": 0.36661211129296234,
      "grad_norm": 5.6875,
      "learning_rate": 1.6794413042615168e-05,
      "loss": 0.8002,
      "step": 840
    },
    {
      "epoch": 0.3709765411893072,
      "grad_norm": 18.5,
      "learning_rate": 1.6691306063588583e-05,
      "loss": 0.8296,
      "step": 850
    },
    {
      "epoch": 0.37534097108565195,
      "grad_norm": 42.0,
      "learning_rate": 1.6586894601186804e-05,
      "loss": 0.8742,
      "step": 860
    },
    {
      "epoch": 0.37970540098199673,
      "grad_norm": 9.5,
      "learning_rate": 1.6481199010631312e-05,
      "loss": 0.779,
      "step": 870
    },
    {
      "epoch": 0.3840698308783415,
      "grad_norm": 6.0625,
      "learning_rate": 1.63742398974869e-05,
      "loss": 0.747,
      "step": 880
    },
    {
      "epoch": 0.3884342607746863,
      "grad_norm": 9.375,
      "learning_rate": 1.6266038113644605e-05,
      "loss": 0.8371,
      "step": 890
    },
    {
      "epoch": 0.39279869067103107,
      "grad_norm": 7.9375,
      "learning_rate": 1.6156614753256583e-05,
      "loss": 0.7598,
      "step": 900
    },
    {
      "epoch": 0.3971631205673759,
      "grad_norm": 14.75,
      "learning_rate": 1.6045991148623752e-05,
      "loss": 0.7371,
      "step": 910
    },
    {
      "epoch": 0.4015275504637207,
      "grad_norm": 5.09375,
      "learning_rate": 1.5934188866037017e-05,
      "loss": 0.7388,
      "step": 920
    },
    {
      "epoch": 0.40589198036006546,
      "grad_norm": 7.96875,
      "learning_rate": 1.5821229701572897e-05,
      "loss": 0.7652,
      "step": 930
    },
    {
      "epoch": 0.41025641025641024,
      "grad_norm": 12.3125,
      "learning_rate": 1.570713567684432e-05,
      "loss": 0.82,
      "step": 940
    },
    {
      "epoch": 0.414620840152755,
      "grad_norm": 9.5,
      "learning_rate": 1.5591929034707468e-05,
      "loss": 0.7934,
      "step": 950
    },
    {
      "epoch": 0.41898527004909986,
      "grad_norm": 8.5625,
      "learning_rate": 1.5475632234925505e-05,
      "loss": 0.7752,
      "step": 960
    },
    {
      "epoch": 0.42334969994544464,
      "grad_norm": 8.375,
      "learning_rate": 1.5358267949789968e-05,
      "loss": 0.7438,
      "step": 970
    },
    {
      "epoch": 0.4277141298417894,
      "grad_norm": 7.6875,
      "learning_rate": 1.5239859059700794e-05,
      "loss": 0.7512,
      "step": 980
    },
    {
      "epoch": 0.4320785597381342,
      "grad_norm": 9.1875,
      "learning_rate": 1.5120428648705716e-05,
      "loss": 0.6996,
      "step": 990
    },
    {
      "epoch": 0.436442989634479,
      "grad_norm": 6.0625,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7038,
      "step": 1000
    },
    {
      "epoch": 0.4408074195308238,
      "grad_norm": 8.9375,
      "learning_rate": 1.4878596591387329e-05,
      "loss": 0.7495,
      "step": 1010
    },
    {
      "epoch": 0.4451718494271686,
      "grad_norm": 6.59375,
      "learning_rate": 1.4756242090702756e-05,
      "loss": 0.7794,
      "step": 1020
    },
    {
      "epoch": 0.44953627932351337,
      "grad_norm": 10.0625,
      "learning_rate": 1.463296035119862e-05,
      "loss": 0.747,
      "step": 1030
    },
    {
      "epoch": 0.45390070921985815,
      "grad_norm": 11.4375,
      "learning_rate": 1.4508775406894308e-05,
      "loss": 0.7394,
      "step": 1040
    },
    {
      "epoch": 0.4582651391162029,
      "grad_norm": 8.375,
      "learning_rate": 1.4383711467890776e-05,
      "loss": 0.6999,
      "step": 1050
    },
    {
      "epoch": 0.46262956901254776,
      "grad_norm": 11.9375,
      "learning_rate": 1.4257792915650728e-05,
      "loss": 0.7475,
      "step": 1060
    },
    {
      "epoch": 0.46699399890889254,
      "grad_norm": 4.125,
      "learning_rate": 1.413104429824542e-05,
      "loss": 0.6836,
      "step": 1070
    },
    {
      "epoch": 0.4713584288052373,
      "grad_norm": 6.5,
      "learning_rate": 1.4003490325568953e-05,
      "loss": 0.6974,
      "step": 1080
    },
    {
      "epoch": 0.4757228587015821,
      "grad_norm": 8.375,
      "learning_rate": 1.3875155864521031e-05,
      "loss": 0.7776,
      "step": 1090
    },
    {
      "epoch": 0.4800872885979269,
      "grad_norm": 11.125,
      "learning_rate": 1.3746065934159123e-05,
      "loss": 0.7633,
      "step": 1100
    },
    {
      "epoch": 0.4844517184942717,
      "grad_norm": 6.09375,
      "learning_rate": 1.3616245700820922e-05,
      "loss": 0.7146,
      "step": 1110
    },
    {
      "epoch": 0.4888161483906165,
      "grad_norm": 54.25,
      "learning_rate": 1.3485720473218153e-05,
      "loss": 0.7503,
      "step": 1120
    },
    {
      "epoch": 0.49318057828696127,
      "grad_norm": 9.0,
      "learning_rate": 1.3354515697502552e-05,
      "loss": 0.6919,
      "step": 1130
    },
    {
      "epoch": 0.49754500818330605,
      "grad_norm": 5.65625,
      "learning_rate": 1.3222656952305113e-05,
      "loss": 0.702,
      "step": 1140
    },
    {
      "epoch": 0.5019094380796508,
      "grad_norm": 8.125,
      "learning_rate": 1.3090169943749475e-05,
      "loss": 0.7013,
      "step": 1150
    },
    {
      "epoch": 0.5062738679759956,
      "grad_norm": 7.125,
      "learning_rate": 1.2957080500440469e-05,
      "loss": 0.6457,
      "step": 1160
    },
    {
      "epoch": 0.5106382978723404,
      "grad_norm": 3.59375,
      "learning_rate": 1.2823414568428767e-05,
      "loss": 0.7375,
      "step": 1170
    },
    {
      "epoch": 0.5150027277686852,
      "grad_norm": 7.53125,
      "learning_rate": 1.2689198206152657e-05,
      "loss": 0.6741,
      "step": 1180
    },
    {
      "epoch": 0.51936715766503,
      "grad_norm": 5.5625,
      "learning_rate": 1.2554457579357906e-05,
      "loss": 0.6753,
      "step": 1190
    },
    {
      "epoch": 0.5237315875613748,
      "grad_norm": 15.5625,
      "learning_rate": 1.2419218955996677e-05,
      "loss": 0.715,
      "step": 1200
    },
    {
      "epoch": 0.5280960174577196,
      "grad_norm": 15.4375,
      "learning_rate": 1.2283508701106559e-05,
      "loss": 0.6916,
      "step": 1210
    },
    {
      "epoch": 0.5324604473540644,
      "grad_norm": 5.90625,
      "learning_rate": 1.2147353271670634e-05,
      "loss": 0.7213,
      "step": 1220
    },
    {
      "epoch": 0.5368248772504092,
      "grad_norm": 3.3125,
      "learning_rate": 1.2010779211459649e-05,
      "loss": 0.6672,
      "step": 1230
    },
    {
      "epoch": 0.541189307146754,
      "grad_norm": 4.46875,
      "learning_rate": 1.187381314585725e-05,
      "loss": 0.6987,
      "step": 1240
    },
    {
      "epoch": 0.5455537370430987,
      "grad_norm": 9.3125,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.6964,
      "step": 1250
    },
    {
      "epoch": 0.5499181669394435,
      "grad_norm": 11.625,
      "learning_rate": 1.159881187691835e-05,
      "loss": 0.689,
      "step": 1260
    },
    {
      "epoch": 0.5542825968357883,
      "grad_norm": 8.4375,
      "learning_rate": 1.1460830285624119e-05,
      "loss": 0.662,
      "step": 1270
    },
    {
      "epoch": 0.5586470267321331,
      "grad_norm": 12.875,
      "learning_rate": 1.1322563902571227e-05,
      "loss": 0.6413,
      "step": 1280
    },
    {
      "epoch": 0.563011456628478,
      "grad_norm": 5.78125,
      "learning_rate": 1.1184039683065014e-05,
      "loss": 0.6609,
      "step": 1290
    },
    {
      "epoch": 0.5673758865248227,
      "grad_norm": 6.0,
      "learning_rate": 1.1045284632676535e-05,
      "loss": 0.6716,
      "step": 1300
    },
    {
      "epoch": 0.5717403164211675,
      "grad_norm": 9.0,
      "learning_rate": 1.0906325801977804e-05,
      "loss": 0.692,
      "step": 1310
    },
    {
      "epoch": 0.5761047463175123,
      "grad_norm": 4.1875,
      "learning_rate": 1.0767190281268187e-05,
      "loss": 0.5993,
      "step": 1320
    },
    {
      "epoch": 0.5804691762138571,
      "grad_norm": 3.953125,
      "learning_rate": 1.0627905195293135e-05,
      "loss": 0.6846,
      "step": 1330
    },
    {
      "epoch": 0.5848336061102019,
      "grad_norm": 5.65625,
      "learning_rate": 1.0488497697956134e-05,
      "loss": 0.6426,
      "step": 1340
    },
    {
      "epoch": 0.5891980360065466,
      "grad_norm": 3.796875,
      "learning_rate": 1.0348994967025012e-05,
      "loss": 0.7003,
      "step": 1350
    },
    {
      "epoch": 0.5935624659028914,
      "grad_norm": 4.53125,
      "learning_rate": 1.0209424198833571e-05,
      "loss": 0.6536,
      "step": 1360
    },
    {
      "epoch": 0.5979268957992362,
      "grad_norm": 6.0625,
      "learning_rate": 1.0069812602979617e-05,
      "loss": 0.6849,
      "step": 1370
    },
    {
      "epoch": 0.602291325695581,
      "grad_norm": 4.34375,
      "learning_rate": 9.930187397020385e-06,
      "loss": 0.672,
      "step": 1380
    },
    {
      "epoch": 0.6066557555919257,
      "grad_norm": 7.84375,
      "learning_rate": 9.790575801166432e-06,
      "loss": 0.715,
      "step": 1390
    },
    {
      "epoch": 0.6110201854882706,
      "grad_norm": 3.078125,
      "learning_rate": 9.651005032974994e-06,
      "loss": 0.7157,
      "step": 1400
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 9.1875,
      "learning_rate": 9.511502302043867e-06,
      "loss": 0.6274,
      "step": 1410
    },
    {
      "epoch": 0.6197490452809602,
      "grad_norm": 3.84375,
      "learning_rate": 9.372094804706867e-06,
      "loss": 0.6586,
      "step": 1420
    },
    {
      "epoch": 0.624113475177305,
      "grad_norm": 4.15625,
      "learning_rate": 9.232809718731815e-06,
      "loss": 0.623,
      "step": 1430
    },
    {
      "epoch": 0.6284779050736498,
      "grad_norm": 4.4375,
      "learning_rate": 9.093674198022201e-06,
      "loss": 0.6606,
      "step": 1440
    },
    {
      "epoch": 0.6328423349699945,
      "grad_norm": 8.25,
      "learning_rate": 8.954715367323468e-06,
      "loss": 0.6573,
      "step": 1450
    },
    {
      "epoch": 0.6372067648663393,
      "grad_norm": 3.671875,
      "learning_rate": 8.815960316934991e-06,
      "loss": 0.6806,
      "step": 1460
    },
    {
      "epoch": 0.6415711947626841,
      "grad_norm": 9.6875,
      "learning_rate": 8.677436097428775e-06,
      "loss": 0.6407,
      "step": 1470
    },
    {
      "epoch": 0.6459356246590289,
      "grad_norm": 8.5,
      "learning_rate": 8.539169714375885e-06,
      "loss": 0.7079,
      "step": 1480
    },
    {
      "epoch": 0.6503000545553737,
      "grad_norm": 4.3125,
      "learning_rate": 8.401188123081653e-06,
      "loss": 0.6743,
      "step": 1490
    },
    {
      "epoch": 0.6546644844517185,
      "grad_norm": 5.28125,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.5891,
      "step": 1500
    },
    {
      "epoch": 0.6590289143480633,
      "grad_norm": 2.890625,
      "learning_rate": 8.126186854142752e-06,
      "loss": 0.6864,
      "step": 1510
    },
    {
      "epoch": 0.6633933442444081,
      "grad_norm": 4.46875,
      "learning_rate": 7.989220788540356e-06,
      "loss": 0.6659,
      "step": 1520
    },
    {
      "epoch": 0.6677577741407529,
      "grad_norm": 3.78125,
      "learning_rate": 7.852646728329368e-06,
      "loss": 0.6588,
      "step": 1530
    },
    {
      "epoch": 0.6721222040370977,
      "grad_norm": 2.953125,
      "learning_rate": 7.716491298893443e-06,
      "loss": 0.6609,
      "step": 1540
    },
    {
      "epoch": 0.6764866339334424,
      "grad_norm": 4.375,
      "learning_rate": 7.580781044003324e-06,
      "loss": 0.6429,
      "step": 1550
    },
    {
      "epoch": 0.6808510638297872,
      "grad_norm": 6.53125,
      "learning_rate": 7.445542420642097e-06,
      "loss": 0.653,
      "step": 1560
    },
    {
      "epoch": 0.685215493726132,
      "grad_norm": 4.65625,
      "learning_rate": 7.310801793847344e-06,
      "loss": 0.6525,
      "step": 1570
    },
    {
      "epoch": 0.6895799236224768,
      "grad_norm": 3.828125,
      "learning_rate": 7.176585431571235e-06,
      "loss": 0.634,
      "step": 1580
    },
    {
      "epoch": 0.6939443535188216,
      "grad_norm": 3.53125,
      "learning_rate": 7.042919499559538e-06,
      "loss": 0.6536,
      "step": 1590
    },
    {
      "epoch": 0.6983087834151664,
      "grad_norm": 3.0,
      "learning_rate": 6.909830056250527e-06,
      "loss": 0.6663,
      "step": 1600
    },
    {
      "epoch": 0.7026732133115112,
      "grad_norm": 2.765625,
      "learning_rate": 6.777343047694891e-06,
      "loss": 0.6167,
      "step": 1610
    },
    {
      "epoch": 0.707037643207856,
      "grad_norm": 6.53125,
      "learning_rate": 6.645484302497452e-06,
      "loss": 0.5728,
      "step": 1620
    },
    {
      "epoch": 0.7114020731042008,
      "grad_norm": 3.609375,
      "learning_rate": 6.5142795267818505e-06,
      "loss": 0.6451,
      "step": 1630
    },
    {
      "epoch": 0.7157665030005456,
      "grad_norm": 2.6875,
      "learning_rate": 6.383754299179079e-06,
      "loss": 0.6403,
      "step": 1640
    },
    {
      "epoch": 0.7201309328968903,
      "grad_norm": 3.90625,
      "learning_rate": 6.25393406584088e-06,
      "loss": 0.6067,
      "step": 1650
    },
    {
      "epoch": 0.7244953627932351,
      "grad_norm": 3.375,
      "learning_rate": 6.124844135478971e-06,
      "loss": 0.6498,
      "step": 1660
    },
    {
      "epoch": 0.7288597926895799,
      "grad_norm": 9.75,
      "learning_rate": 5.996509674431053e-06,
      "loss": 0.6422,
      "step": 1670
    },
    {
      "epoch": 0.7332242225859247,
      "grad_norm": 4.09375,
      "learning_rate": 5.868955701754584e-06,
      "loss": 0.641,
      "step": 1680
    },
    {
      "epoch": 0.7375886524822695,
      "grad_norm": 3.4375,
      "learning_rate": 5.742207084349274e-06,
      "loss": 0.574,
      "step": 1690
    },
    {
      "epoch": 0.7419530823786143,
      "grad_norm": 3.03125,
      "learning_rate": 5.616288532109225e-06,
      "loss": 0.6205,
      "step": 1700
    },
    {
      "epoch": 0.7463175122749591,
      "grad_norm": 3.78125,
      "learning_rate": 5.491224593105695e-06,
      "loss": 0.6052,
      "step": 1710
    },
    {
      "epoch": 0.7506819421713039,
      "grad_norm": 4.25,
      "learning_rate": 5.367039648801386e-06,
      "loss": 0.6371,
      "step": 1720
    },
    {
      "epoch": 0.7550463720676487,
      "grad_norm": 5.875,
      "learning_rate": 5.243757909297247e-06,
      "loss": 0.6435,
      "step": 1730
    },
    {
      "epoch": 0.7594108019639935,
      "grad_norm": 6.15625,
      "learning_rate": 5.121403408612672e-06,
      "loss": 0.6427,
      "step": 1740
    },
    {
      "epoch": 0.7637752318603382,
      "grad_norm": 3.5625,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.6426,
      "step": 1750
    },
    {
      "epoch": 0.768139661756683,
      "grad_norm": 3.703125,
      "learning_rate": 4.879571351294287e-06,
      "loss": 0.6327,
      "step": 1760
    },
    {
      "epoch": 0.7725040916530278,
      "grad_norm": 3.828125,
      "learning_rate": 4.76014094029921e-06,
      "loss": 0.6088,
      "step": 1770
    },
    {
      "epoch": 0.7768685215493726,
      "grad_norm": 3.046875,
      "learning_rate": 4.641732050210032e-06,
      "loss": 0.6302,
      "step": 1780
    },
    {
      "epoch": 0.7812329514457174,
      "grad_norm": 3.09375,
      "learning_rate": 4.524367765074499e-06,
      "loss": 0.6669,
      "step": 1790
    },
    {
      "epoch": 0.7855973813420621,
      "grad_norm": 4.59375,
      "learning_rate": 4.408070965292534e-06,
      "loss": 0.663,
      "step": 1800
    },
    {
      "epoch": 0.789961811238407,
      "grad_norm": 2.890625,
      "learning_rate": 4.292864323155684e-06,
      "loss": 0.6331,
      "step": 1810
    },
    {
      "epoch": 0.7943262411347518,
      "grad_norm": 2.796875,
      "learning_rate": 4.178770298427107e-06,
      "loss": 0.6331,
      "step": 1820
    },
    {
      "epoch": 0.7986906710310966,
      "grad_norm": 2.796875,
      "learning_rate": 4.065811133962987e-06,
      "loss": 0.6317,
      "step": 1830
    },
    {
      "epoch": 0.8030551009274414,
      "grad_norm": 2.59375,
      "learning_rate": 3.954008851376252e-06,
      "loss": 0.6493,
      "step": 1840
    },
    {
      "epoch": 0.8074195308237861,
      "grad_norm": 4.1875,
      "learning_rate": 3.8433852467434175e-06,
      "loss": 0.5514,
      "step": 1850
    },
    {
      "epoch": 0.8117839607201309,
      "grad_norm": 3.203125,
      "learning_rate": 3.7339618863553983e-06,
      "loss": 0.5782,
      "step": 1860
    },
    {
      "epoch": 0.8161483906164757,
      "grad_norm": 2.84375,
      "learning_rate": 3.625760102513103e-06,
      "loss": 0.6039,
      "step": 1870
    },
    {
      "epoch": 0.8205128205128205,
      "grad_norm": 5.6875,
      "learning_rate": 3.5188009893686916e-06,
      "loss": 0.6285,
      "step": 1880
    },
    {
      "epoch": 0.8248772504091653,
      "grad_norm": 3.890625,
      "learning_rate": 3.4131053988131947e-06,
      "loss": 0.6174,
      "step": 1890
    },
    {
      "epoch": 0.82924168030551,
      "grad_norm": 2.78125,
      "learning_rate": 3.308693936411421e-06,
      "loss": 0.5981,
      "step": 1900
    },
    {
      "epoch": 0.8336061102018549,
      "grad_norm": 3.46875,
      "learning_rate": 3.2055869573848374e-06,
      "loss": 0.623,
      "step": 1910
    },
    {
      "epoch": 0.8379705400981997,
      "grad_norm": 2.609375,
      "learning_rate": 3.103804562643302e-06,
      "loss": 0.5816,
      "step": 1920
    },
    {
      "epoch": 0.8423349699945445,
      "grad_norm": 2.1875,
      "learning_rate": 3.003366594866345e-06,
      "loss": 0.6548,
      "step": 1930
    },
    {
      "epoch": 0.8466993998908893,
      "grad_norm": 2.296875,
      "learning_rate": 2.9042926346347932e-06,
      "loss": 0.5988,
      "step": 1940
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 2.578125,
      "learning_rate": 2.8066019966134907e-06,
      "loss": 0.6785,
      "step": 1950
    },
    {
      "epoch": 0.8554282596835788,
      "grad_norm": 2.78125,
      "learning_rate": 2.7103137257858867e-06,
      "loss": 0.5518,
      "step": 1960
    },
    {
      "epoch": 0.8597926895799236,
      "grad_norm": 5.28125,
      "learning_rate": 2.615446593741161e-06,
      "loss": 0.542,
      "step": 1970
    },
    {
      "epoch": 0.8641571194762684,
      "grad_norm": 4.25,
      "learning_rate": 2.522019095014683e-06,
      "loss": 0.6423,
      "step": 1980
    },
    {
      "epoch": 0.8685215493726132,
      "grad_norm": 2.375,
      "learning_rate": 2.4300494434824373e-06,
      "loss": 0.6308,
      "step": 1990
    },
    {
      "epoch": 0.872885979268958,
      "grad_norm": 5.53125,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.6274,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 2500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.65313243152384e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}