{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2292,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004363953742090334,
      "grad_norm": 26.499397821507202,
      "learning_rate": 3.91304347826087e-07,
      "loss": 1.5146,
      "step": 10
    },
    {
      "epoch": 0.008727907484180668,
      "grad_norm": 18.696221795032976,
      "learning_rate": 8.260869565217392e-07,
      "loss": 1.3975,
      "step": 20
    },
    {
      "epoch": 0.013091861226271002,
      "grad_norm": 5.195746812383845,
      "learning_rate": 1.2608695652173913e-06,
      "loss": 1.0301,
      "step": 30
    },
    {
      "epoch": 0.017455814968361336,
      "grad_norm": 2.9388986813240305,
      "learning_rate": 1.6956521739130435e-06,
      "loss": 0.9338,
      "step": 40
    },
    {
      "epoch": 0.02181976871045167,
      "grad_norm": 2.157896980678217,
      "learning_rate": 2.130434782608696e-06,
      "loss": 0.834,
      "step": 50
    },
    {
      "epoch": 0.026183722452542003,
      "grad_norm": 2.030356833343019,
      "learning_rate": 2.5652173913043484e-06,
      "loss": 0.7817,
      "step": 60
    },
    {
      "epoch": 0.030547676194632337,
      "grad_norm": 2.010110703082707,
      "learning_rate": 3e-06,
      "loss": 0.7423,
      "step": 70
    },
    {
      "epoch": 0.03491162993672267,
      "grad_norm": 2.036582614030934,
      "learning_rate": 3.4347826086956526e-06,
      "loss": 0.7193,
      "step": 80
    },
    {
      "epoch": 0.039275583678813,
      "grad_norm": 1.9487163653762136,
      "learning_rate": 3.869565217391304e-06,
      "loss": 0.6965,
      "step": 90
    },
    {
      "epoch": 0.04363953742090334,
      "grad_norm": 1.8884072181165537,
      "learning_rate": 4.304347826086957e-06,
      "loss": 0.6863,
      "step": 100
    },
    {
      "epoch": 0.04800349116299367,
      "grad_norm": 1.9672176130957406,
      "learning_rate": 4.739130434782609e-06,
      "loss": 0.6761,
      "step": 110
    },
    {
      "epoch": 0.05236744490508401,
      "grad_norm": 1.9175860223528627,
      "learning_rate": 5.173913043478262e-06,
      "loss": 0.669,
      "step": 120
    },
    {
      "epoch": 0.05673139864717434,
      "grad_norm": 4.64170340673796,
      "learning_rate": 5.608695652173914e-06,
      "loss": 0.6553,
      "step": 130
    },
    {
      "epoch": 0.061095352389264675,
      "grad_norm": 2.4419914657221766,
      "learning_rate": 6.043478260869565e-06,
      "loss": 0.6635,
      "step": 140
    },
    {
      "epoch": 0.065459306131355,
      "grad_norm": 2.219811974469962,
      "learning_rate": 6.478260869565218e-06,
      "loss": 0.6417,
      "step": 150
    },
    {
      "epoch": 0.06982325987344534,
      "grad_norm": 1.8731532270945315,
      "learning_rate": 6.91304347826087e-06,
      "loss": 0.6427,
      "step": 160
    },
    {
      "epoch": 0.07418721361553568,
      "grad_norm": 1.9110914927729745,
      "learning_rate": 7.347826086956522e-06,
      "loss": 0.6442,
      "step": 170
    },
    {
      "epoch": 0.078551167357626,
      "grad_norm": 1.9295608959806378,
      "learning_rate": 7.782608695652174e-06,
      "loss": 0.6369,
      "step": 180
    },
    {
      "epoch": 0.08291512109971634,
      "grad_norm": 1.7914358748612738,
      "learning_rate": 8.217391304347827e-06,
      "loss": 0.639,
      "step": 190
    },
    {
      "epoch": 0.08727907484180668,
      "grad_norm": 1.9446891106680222,
      "learning_rate": 8.65217391304348e-06,
      "loss": 0.6414,
      "step": 200
    },
    {
      "epoch": 0.09164302858389702,
      "grad_norm": 2.1347110290826032,
      "learning_rate": 9.086956521739132e-06,
      "loss": 0.6392,
      "step": 210
    },
    {
      "epoch": 0.09600698232598734,
      "grad_norm": 1.9030077411906519,
      "learning_rate": 9.521739130434784e-06,
      "loss": 0.6407,
      "step": 220
    },
    {
      "epoch": 0.10037093606807768,
      "grad_norm": 1.9072699967620688,
      "learning_rate": 9.956521739130436e-06,
      "loss": 0.6296,
      "step": 230
    },
    {
      "epoch": 0.10473488981016801,
      "grad_norm": 6.11411147753339,
      "learning_rate": 9.999529953691501e-06,
      "loss": 0.6364,
      "step": 240
    },
    {
      "epoch": 0.10909884355225835,
      "grad_norm": 1.9274840356722631,
      "learning_rate": 9.997905215717636e-06,
      "loss": 0.6271,
      "step": 250
    },
    {
      "epoch": 0.11346279729434867,
      "grad_norm": 1.6084651209610894,
      "learning_rate": 9.995120360083208e-06,
      "loss": 0.625,
      "step": 260
    },
    {
      "epoch": 0.11782675103643901,
      "grad_norm": 1.9201552574955696,
      "learning_rate": 9.991176033211077e-06,
      "loss": 0.6271,
      "step": 270
    },
    {
      "epoch": 0.12219070477852935,
      "grad_norm": 1.8527351849291473,
      "learning_rate": 9.98607315066148e-06,
      "loss": 0.62,
      "step": 280
    },
    {
      "epoch": 0.12655465852061967,
      "grad_norm": 2.1728262084174137,
      "learning_rate": 9.979812896919497e-06,
      "loss": 0.6234,
      "step": 290
    },
    {
      "epoch": 0.13091861226271,
      "grad_norm": 2.5508242501465848,
      "learning_rate": 9.97239672512011e-06,
      "loss": 0.6228,
      "step": 300
    },
    {
      "epoch": 0.13528256600480035,
      "grad_norm": 1.8024605834099054,
      "learning_rate": 9.963826356710913e-06,
      "loss": 0.6189,
      "step": 310
    },
    {
      "epoch": 0.13964651974689068,
      "grad_norm": 1.7899544274401067,
      "learning_rate": 9.954103781052508e-06,
      "loss": 0.622,
      "step": 320
    },
    {
      "epoch": 0.14401047348898102,
      "grad_norm": 2.6634929838451162,
      "learning_rate": 9.943231254956749e-06,
      "loss": 0.6215,
      "step": 330
    },
    {
      "epoch": 0.14837442723107136,
      "grad_norm": 1.6887392545009612,
      "learning_rate": 9.93121130216288e-06,
      "loss": 0.6095,
      "step": 340
    },
    {
      "epoch": 0.1527383809731617,
      "grad_norm": 1.564527068427684,
      "learning_rate": 9.918046712751731e-06,
      "loss": 0.6073,
      "step": 350
    },
    {
      "epoch": 0.157102334715252,
      "grad_norm": 1.7759578419290645,
      "learning_rate": 9.903740542498071e-06,
      "loss": 0.6095,
      "step": 360
    },
    {
      "epoch": 0.16146628845734234,
      "grad_norm": 2.604040660939313,
      "learning_rate": 9.888296112161308e-06,
      "loss": 0.6039,
      "step": 370
    },
    {
      "epoch": 0.16583024219943268,
      "grad_norm": 1.7552723705971771,
      "learning_rate": 9.87171700671467e-06,
      "loss": 0.6049,
      "step": 380
    },
    {
      "epoch": 0.17019419594152302,
      "grad_norm": 2.786695194524141,
      "learning_rate": 9.854007074513056e-06,
      "loss": 0.6069,
      "step": 390
    },
    {
      "epoch": 0.17455814968361336,
      "grad_norm": 1.6966428135982474,
      "learning_rate": 9.835170426399757e-06,
      "loss": 0.6067,
      "step": 400
    },
    {
      "epoch": 0.1789221034257037,
      "grad_norm": 3.1900936362907073,
      "learning_rate": 9.815211434752236e-06,
      "loss": 0.6021,
      "step": 410
    },
    {
      "epoch": 0.18328605716779403,
      "grad_norm": 2.0950790085022795,
      "learning_rate": 9.79413473246722e-06,
      "loss": 0.6064,
      "step": 420
    },
    {
      "epoch": 0.18765001090988437,
      "grad_norm": 1.8573324559272892,
      "learning_rate": 9.771945211885294e-06,
      "loss": 0.5962,
      "step": 430
    },
    {
      "epoch": 0.19201396465197468,
      "grad_norm": 1.7147053708189852,
      "learning_rate": 9.748648023655302e-06,
      "loss": 0.6005,
      "step": 440
    },
    {
      "epoch": 0.19637791839406502,
      "grad_norm": 1.620349780247674,
      "learning_rate": 9.724248575538764e-06,
      "loss": 0.5922,
      "step": 450
    },
    {
      "epoch": 0.20074187213615535,
      "grad_norm": 1.654417334838078,
      "learning_rate": 9.698752531154623e-06,
      "loss": 0.5977,
      "step": 460
    },
    {
      "epoch": 0.2051058258782457,
      "grad_norm": 3.7007534847341335,
      "learning_rate": 9.672165808664609e-06,
      "loss": 0.5989,
      "step": 470
    },
    {
      "epoch": 0.20946977962033603,
      "grad_norm": 1.8306160811104213,
      "learning_rate": 9.6444945793995e-06,
      "loss": 0.5914,
      "step": 480
    },
    {
      "epoch": 0.21383373336242636,
      "grad_norm": 1.9518306293763308,
      "learning_rate": 9.61574526642664e-06,
      "loss": 0.5883,
      "step": 490
    },
    {
      "epoch": 0.2181976871045167,
      "grad_norm": 1.4825723400886388,
      "learning_rate": 9.585924543058997e-06,
      "loss": 0.591,
      "step": 500
    },
    {
      "epoch": 0.2181976871045167,
      "eval_loss": 0.5827435255050659,
      "eval_runtime": 69.7858,
      "eval_samples_per_second": 53.062,
      "eval_steps_per_second": 3.324,
      "step": 500
    },
    {
      "epoch": 0.22256164084660704,
      "grad_norm": 1.480013652739435,
      "learning_rate": 9.555039331306164e-06,
      "loss": 0.5893,
      "step": 510
    },
    {
      "epoch": 0.22692559458869735,
      "grad_norm": 1.697064020550412,
      "learning_rate": 9.523096800267602e-06,
      "loss": 0.5967,
      "step": 520
    },
    {
      "epoch": 0.2312895483307877,
      "grad_norm": 1.4577169028061336,
      "learning_rate": 9.490104364468557e-06,
      "loss": 0.5876,
      "step": 530
    },
    {
      "epoch": 0.23565350207287802,
      "grad_norm": 1.407863250608549,
      "learning_rate": 9.456069682138981e-06,
      "loss": 0.5879,
      "step": 540
    },
    {
      "epoch": 0.24001745581496836,
      "grad_norm": 1.5047099371255703,
      "learning_rate": 9.421000653435908e-06,
      "loss": 0.5898,
      "step": 550
    },
    {
      "epoch": 0.2443814095570587,
      "grad_norm": 1.5343210297318903,
      "learning_rate": 9.38490541860966e-06,
      "loss": 0.5938,
      "step": 560
    },
    {
      "epoch": 0.24874536329914904,
      "grad_norm": 1.4956667878525625,
      "learning_rate": 9.347792356114324e-06,
      "loss": 0.5893,
      "step": 570
    },
    {
      "epoch": 0.25310931704123935,
      "grad_norm": 1.5101062644216703,
      "learning_rate": 9.30967008066294e-06,
      "loss": 0.5976,
      "step": 580
    },
    {
      "epoch": 0.2574732707833297,
      "grad_norm": 1.5236342080851415,
      "learning_rate": 9.27054744122785e-06,
      "loss": 0.592,
      "step": 590
    },
    {
      "epoch": 0.26183722452542,
      "grad_norm": 1.6140869174258692,
      "learning_rate": 9.230433518986664e-06,
      "loss": 0.5888,
      "step": 600
    },
    {
      "epoch": 0.2662011782675104,
      "grad_norm": 1.419486174653226,
      "learning_rate": 9.189337625214324e-06,
      "loss": 0.5842,
      "step": 610
    },
    {
      "epoch": 0.2705651320096007,
      "grad_norm": 1.5576946265406408,
      "learning_rate": 9.147269299121782e-06,
      "loss": 0.582,
      "step": 620
    },
    {
      "epoch": 0.274929085751691,
      "grad_norm": 1.4457336642256946,
      "learning_rate": 9.104238305641731e-06,
      "loss": 0.579,
      "step": 630
    },
    {
      "epoch": 0.27929303949378137,
      "grad_norm": 1.5422960552682006,
      "learning_rate": 9.060254633161973e-06,
      "loss": 0.5889,
      "step": 640
    },
    {
      "epoch": 0.2836569932358717,
      "grad_norm": 1.5024597086356501,
      "learning_rate": 9.015328491206901e-06,
      "loss": 0.5793,
      "step": 650
    },
    {
      "epoch": 0.28802094697796204,
      "grad_norm": 1.3686070974882203,
      "learning_rate": 8.969470308067662e-06,
      "loss": 0.5807,
      "step": 660
    },
    {
      "epoch": 0.29238490072005235,
      "grad_norm": 2.6998427178862894,
      "learning_rate": 8.922690728381527e-06,
      "loss": 0.5748,
      "step": 670
    },
    {
      "epoch": 0.2967488544621427,
      "grad_norm": 2.065649261696864,
      "learning_rate": 8.875000610661052e-06,
      "loss": 0.5688,
      "step": 680
    },
    {
      "epoch": 0.30111280820423303,
      "grad_norm": 1.5958849520041398,
      "learning_rate": 8.826411024773595e-06,
      "loss": 0.5727,
      "step": 690
    },
    {
      "epoch": 0.3054767619463234,
      "grad_norm": 1.4305834347213016,
      "learning_rate": 8.776933249371769e-06,
      "loss": 0.5621,
      "step": 700
    },
    {
      "epoch": 0.3098407156884137,
      "grad_norm": 1.4826756109906936,
      "learning_rate": 8.726578769275426e-06,
      "loss": 0.57,
      "step": 710
    },
    {
      "epoch": 0.314204669430504,
      "grad_norm": 1.4919661778521895,
      "learning_rate": 8.675359272805796e-06,
      "loss": 0.5733,
      "step": 720
    },
    {
      "epoch": 0.3185686231725944,
      "grad_norm": 1.539486119661036,
      "learning_rate": 8.62328664907239e-06,
      "loss": 0.5617,
      "step": 730
    },
    {
      "epoch": 0.3229325769146847,
      "grad_norm": 2.412337158368947,
      "learning_rate": 8.570372985213283e-06,
      "loss": 0.5748,
      "step": 740
    },
    {
      "epoch": 0.32729653065677505,
      "grad_norm": 1.4949607784946515,
      "learning_rate": 8.516630563589436e-06,
      "loss": 0.5765,
      "step": 750
    },
    {
      "epoch": 0.33166048439886536,
      "grad_norm": 2.1389988555015527,
      "learning_rate": 8.462071858933717e-06,
      "loss": 0.5717,
      "step": 760
    },
    {
      "epoch": 0.33602443814095573,
      "grad_norm": 1.7992243083658257,
      "learning_rate": 8.406709535455242e-06,
      "loss": 0.5728,
      "step": 770
    },
    {
      "epoch": 0.34038839188304604,
      "grad_norm": 1.7459966293740241,
      "learning_rate": 8.35055644389976e-06,
      "loss": 0.5621,
      "step": 780
    },
    {
      "epoch": 0.34475234562513635,
      "grad_norm": 1.6584257041747206,
      "learning_rate": 8.293625618566723e-06,
      "loss": 0.5676,
      "step": 790
    },
    {
      "epoch": 0.3491162993672267,
      "grad_norm": 2.234742145469745,
      "learning_rate": 8.23593027428375e-06,
      "loss": 0.5657,
      "step": 800
    },
    {
      "epoch": 0.353480253109317,
      "grad_norm": 1.4160485895062545,
      "learning_rate": 8.177483803339203e-06,
      "loss": 0.5739,
      "step": 810
    },
    {
      "epoch": 0.3578442068514074,
      "grad_norm": 1.6236310717790317,
      "learning_rate": 8.118299772373546e-06,
      "loss": 0.5708,
      "step": 820
    },
    {
      "epoch": 0.3622081605934977,
      "grad_norm": 1.4409935142518557,
      "learning_rate": 8.05839191923025e-06,
      "loss": 0.5652,
      "step": 830
    },
    {
      "epoch": 0.36657211433558806,
      "grad_norm": 1.7422827738179059,
      "learning_rate": 7.997774149766962e-06,
      "loss": 0.5665,
      "step": 840
    },
    {
      "epoch": 0.37093606807767837,
      "grad_norm": 1.4654585573940748,
      "learning_rate": 7.936460534627648e-06,
      "loss": 0.5592,
      "step": 850
    },
    {
      "epoch": 0.37530002181976874,
      "grad_norm": 1.5684193622827431,
      "learning_rate": 7.874465305976527e-06,
      "loss": 0.558,
      "step": 860
    },
    {
      "epoch": 0.37966397556185905,
      "grad_norm": 1.8405640707393334,
      "learning_rate": 7.81180285419447e-06,
      "loss": 0.5653,
      "step": 870
    },
    {
      "epoch": 0.38402792930394936,
      "grad_norm": 1.475228530845922,
      "learning_rate": 7.748487724538707e-06,
      "loss": 0.5587,
      "step": 880
    },
    {
      "epoch": 0.3883918830460397,
      "grad_norm": 1.331005791046321,
      "learning_rate": 7.684534613766566e-06,
      "loss": 0.574,
      "step": 890
    },
    {
      "epoch": 0.39275583678813003,
      "grad_norm": 1.3582718830868108,
      "learning_rate": 7.619958366724043e-06,
      "loss": 0.5657,
      "step": 900
    },
    {
      "epoch": 0.3971197905302204,
      "grad_norm": 1.3639931742020133,
      "learning_rate": 7.554773972900024e-06,
      "loss": 0.5642,
      "step": 910
    },
    {
      "epoch": 0.4014837442723107,
      "grad_norm": 1.4421529537691287,
      "learning_rate": 7.488996562946886e-06,
      "loss": 0.5585,
      "step": 920
    },
    {
      "epoch": 0.40584769801440107,
      "grad_norm": 1.4453139064733957,
      "learning_rate": 7.422641405168386e-06,
      "loss": 0.5611,
      "step": 930
    },
    {
      "epoch": 0.4102116517564914,
      "grad_norm": 1.41874403862543,
      "learning_rate": 7.355723901975546e-06,
      "loss": 0.5556,
      "step": 940
    },
    {
      "epoch": 0.4145756054985817,
      "grad_norm": 1.569355186279227,
      "learning_rate": 7.288259586311455e-06,
      "loss": 0.5638,
      "step": 950
    },
    {
      "epoch": 0.41893955924067205,
      "grad_norm": 1.340010425244894,
      "learning_rate": 7.22026411804573e-06,
      "loss": 0.554,
      "step": 960
    },
    {
      "epoch": 0.42330351298276236,
      "grad_norm": 1.4102271417835663,
      "learning_rate": 7.151753280339554e-06,
      "loss": 0.5574,
      "step": 970
    },
    {
      "epoch": 0.42766746672485273,
      "grad_norm": 1.4168840313026727,
      "learning_rate": 7.082742975982057e-06,
      "loss": 0.5613,
      "step": 980
    },
    {
      "epoch": 0.43203142046694304,
      "grad_norm": 1.4932003760222727,
      "learning_rate": 7.0132492236989694e-06,
      "loss": 0.5549,
      "step": 990
    },
    {
      "epoch": 0.4363953742090334,
      "grad_norm": 1.3727246781695526,
      "learning_rate": 6.94328815443432e-06,
      "loss": 0.5605,
      "step": 1000
    },
    {
      "epoch": 0.4363953742090334,
      "eval_loss": 0.545971155166626,
      "eval_runtime": 68.1299,
      "eval_samples_per_second": 54.352,
      "eval_steps_per_second": 3.405,
      "step": 1000
    },
    {
      "epoch": 0.4407593279511237,
      "grad_norm": 1.5610600215808323,
      "learning_rate": 6.872876007606127e-06,
      "loss": 0.5505,
      "step": 1010
    },
    {
      "epoch": 0.4451232816932141,
      "grad_norm": 1.2665441451258144,
      "learning_rate": 6.802029127336884e-06,
      "loss": 0.5511,
      "step": 1020
    },
    {
      "epoch": 0.4494872354353044,
      "grad_norm": 1.4767565256072326,
      "learning_rate": 6.73076395865975e-06,
      "loss": 0.5479,
      "step": 1030
    },
    {
      "epoch": 0.4538511891773947,
      "grad_norm": 1.3450898726192824,
      "learning_rate": 6.6590970437013135e-06,
      "loss": 0.5549,
      "step": 1040
    },
    {
      "epoch": 0.45821514291948506,
      "grad_norm": 1.3691285786320522,
      "learning_rate": 6.587045017841828e-06,
      "loss": 0.5566,
      "step": 1050
    },
    {
      "epoch": 0.4625790966615754,
      "grad_norm": 1.407701866024824,
      "learning_rate": 6.514624605853785e-06,
      "loss": 0.5519,
      "step": 1060
    },
    {
      "epoch": 0.46694305040366574,
      "grad_norm": 1.7016011593861633,
      "learning_rate": 6.441852618019757e-06,
      "loss": 0.5453,
      "step": 1070
    },
    {
      "epoch": 0.47130700414575605,
      "grad_norm": 2.0081488788972495,
      "learning_rate": 6.368745946230371e-06,
      "loss": 0.5509,
      "step": 1080
    },
    {
      "epoch": 0.4756709578878464,
      "grad_norm": 1.4313272433739415,
      "learning_rate": 6.295321560063358e-06,
      "loss": 0.5558,
      "step": 1090
    },
    {
      "epoch": 0.4800349116299367,
      "grad_norm": 1.3908306510503996,
      "learning_rate": 6.221596502844558e-06,
      "loss": 0.5496,
      "step": 1100
    },
    {
      "epoch": 0.48439886537202703,
      "grad_norm": 1.4083635646126667,
      "learning_rate": 6.147587887691812e-06,
      "loss": 0.5447,
      "step": 1110
    },
    {
      "epoch": 0.4887628191141174,
      "grad_norm": 17.48008864471473,
      "learning_rate": 6.073312893542644e-06,
      "loss": 0.5507,
      "step": 1120
    },
    {
      "epoch": 0.4931267728562077,
      "grad_norm": 1.7242442516057832,
      "learning_rate": 5.998788761166689e-06,
      "loss": 0.5497,
      "step": 1130
    },
    {
      "epoch": 0.49749072659829807,
      "grad_norm": 1.6226845226734892,
      "learning_rate": 5.9240327891637296e-06,
      "loss": 0.5521,
      "step": 1140
    },
    {
      "epoch": 0.5018546803403884,
      "grad_norm": 1.6604002116905412,
      "learning_rate": 5.849062329948353e-06,
      "loss": 0.5451,
      "step": 1150
    },
    {
      "epoch": 0.5062186340824787,
      "grad_norm": 1.5873724924653319,
      "learning_rate": 5.773894785722082e-06,
      "loss": 0.547,
      "step": 1160
    },
    {
      "epoch": 0.5105825878245691,
      "grad_norm": 1.4609962799328036,
      "learning_rate": 5.698547604433963e-06,
      "loss": 0.551,
      "step": 1170
    },
    {
      "epoch": 0.5149465415666594,
      "grad_norm": 2.35781566677412,
      "learning_rate": 5.623038275730543e-06,
      "loss": 0.5482,
      "step": 1180
    },
    {
      "epoch": 0.5193104953087497,
      "grad_norm": 1.7956175108962174,
      "learning_rate": 5.547384326896152e-06,
      "loss": 0.5378,
      "step": 1190
    },
    {
      "epoch": 0.52367444905084,
      "grad_norm": 1.3754392835784166,
      "learning_rate": 5.4716033187844565e-06,
      "loss": 0.5289,
      "step": 1200
    },
    {
      "epoch": 0.5280384027929304,
      "grad_norm": 1.351125298582442,
      "learning_rate": 5.3957128417422196e-06,
      "loss": 0.5375,
      "step": 1210
    },
    {
      "epoch": 0.5324023565350208,
      "grad_norm": 1.2885359407090353,
      "learning_rate": 5.319730511526225e-06,
      "loss": 0.5484,
      "step": 1220
    },
    {
      "epoch": 0.5367663102771111,
      "grad_norm": 1.4931791483959027,
      "learning_rate": 5.243673965214276e-06,
      "loss": 0.5352,
      "step": 1230
    },
    {
      "epoch": 0.5411302640192014,
      "grad_norm": 1.3301082529926556,
      "learning_rate": 5.167560857111286e-06,
      "loss": 0.5407,
      "step": 1240
    },
    {
      "epoch": 0.5454942177612917,
      "grad_norm": 1.3023885557832229,
      "learning_rate": 5.091408854651327e-06,
      "loss": 0.5354,
      "step": 1250
    },
    {
      "epoch": 0.549858171503382,
      "grad_norm": 1.8840653667484866,
      "learning_rate": 5.0152356342966624e-06,
      "loss": 0.5425,
      "step": 1260
    },
    {
      "epoch": 0.5542221252454724,
      "grad_norm": 1.3585936189959857,
      "learning_rate": 4.939058877434672e-06,
      "loss": 0.5386,
      "step": 1270
    },
    {
      "epoch": 0.5585860789875627,
      "grad_norm": 1.4612137856640806,
      "learning_rate": 4.862896266273627e-06,
      "loss": 0.5512,
      "step": 1280
    },
    {
      "epoch": 0.562950032729653,
      "grad_norm": 1.3030406905808916,
      "learning_rate": 4.786765479738293e-06,
      "loss": 0.5422,
      "step": 1290
    },
    {
      "epoch": 0.5673139864717434,
      "grad_norm": 1.3169625647849659,
      "learning_rate": 4.7106841893662755e-06,
      "loss": 0.5377,
      "step": 1300
    },
    {
      "epoch": 0.5716779402138338,
      "grad_norm": 1.4365133088202295,
      "learning_rate": 4.634670055206092e-06,
      "loss": 0.5429,
      "step": 1310
    },
    {
      "epoch": 0.5760418939559241,
      "grad_norm": 1.400666928678306,
      "learning_rate": 4.5587407217179094e-06,
      "loss": 0.5441,
      "step": 1320
    },
    {
      "epoch": 0.5804058476980144,
      "grad_norm": 1.4020862611201026,
      "learning_rate": 4.482913813677896e-06,
      "loss": 0.5372,
      "step": 1330
    },
    {
      "epoch": 0.5847698014401047,
      "grad_norm": 1.3899074056942011,
      "learning_rate": 4.407206932087143e-06,
      "loss": 0.5431,
      "step": 1340
    },
    {
      "epoch": 0.589133755182195,
      "grad_norm": 1.2780180293099543,
      "learning_rate": 4.331637650086111e-06,
      "loss": 0.5354,
      "step": 1350
    },
    {
      "epoch": 0.5934977089242854,
      "grad_norm": 1.3047906646065353,
      "learning_rate": 4.256223508875536e-06,
      "loss": 0.536,
      "step": 1360
    },
    {
      "epoch": 0.5978616626663757,
      "grad_norm": 2.2450550137169025,
      "learning_rate": 4.180982013644749e-06,
      "loss": 0.5345,
      "step": 1370
    },
    {
      "epoch": 0.6022256164084661,
      "grad_norm": 1.2604878888770283,
      "learning_rate": 4.105930629508369e-06,
      "loss": 0.5349,
      "step": 1380
    },
    {
      "epoch": 0.6065895701505564,
      "grad_norm": 1.4408134764916174,
      "learning_rate": 4.0310867774522724e-06,
      "loss": 0.5338,
      "step": 1390
    },
    {
      "epoch": 0.6109535238926468,
      "grad_norm": 1.194095025493141,
      "learning_rate": 3.95646783028983e-06,
      "loss": 0.5332,
      "step": 1400
    },
    {
      "epoch": 0.6153174776347371,
      "grad_norm": 1.3682237212576729,
      "learning_rate": 3.882091108629311e-06,
      "loss": 0.5305,
      "step": 1410
    },
    {
      "epoch": 0.6196814313768274,
      "grad_norm": 1.3808616625009607,
      "learning_rate": 3.807973876853414e-06,
      "loss": 0.5333,
      "step": 1420
    },
    {
      "epoch": 0.6240453851189177,
      "grad_norm": 1.3166679869381683,
      "learning_rate": 3.734133339111844e-06,
      "loss": 0.5305,
      "step": 1430
    },
    {
      "epoch": 0.628409338861008,
      "grad_norm": 2.075662539665934,
      "learning_rate": 3.660586635327869e-06,
      "loss": 0.5386,
      "step": 1440
    },
    {
      "epoch": 0.6327732926030984,
      "grad_norm": 1.3904648380650928,
      "learning_rate": 3.587350837219788e-06,
      "loss": 0.5396,
      "step": 1450
    },
    {
      "epoch": 0.6371372463451888,
      "grad_norm": 1.492731925374217,
      "learning_rate": 3.5144429443382356e-06,
      "loss": 0.5366,
      "step": 1460
    },
    {
      "epoch": 0.6415012000872791,
      "grad_norm": 1.2451404313444014,
      "learning_rate": 3.4418798801202256e-06,
      "loss": 0.5255,
      "step": 1470
    },
    {
      "epoch": 0.6458651538293694,
      "grad_norm": 1.5668762418599276,
      "learning_rate": 3.3696784879608747e-06,
      "loss": 0.5307,
      "step": 1480
    },
    {
      "epoch": 0.6502291075714598,
      "grad_norm": 1.4372909935764755,
      "learning_rate": 3.2978555273037006e-06,
      "loss": 0.5269,
      "step": 1490
    },
    {
      "epoch": 0.6545930613135501,
      "grad_norm": 1.3776181734494666,
      "learning_rate": 3.2264276697504026e-06,
      "loss": 0.5252,
      "step": 1500
    },
    {
      "epoch": 0.6545930613135501,
      "eval_loss": 0.5199297666549683,
      "eval_runtime": 69.8992,
      "eval_samples_per_second": 52.976,
      "eval_steps_per_second": 3.319,
      "step": 1500
    },
    {
      "epoch": 0.6589570150556404,
      "grad_norm": 1.4365144824102365,
      "learning_rate": 3.1554114951910387e-06,
      "loss": 0.5312,
      "step": 1510
    },
    {
      "epoch": 0.6633209687977307,
      "grad_norm": 1.2422304913068782,
      "learning_rate": 3.0848234879554916e-06,
      "loss": 0.5283,
      "step": 1520
    },
    {
      "epoch": 0.667684922539821,
      "grad_norm": 1.2495828228531431,
      "learning_rate": 3.0146800329871107e-06,
      "loss": 0.5359,
      "step": 1530
    },
    {
      "epoch": 0.6720488762819115,
      "grad_norm": 1.3009383931934748,
      "learning_rate": 2.944997412039422e-06,
      "loss": 0.5286,
      "step": 1540
    },
    {
      "epoch": 0.6764128300240018,
      "grad_norm": 1.3215444464354889,
      "learning_rate": 2.8757917998968042e-06,
      "loss": 0.5345,
      "step": 1550
    },
    {
      "epoch": 0.6807767837660921,
      "grad_norm": 1.199193600939233,
      "learning_rate": 2.807079260619976e-06,
      "loss": 0.5226,
      "step": 1560
    },
    {
      "epoch": 0.6851407375081824,
      "grad_norm": 1.3916367911407932,
      "learning_rate": 2.7388757438171953e-06,
      "loss": 0.5294,
      "step": 1570
    },
    {
      "epoch": 0.6895046912502727,
      "grad_norm": 1.2821114411807195,
      "learning_rate": 2.6711970809420327e-06,
      "loss": 0.5243,
      "step": 1580
    },
    {
      "epoch": 0.6938686449923631,
      "grad_norm": 1.2276903910497599,
      "learning_rate": 2.6040589816185534e-06,
      "loss": 0.5227,
      "step": 1590
    },
    {
      "epoch": 0.6982325987344534,
      "grad_norm": 1.2534508386414063,
      "learning_rate": 2.5374770299947837e-06,
      "loss": 0.5336,
      "step": 1600
    },
    {
      "epoch": 0.7025965524765437,
      "grad_norm": 1.2952567981586396,
      "learning_rate": 2.471466681125316e-06,
      "loss": 0.5242,
      "step": 1610
    },
    {
      "epoch": 0.706960506218634,
      "grad_norm": 1.2491075248862247,
      "learning_rate": 2.4060432573838686e-06,
      "loss": 0.5229,
      "step": 1620
    },
    {
      "epoch": 0.7113244599607245,
      "grad_norm": 1.3070997549733596,
      "learning_rate": 2.3412219449066316e-06,
      "loss": 0.5238,
      "step": 1630
    },
    {
      "epoch": 0.7156884137028148,
      "grad_norm": 1.2500083040545036,
      "learning_rate": 2.2770177900672658e-06,
      "loss": 0.527,
      "step": 1640
    },
    {
      "epoch": 0.7200523674449051,
      "grad_norm": 1.2464501526769323,
      "learning_rate": 2.213445695984318e-06,
      "loss": 0.5208,
      "step": 1650
    },
    {
      "epoch": 0.7244163211869954,
      "grad_norm": 1.2752274285884029,
      "learning_rate": 2.150520419061896e-06,
      "loss": 0.5268,
      "step": 1660
    },
    {
      "epoch": 0.7287802749290857,
      "grad_norm": 1.243680862493251,
      "learning_rate": 2.0882565655644054e-06,
      "loss": 0.5233,
      "step": 1670
    },
    {
      "epoch": 0.7331442286711761,
      "grad_norm": 1.6747464378055472,
      "learning_rate": 2.026668588226133e-06,
      "loss": 0.5254,
      "step": 1680
    },
    {
      "epoch": 0.7375081824132664,
      "grad_norm": 1.2620103056544976,
      "learning_rate": 1.965770782896455e-06,
      "loss": 0.5224,
      "step": 1690
    },
    {
      "epoch": 0.7418721361553567,
      "grad_norm": 1.3278419822937977,
      "learning_rate": 1.9055772852214916e-06,
      "loss": 0.515,
      "step": 1700
    },
    {
      "epoch": 0.746236089897447,
      "grad_norm": 1.6987330727891572,
      "learning_rate": 1.8461020673629172e-06,
      "loss": 0.5144,
      "step": 1710
    },
    {
      "epoch": 0.7506000436395375,
      "grad_norm": 1.3497988296955354,
      "learning_rate": 1.7873589347547321e-06,
      "loss": 0.5173,
      "step": 1720
    },
    {
      "epoch": 0.7549639973816278,
      "grad_norm": 1.6208371219517643,
      "learning_rate": 1.729361522898737e-06,
      "loss": 0.5228,
      "step": 1730
    },
    {
      "epoch": 0.7593279511237181,
      "grad_norm": 1.308791373132814,
      "learning_rate": 1.6721232941994526e-06,
      "loss": 0.5282,
      "step": 1740
    },
    {
      "epoch": 0.7636919048658084,
      "grad_norm": 1.2618118893101176,
      "learning_rate": 1.6156575348392122e-06,
      "loss": 0.5208,
      "step": 1750
    },
    {
      "epoch": 0.7680558586078987,
      "grad_norm": 1.3104736125525216,
      "learning_rate": 1.559977351694158e-06,
      "loss": 0.5208,
      "step": 1760
    },
    {
      "epoch": 0.7724198123499891,
      "grad_norm": 1.2487367116120625,
      "learning_rate": 1.5050956692918739e-06,
      "loss": 0.5186,
      "step": 1770
    },
    {
      "epoch": 0.7767837660920794,
      "grad_norm": 1.156634335046638,
      "learning_rate": 1.4510252268113263e-06,
      "loss": 0.5192,
      "step": 1780
    },
    {
      "epoch": 0.7811477198341698,
      "grad_norm": 1.3021570425828364,
      "learning_rate": 1.3977785751258345e-06,
      "loss": 0.5211,
      "step": 1790
    },
    {
      "epoch": 0.7855116735762601,
      "grad_norm": 1.2504588787454711,
      "learning_rate": 1.34536807388976e-06,
      "loss": 0.5159,
      "step": 1800
    },
    {
      "epoch": 0.7898756273183504,
      "grad_norm": 1.991690391624192,
      "learning_rate": 1.2938058886695643e-06,
      "loss": 0.5169,
      "step": 1810
    },
    {
      "epoch": 0.7942395810604408,
      "grad_norm": 1.4484577791210265,
      "learning_rate": 1.2431039881199374e-06,
      "loss": 0.5246,
      "step": 1820
    },
    {
      "epoch": 0.7986035348025311,
      "grad_norm": 1.2031874504717428,
      "learning_rate": 1.1932741412056187e-06,
      "loss": 0.5238,
      "step": 1830
    },
    {
      "epoch": 0.8029674885446214,
      "grad_norm": 1.4251143516882596,
      "learning_rate": 1.1443279144695746e-06,
      "loss": 0.5202,
      "step": 1840
    },
    {
      "epoch": 0.8073314422867117,
      "grad_norm": 1.428433930967254,
      "learning_rate": 1.0962766693481686e-06,
      "loss": 0.5178,
      "step": 1850
    },
    {
      "epoch": 0.8116953960288021,
      "grad_norm": 1.3957444052098604,
      "learning_rate": 1.049131559533933e-06,
      "loss": 0.5142,
      "step": 1860
    },
    {
      "epoch": 0.8160593497708925,
      "grad_norm": 1.7513113973659105,
      "learning_rate": 1.002903528386564e-06,
      "loss": 0.5188,
      "step": 1870
    },
    {
      "epoch": 0.8204233035129828,
      "grad_norm": 1.3254734615581885,
      "learning_rate": 9.576033063927398e-07,
      "loss": 0.5207,
      "step": 1880
    },
    {
      "epoch": 0.8247872572550731,
      "grad_norm": 1.3416379717629452,
      "learning_rate": 9.132414086753578e-07,
      "loss": 0.5134,
      "step": 1890
    },
    {
      "epoch": 0.8291512109971634,
      "grad_norm": 1.2636084433299684,
      "learning_rate": 8.698281325527502e-07,
      "loss": 0.5144,
      "step": 1900
    },
    {
      "epoch": 0.8335151647392538,
      "grad_norm": 1.718038484652991,
      "learning_rate": 8.273735551484613e-07,
      "loss": 0.5158,
      "step": 1910
    },
    {
      "epoch": 0.8378791184813441,
      "grad_norm": 1.365124227466459,
      "learning_rate": 7.858875310521447e-07,
      "loss": 0.5116,
      "step": 1920
    },
    {
      "epoch": 0.8422430722234344,
      "grad_norm": 1.2857842722903012,
      "learning_rate": 7.45379690032102e-07,
      "loss": 0.5213,
      "step": 1930
    },
    {
      "epoch": 0.8466070259655247,
      "grad_norm": 1.8949074875975938,
      "learning_rate": 7.058594348000142e-07,
      "loss": 0.5126,
      "step": 1940
    },
    {
      "epoch": 0.8509709797076151,
      "grad_norm": 2.0508297274530425,
      "learning_rate": 6.673359388283796e-07,
      "loss": 0.5118,
      "step": 1950
    },
    {
      "epoch": 0.8553349334497055,
      "grad_norm": 1.2982410941631513,
      "learning_rate": 6.29818144221161e-07,
      "loss": 0.5256,
      "step": 1960
    },
    {
      "epoch": 0.8596988871917958,
      "grad_norm": 1.3881080528727965,
      "learning_rate": 5.933147596381295e-07,
      "loss": 0.5178,
      "step": 1970
    },
    {
      "epoch": 0.8640628409338861,
      "grad_norm": 1.2689567204311927,
      "learning_rate": 5.578342582734153e-07,
      "loss": 0.5106,
      "step": 1980
    },
    {
      "epoch": 0.8684267946759764,
      "grad_norm": 1.3514399003622013,
      "learning_rate": 5.233848758886945e-07,
      "loss": 0.515,
      "step": 1990
    },
    {
      "epoch": 0.8727907484180668,
      "grad_norm": 1.8878476820927994,
      "learning_rate": 4.899746089015006e-07,
      "loss": 0.5075,
      "step": 2000
    },
    {
      "epoch": 0.8727907484180668,
      "eval_loss": 0.5055009722709656,
      "eval_runtime": 68.9834,
      "eval_samples_per_second": 53.68,
      "eval_steps_per_second": 3.363,
      "step": 2000
    },
    {
      "epoch": 0.8771547021601571,
      "grad_norm": 1.1864343347355735,
      "learning_rate": 4.57611212529091e-07,
      "loss": 0.5118,
      "step": 2010
    },
    {
      "epoch": 0.8815186559022474,
      "grad_norm": 1.3391728742114728,
      "learning_rate": 4.263021989882965e-07,
      "loss": 0.5209,
      "step": 2020
    },
    {
      "epoch": 0.8858826096443377,
      "grad_norm": 1.265995062918622,
      "learning_rate": 3.9605483575178517e-07,
      "loss": 0.5172,
      "step": 2030
    },
    {
      "epoch": 0.8902465633864282,
      "grad_norm": 1.3612996127522512,
      "learning_rate": 3.6687614386112403e-07,
      "loss": 0.51,
      "step": 2040
    },
    {
      "epoch": 0.8946105171285185,
      "grad_norm": 1.1822823920287409,
      "learning_rate": 3.38772896297056e-07,
      "loss": 0.512,
      "step": 2050
    },
    {
      "epoch": 0.8989744708706088,
      "grad_norm": 1.4239619500028629,
      "learning_rate": 3.117516164073459e-07,
      "loss": 0.519,
      "step": 2060
    },
    {
      "epoch": 0.9033384246126991,
      "grad_norm": 1.3503881714270254,
      "learning_rate": 2.858185763925764e-07,
      "loss": 0.5104,
      "step": 2070
    },
    {
      "epoch": 0.9077023783547894,
      "grad_norm": 1.3564880256851914,
      "learning_rate": 2.6097979585024193e-07,
      "loss": 0.5136,
      "step": 2080
    },
    {
      "epoch": 0.9120663320968798,
      "grad_norm": 1.2482713788944904,
      "learning_rate": 2.3724104037747553e-07,
      "loss": 0.5085,
      "step": 2090
    },
    {
      "epoch": 0.9164302858389701,
      "grad_norm": 1.4746577210230283,
      "learning_rate": 2.146078202327284e-07,
      "loss": 0.5159,
      "step": 2100
    },
    {
      "epoch": 0.9207942395810604,
      "grad_norm": 2.19277205698307,
      "learning_rate": 1.9308538905673135e-07,
      "loss": 0.5118,
      "step": 2110
    },
    {
      "epoch": 0.9251581933231507,
      "grad_norm": 1.3172149163534492,
      "learning_rate": 1.7267874265301187e-07,
      "loss": 0.5139,
      "step": 2120
    },
    {
      "epoch": 0.9295221470652411,
      "grad_norm": 1.217984783068752,
      "learning_rate": 1.533926178282641e-07,
      "loss": 0.5084,
      "step": 2130
    },
    {
      "epoch": 0.9338861008073315,
      "grad_norm": 1.3602984814635521,
      "learning_rate": 1.3523149129283987e-07,
      "loss": 0.5222,
      "step": 2140
    },
    {
      "epoch": 0.9382500545494218,
      "grad_norm": 1.3460153750821495,
      "learning_rate": 1.1819957862160835e-07,
      "loss": 0.5069,
      "step": 2150
    },
    {
      "epoch": 0.9426140082915121,
      "grad_norm": 1.4510294571460467,
      "learning_rate": 1.0230083327543661e-07,
      "loss": 0.5142,
      "step": 2160
    },
    {
      "epoch": 0.9469779620336024,
      "grad_norm": 1.396143496017602,
      "learning_rate": 8.753894568350573e-08,
      "loss": 0.5155,
      "step": 2170
    },
    {
      "epoch": 0.9513419157756928,
      "grad_norm": 1.437628267014867,
      "learning_rate": 7.391734238668846e-08,
      "loss": 0.5074,
      "step": 2180
    },
    {
      "epoch": 0.9557058695177831,
      "grad_norm": 1.3148582495813304,
      "learning_rate": 6.143918524217696e-08,
      "loss": 0.5176,
      "step": 2190
    },
    {
      "epoch": 0.9600698232598734,
      "grad_norm": 1.267741565404716,
      "learning_rate": 5.01073706895483e-08,
      "loss": 0.5137,
      "step": 2200
    },
    {
      "epoch": 0.9644337770019638,
      "grad_norm": 1.2859440475396895,
      "learning_rate": 3.992452907844502e-08,
      "loss": 0.5142,
      "step": 2210
    },
    {
      "epoch": 0.9687977307440541,
      "grad_norm": 1.2296422531635047,
      "learning_rate": 3.08930240580102e-08,
      "loss": 0.5025,
      "step": 2220
    },
    {
      "epoch": 0.9731616844861445,
      "grad_norm": 1.3046489064561737,
      "learning_rate": 2.3014952028240223e-08,
      "loss": 0.5138,
      "step": 2230
    },
    {
      "epoch": 0.9775256382282348,
      "grad_norm": 1.238897807436414,
      "learning_rate": 1.6292141653366232e-08,
      "loss": 0.5067,
      "step": 2240
    },
    {
      "epoch": 0.9818895919703251,
      "grad_norm": 1.2689478765861664,
      "learning_rate": 1.0726153437379816e-08,
      "loss": 0.5188,
      "step": 2250
    },
    {
      "epoch": 0.9862535457124154,
      "grad_norm": 2.492246786468264,
      "learning_rate": 6.3182793618110634e-09,
      "loss": 0.5151,
      "step": 2260
    },
    {
      "epoch": 0.9906174994545058,
      "grad_norm": 1.2722668029711148,
      "learning_rate": 3.0695425858329187e-09,
      "loss": 0.5197,
      "step": 2270
    },
    {
      "epoch": 0.9949814531965961,
      "grad_norm": 1.3363964992431479,
      "learning_rate": 9.806972087605949e-10,
      "loss": 0.5223,
      "step": 2280
    },
    {
      "epoch": 0.9993454069386865,
      "grad_norm": 1.5575122083398747,
      "learning_rate": 5.222809501492787e-11,
      "loss": 0.5178,
      "step": 2290
    },
    {
      "epoch": 1.0,
      "step": 2292,
      "total_flos": 2450977789313024.0,
      "train_loss": 0.5714389552828737,
      "train_runtime": 14627.8115,
      "train_samples_per_second": 25.06,
      "train_steps_per_second": 0.157
    }
  ],
  "logging_steps": 10,
  "max_steps": 2292,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2450977789313024.0,
  "train_batch_size": 10,
  "trial_name": null,
  "trial_params": null
}