{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 24920,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 2.4730186462402344,
      "learning_rate": 2e-05,
      "loss": 0.3314,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.50852108001709,
      "learning_rate": 1.9999872447769624e-05,
      "loss": 0.0313,
      "step": 200
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.40423300862312317,
      "learning_rate": 1.9999489794332404e-05,
      "loss": 0.0226,
      "step": 300
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.4923788011074066,
      "learning_rate": 1.9998852049449998e-05,
      "loss": 0.0144,
      "step": 400
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.039026230573654175,
      "learning_rate": 1.9997959229391567e-05,
      "loss": 0.0088,
      "step": 500
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.13339382410049438,
      "learning_rate": 1.9996811356933346e-05,
      "loss": 0.0097,
      "step": 600
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.03016798384487629,
      "learning_rate": 1.9995408461358074e-05,
      "loss": 0.0063,
      "step": 700
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.12673313915729523,
      "learning_rate": 1.9993750578454248e-05,
      "loss": 0.0084,
      "step": 800
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.012224284932017326,
      "learning_rate": 1.999183775051519e-05,
      "loss": 0.006,
      "step": 900
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.144702434539795,
      "learning_rate": 1.9989670026338002e-05,
      "loss": 0.0076,
      "step": 1000
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.02996153011918068,
      "learning_rate": 1.9987247461222297e-05,
      "loss": 0.0052,
      "step": 1100
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.008165411651134491,
      "learning_rate": 1.9984570116968785e-05,
      "loss": 0.0052,
      "step": 1200
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.07781720906496048,
      "learning_rate": 1.9981638061877714e-05,
      "loss": 0.0056,
      "step": 1300
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.016230745241045952,
      "learning_rate": 1.9978451370747122e-05,
      "loss": 0.0052,
      "step": 1400
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.58638858795166,
      "learning_rate": 1.997501012487091e-05,
      "loss": 0.0037,
      "step": 1500
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.029858984053134918,
      "learning_rate": 1.9971314412036807e-05,
      "loss": 0.0037,
      "step": 1600
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.2630612850189209,
      "learning_rate": 1.996736432652409e-05,
      "loss": 0.0046,
      "step": 1700
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.013704606331884861,
      "learning_rate": 1.9963159969101207e-05,
      "loss": 0.0056,
      "step": 1800
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.09422887861728668,
      "learning_rate": 1.9958701447023188e-05,
      "loss": 0.0041,
      "step": 1900
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.02264581061899662,
      "learning_rate": 1.9953988874028917e-05,
      "loss": 0.0039,
      "step": 2000
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.020697351545095444,
      "learning_rate": 1.994902237033824e-05,
      "loss": 0.0048,
      "step": 2100
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.1434525102376938,
      "learning_rate": 1.9943802062648877e-05,
      "loss": 0.0051,
      "step": 2200
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.06430936604738235,
      "learning_rate": 1.9938328084133206e-05,
      "loss": 0.0023,
      "step": 2300
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.008740647695958614,
      "learning_rate": 1.9932600574434864e-05,
      "loss": 0.004,
      "step": 2400
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.014448602683842182,
      "learning_rate": 1.9926619679665175e-05,
      "loss": 0.0043,
      "step": 2500
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.02901746705174446,
      "learning_rate": 1.9920385552399434e-05,
      "loss": 0.0019,
      "step": 2600
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6094048023223877,
      "learning_rate": 1.9913898351673006e-05,
      "loss": 0.0022,
      "step": 2700
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.36414438486099243,
      "learning_rate": 1.990715824297728e-05,
      "loss": 0.0056,
      "step": 2800
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.1421067714691162,
      "learning_rate": 1.9900165398255434e-05,
      "loss": 0.003,
      "step": 2900
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.028148509562015533,
      "learning_rate": 1.9892919995898052e-05,
      "loss": 0.0016,
      "step": 3000
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.015177663415670395,
      "learning_rate": 1.9885422220738583e-05,
      "loss": 0.004,
      "step": 3100
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.25167518854141235,
      "learning_rate": 1.9877672264048618e-05,
      "loss": 0.0027,
      "step": 3200
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.03997454047203064,
      "learning_rate": 1.9869670323533005e-05,
      "loss": 0.003,
      "step": 3300
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.0028598604258149862,
      "learning_rate": 1.986141660332482e-05,
      "loss": 0.0019,
      "step": 3400
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.1343453824520111,
      "learning_rate": 1.9852911313980146e-05,
      "loss": 0.0034,
      "step": 3500
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.12534217536449432,
      "learning_rate": 1.9844154672472707e-05,
      "loss": 0.0026,
      "step": 3600
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.24615176022052765,
      "learning_rate": 1.9835146902188336e-05,
      "loss": 0.0018,
      "step": 3700
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.013632328249514103,
      "learning_rate": 1.9825888232919268e-05,
      "loss": 0.0019,
      "step": 3800
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.4895065724849701,
      "learning_rate": 1.9816378900858288e-05,
      "loss": 0.0037,
      "step": 3900
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3317159414291382,
      "learning_rate": 1.98066191485927e-05,
      "loss": 0.0022,
      "step": 4000
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.008369619026780128,
      "learning_rate": 1.9796609225098136e-05,
      "loss": 0.0008,
      "step": 4100
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.007471293676644564,
      "learning_rate": 1.9786349385732212e-05,
      "loss": 0.0031,
      "step": 4200
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.011221354827284813,
      "learning_rate": 1.9775839892228004e-05,
      "loss": 0.0013,
      "step": 4300
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.006684939842671156,
      "learning_rate": 1.976508101268738e-05,
      "loss": 0.0024,
      "step": 4400
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.7115832567214966,
      "learning_rate": 1.9754073021574153e-05,
      "loss": 0.0027,
      "step": 4500
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.031222190707921982,
      "learning_rate": 1.9742816199707096e-05,
      "loss": 0.0013,
      "step": 4600
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.033377427607774734,
      "learning_rate": 1.9731310834252747e-05,
      "loss": 0.0046,
      "step": 4700
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3707186281681061,
      "learning_rate": 1.9719557218718116e-05,
      "loss": 0.0025,
      "step": 4800
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.030659371986985207,
      "learning_rate": 1.970755565294318e-05,
      "loss": 0.0033,
      "step": 4900
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.031716570258140564,
      "learning_rate": 1.969530644309323e-05,
      "loss": 0.0017,
      "step": 5000
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.018681248649954796,
      "learning_rate": 1.9682809901651074e-05,
      "loss": 0.0025,
      "step": 5100
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.031807415187358856,
      "learning_rate": 1.9670066347409063e-05,
      "loss": 0.0029,
      "step": 5200
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.00439372006803751,
      "learning_rate": 1.9657076105460945e-05,
      "loss": 0.0024,
      "step": 5300
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2656748592853546,
      "learning_rate": 1.964383950719359e-05,
      "loss": 0.0019,
      "step": 5400
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.18905937671661377,
      "learning_rate": 1.9630356890278527e-05,
      "loss": 0.0015,
      "step": 5500
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1290869414806366,
      "learning_rate": 1.9616628598663322e-05,
      "loss": 0.0034,
      "step": 5600
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.004318041726946831,
      "learning_rate": 1.9602654982562822e-05,
      "loss": 0.0014,
      "step": 5700
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.15103724598884583,
      "learning_rate": 1.9588436398450206e-05,
      "loss": 0.0013,
      "step": 5800
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.039078064262866974,
      "learning_rate": 1.9573973209047893e-05,
      "loss": 0.0015,
      "step": 5900
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.7625222206115723,
      "learning_rate": 1.9559265783318304e-05,
      "loss": 0.0018,
      "step": 6000
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.32574662566185,
      "learning_rate": 1.9544314496454423e-05,
      "loss": 0.0023,
      "step": 6100
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.014155671931803226,
      "learning_rate": 1.9529119729870253e-05,
      "loss": 0.0011,
      "step": 6200
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.007221277803182602,
      "learning_rate": 1.9513681871191063e-05,
      "loss": 0.0033,
      "step": 6300
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.013470535166561604,
      "learning_rate": 1.949800131424352e-05,
      "loss": 0.003,
      "step": 6400
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.003437698120251298,
      "learning_rate": 1.9482078459045617e-05,
      "loss": 0.0012,
      "step": 6500
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.006796371191740036,
      "learning_rate": 1.9465913711796502e-05,
      "loss": 0.001,
      "step": 6600
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.7563895583152771,
      "learning_rate": 1.9449507484866084e-05,
      "loss": 0.0018,
      "step": 6700
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.016867786645889282,
      "learning_rate": 1.9432860196784533e-05,
      "loss": 0.0016,
      "step": 6800
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.006416236516088247,
      "learning_rate": 1.941597227223159e-05,
      "loss": 0.0008,
      "step": 6900
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.0008421270758844912,
      "learning_rate": 1.9398844142025746e-05,
      "loss": 0.0011,
      "step": 7000
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.012384007684886456,
      "learning_rate": 1.9381476243113243e-05,
      "loss": 0.0017,
      "step": 7100
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.004778598435223103,
      "learning_rate": 1.9363869018556928e-05,
      "loss": 0.0016,
      "step": 7200
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.00394839234650135,
      "learning_rate": 1.9346022917524958e-05,
      "loss": 0.0017,
      "step": 7300
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.006818379741162062,
      "learning_rate": 1.9327938395279325e-05,
      "loss": 0.0019,
      "step": 7400
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.028212955221533775,
      "learning_rate": 1.9309615913164262e-05,
      "loss": 0.0013,
      "step": 7500
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.017115216702222824,
      "learning_rate": 1.9291055938594464e-05,
      "loss": 0.0017,
      "step": 7600
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.02033141627907753,
      "learning_rate": 1.9272258945043154e-05,
      "loss": 0.0006,
      "step": 7700
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.2443268299102783,
      "learning_rate": 1.9253225412030028e-05,
      "loss": 0.0011,
      "step": 7800
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.0025956807658076286,
      "learning_rate": 1.9233955825109e-05,
      "loss": 0.001,
      "step": 7900
    },
    {
      "epoch": 0.64,
      "grad_norm": 13.584095001220703,
      "learning_rate": 1.9214450675855832e-05,
      "loss": 0.0029,
      "step": 8000
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.0022025350481271744,
      "learning_rate": 1.919471046185558e-05,
      "loss": 0.001,
      "step": 8100
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.002522263675928116,
      "learning_rate": 1.917473568668991e-05,
      "loss": 0.0024,
      "step": 8200
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.3481580018997192,
      "learning_rate": 1.9154526859924242e-05,
      "loss": 0.0017,
      "step": 8300
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.010974978096783161,
      "learning_rate": 1.9134084497094766e-05,
      "loss": 0.0008,
      "step": 8400
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.005469560623168945,
      "learning_rate": 1.9113409119695276e-05,
      "loss": 0.0012,
      "step": 8500
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.08603531867265701,
      "learning_rate": 1.9092501255163874e-05,
      "loss": 0.0025,
      "step": 8600
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.0022059655748307705,
      "learning_rate": 1.907136143686951e-05,
      "loss": 0.0013,
      "step": 8700
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.0005606790073215961,
      "learning_rate": 1.904999020409837e-05,
      "loss": 0.0014,
      "step": 8800
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.012699578888714314,
      "learning_rate": 1.902838810204015e-05,
      "loss": 0.0012,
      "step": 8900
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.00312532065436244,
      "learning_rate": 1.90065556817741e-05,
      "loss": 0.0012,
      "step": 9000
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.008270219899713993,
      "learning_rate": 1.8984493500255e-05,
      "loss": 0.0014,
      "step": 9100
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.01274858694523573,
      "learning_rate": 1.8962202120298948e-05,
      "loss": 0.0013,
      "step": 9200
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.012512357905507088,
      "learning_rate": 1.8939682110568982e-05,
      "loss": 0.001,
      "step": 9300
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.0043404679745435715,
      "learning_rate": 1.8916934045560603e-05,
      "loss": 0.0023,
      "step": 9400
    },
    {
      "epoch": 0.76,
      "grad_norm": 3.700385332107544,
      "learning_rate": 1.8893958505587093e-05,
      "loss": 0.0031,
      "step": 9500
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.016660619527101517,
      "learning_rate": 1.8870756076764728e-05,
      "loss": 0.0019,
      "step": 9600
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.047528013586997986,
      "learning_rate": 1.8847327350997814e-05,
      "loss": 0.0008,
      "step": 9700
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.01734699122607708,
      "learning_rate": 1.8823672925963598e-05,
      "loss": 0.0022,
      "step": 9800
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.019478172063827515,
      "learning_rate": 1.879979340509701e-05,
      "loss": 0.0009,
      "step": 9900
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.1385781317949295,
      "learning_rate": 1.877568939757529e-05,
      "loss": 0.0015,
      "step": 10000
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.143458753824234,
      "learning_rate": 1.8751361518302413e-05,
      "loss": 0.0008,
      "step": 10100
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.003875449998304248,
      "learning_rate": 1.8726810387893438e-05,
      "loss": 0.0029,
      "step": 10200
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.0030744324903935194,
      "learning_rate": 1.8702036632658646e-05,
      "loss": 0.0012,
      "step": 10300
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.012408553622663021,
      "learning_rate": 1.867704088458759e-05,
      "loss": 0.0017,
      "step": 10400
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.001399656874127686,
      "learning_rate": 1.8651823781332948e-05,
      "loss": 0.0006,
      "step": 10500
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.001192159834317863,
      "learning_rate": 1.8626385966194275e-05,
      "loss": 0.001,
      "step": 10600
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.6697074770927429,
      "learning_rate": 1.8600728088101587e-05,
      "loss": 0.0021,
      "step": 10700
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.007071709726005793,
      "learning_rate": 1.857485080159879e-05,
      "loss": 0.0016,
      "step": 10800
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.005609508138149977,
      "learning_rate": 1.8548754766827016e-05,
      "loss": 0.0022,
      "step": 10900
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.0166325643658638,
      "learning_rate": 1.852244064950775e-05,
      "loss": 0.0032,
      "step": 11000
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.0896034687757492,
      "learning_rate": 1.8495909120925857e-05,
      "loss": 0.0015,
      "step": 11100
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.24882763624191284,
      "learning_rate": 1.846916085791247e-05,
      "loss": 0.0017,
      "step": 11200
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.013551519252359867,
      "learning_rate": 1.8442196542827712e-05,
      "loss": 0.0007,
      "step": 11300
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.01764736883342266,
      "learning_rate": 1.8415016863543286e-05,
      "loss": 0.0005,
      "step": 11400
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.001765413791872561,
      "learning_rate": 1.8387622513424942e-05,
      "loss": 0.0044,
      "step": 11500
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.0014637404819950461,
      "learning_rate": 1.836001419131476e-05,
      "loss": 0.0012,
      "step": 11600
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.005813998635858297,
      "learning_rate": 1.8332192601513358e-05,
      "loss": 0.0006,
      "step": 11700
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.0008558441768400371,
      "learning_rate": 1.8304158453761904e-05,
      "loss": 0.002,
      "step": 11800
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.0033321972005069256,
      "learning_rate": 1.827591246322401e-05,
      "loss": 0.0005,
      "step": 11900
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5678846836090088,
      "learning_rate": 1.8247455350467496e-05,
      "loss": 0.0014,
      "step": 12000
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.009241198189556599,
      "learning_rate": 1.8218787841446003e-05,
      "loss": 0.0004,
      "step": 12100
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.07260189205408096,
      "learning_rate": 1.8189910667480476e-05,
      "loss": 0.0015,
      "step": 12200
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.0022115109022706747,
      "learning_rate": 1.8160824565240495e-05,
      "loss": 0.0029,
      "step": 12300
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.020148636773228645,
      "learning_rate": 1.8131530276725514e-05,
      "loss": 0.0015,
      "step": 12400
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2839891314506531,
      "learning_rate": 1.8102028549245894e-05,
      "loss": 0.0015,
      "step": 12500
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.004473233129829168,
      "learning_rate": 1.8072320135403862e-05,
      "loss": 0.0012,
      "step": 12600
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.0016079695196822286,
      "learning_rate": 1.804240579307431e-05,
      "loss": 0.0009,
      "step": 12700
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.015555166639387608,
      "learning_rate": 1.8012286285385456e-05,
      "loss": 0.0015,
      "step": 12800
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.014301074668765068,
      "learning_rate": 1.7981962380699376e-05,
      "loss": 0.0006,
      "step": 12900
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.005016946699470282,
      "learning_rate": 1.7951434852592406e-05,
      "loss": 0.0008,
      "step": 13000
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.006782655604183674,
      "learning_rate": 1.79207044798354e-05,
      "loss": 0.0004,
      "step": 13100
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.006875937804579735,
      "learning_rate": 1.788977204637388e-05,
      "loss": 0.0027,
      "step": 13200
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.007184536661952734,
      "learning_rate": 1.7858638341308026e-05,
      "loss": 0.0013,
      "step": 13300
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.0019093825249001384,
      "learning_rate": 1.7827304158872538e-05,
      "loss": 0.0014,
      "step": 13400
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.7438774108886719,
      "learning_rate": 1.779577029841638e-05,
      "loss": 0.0027,
      "step": 13500
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.0496426597237587,
      "learning_rate": 1.776403756438241e-05,
      "loss": 0.0016,
      "step": 13600
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.005185109097510576,
      "learning_rate": 1.773210676628682e-05,
      "loss": 0.0004,
      "step": 13700
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.0010710001224651933,
      "learning_rate": 1.769997871869852e-05,
      "loss": 0.0009,
      "step": 13800
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.013586360029876232,
      "learning_rate": 1.7667654241218332e-05,
      "loss": 0.0028,
      "step": 13900
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.021679196506738663,
      "learning_rate": 1.7635134158458095e-05,
      "loss": 0.0011,
      "step": 14000
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.2464137077331543,
      "learning_rate": 1.7602419300019627e-05,
      "loss": 0.0014,
      "step": 14100
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.0015559865860268474,
      "learning_rate": 1.7569510500473566e-05,
      "loss": 0.0003,
      "step": 14200
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.007820005528628826,
      "learning_rate": 1.753640859933806e-05,
      "loss": 0.001,
      "step": 14300
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.0021381524857133627,
      "learning_rate": 1.7503114441057374e-05,
      "loss": 0.0019,
      "step": 14400
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.004830517340451479,
      "learning_rate": 1.746962887498034e-05,
      "loss": 0.0006,
      "step": 14500
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.004048046190291643,
      "learning_rate": 1.743595275533869e-05,
      "loss": 0.001,
      "step": 14600
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.0010604397393763065,
      "learning_rate": 1.7402086941225246e-05,
      "loss": 0.0007,
      "step": 14700
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.10798583179712296,
      "learning_rate": 1.736803229657204e-05,
      "loss": 0.0016,
      "step": 14800
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.00545172905549407,
      "learning_rate": 1.7333789690128252e-05,
      "loss": 0.0003,
      "step": 14900
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.051401443779468536,
      "learning_rate": 1.7299359995438046e-05,
      "loss": 0.0007,
      "step": 15000
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.0022059613838791847,
      "learning_rate": 1.7264744090818284e-05,
      "loss": 0.0009,
      "step": 15100
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.0031678322702646255,
      "learning_rate": 1.7229942859336142e-05,
      "loss": 0.0003,
      "step": 15200
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.4575800597667694,
      "learning_rate": 1.719495718878655e-05,
      "loss": 0.0008,
      "step": 15300
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.0009642325458116829,
      "learning_rate": 1.7159787971669586e-05,
      "loss": 0.001,
      "step": 15400
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.0024611325934529305,
      "learning_rate": 1.712443610516765e-05,
      "loss": 0.0007,
      "step": 15500
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.007828389294445515,
      "learning_rate": 1.7088902491122636e-05,
      "loss": 0.0003,
      "step": 15600
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.0010621993569657207,
      "learning_rate": 1.7053188036012885e-05,
      "loss": 0.0022,
      "step": 15700
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.003037663409486413,
      "learning_rate": 1.7017293650930083e-05,
      "loss": 0.0011,
      "step": 15800
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.021871333941817284,
      "learning_rate": 1.6981220251555996e-05,
      "loss": 0.0005,
      "step": 15900
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.2676248848438263,
      "learning_rate": 1.6944968758139144e-05,
      "loss": 0.0012,
      "step": 16000
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.002139603951945901,
      "learning_rate": 1.6908540095471288e-05,
      "loss": 0.0016,
      "step": 16100
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.00880183931440115,
      "learning_rate": 1.6871935192863862e-05,
      "loss": 0.0007,
      "step": 16200
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.006027602590620518,
      "learning_rate": 1.6835154984124266e-05,
      "loss": 0.0027,
      "step": 16300
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.002160316100344062,
      "learning_rate": 1.6798200407532025e-05,
      "loss": 0.0004,
      "step": 16400
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.004634434822946787,
      "learning_rate": 1.676107240581488e-05,
      "loss": 0.0005,
      "step": 16500
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.013142119161784649,
      "learning_rate": 1.6723771926124704e-05,
      "loss": 0.0012,
      "step": 16600
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.0017228337237611413,
      "learning_rate": 1.6686299920013388e-05,
      "loss": 0.001,
      "step": 16700
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.011891527101397514,
      "learning_rate": 1.6648657343408517e-05,
      "loss": 0.0004,
      "step": 16800
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.03485196456313133,
      "learning_rate": 1.661084515658901e-05,
      "loss": 0.0008,
      "step": 16900
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.006045693065971136,
      "learning_rate": 1.6572864324160617e-05,
      "loss": 0.0016,
      "step": 17000
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.0319632813334465,
      "learning_rate": 1.6534715815031325e-05,
      "loss": 0.001,
      "step": 17100
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.000651439419016242,
      "learning_rate": 1.649640060238661e-05,
      "loss": 0.0008,
      "step": 17200
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.013289058580994606,
      "learning_rate": 1.645791966366464e-05,
      "loss": 0.0004,
      "step": 17300
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.1903555393218994,
      "learning_rate": 1.6419273980531333e-05,
      "loss": 0.0009,
      "step": 17400
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.0004181715485174209,
      "learning_rate": 1.63804645388553e-05,
      "loss": 0.0008,
      "step": 17500
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.0005597823183052242,
      "learning_rate": 1.6341492328682703e-05,
      "loss": 0.0015,
      "step": 17600
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.003250558627769351,
      "learning_rate": 1.6302358344212025e-05,
      "loss": 0.0015,
      "step": 17700
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.00633718678727746,
      "learning_rate": 1.6263063583768652e-05,
      "loss": 0.0007,
      "step": 17800
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.0012848949991166592,
      "learning_rate": 1.622360904977946e-05,
      "loss": 0.001,
      "step": 17900
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.03510229289531708,
      "learning_rate": 1.6183995748747204e-05,
      "loss": 0.001,
      "step": 18000
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.00044461427023634315,
      "learning_rate": 1.6144224691224868e-05,
      "loss": 0.0005,
      "step": 18100
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.056695085018873215,
      "learning_rate": 1.6104296891789867e-05,
      "loss": 0.0011,
      "step": 18200
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.002499540336430073,
      "learning_rate": 1.606421336901818e-05,
      "loss": 0.0005,
      "step": 18300
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.00433159526437521,
      "learning_rate": 1.6023975145458352e-05,
      "loss": 0.0007,
      "step": 18400
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.0020198116544634104,
      "learning_rate": 1.5983583247605414e-05,
      "loss": 0.0005,
      "step": 18500
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.004110960755497217,
      "learning_rate": 1.5943038705874697e-05,
      "loss": 0.001,
      "step": 18600
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.0036564720794558525,
      "learning_rate": 1.590234255457555e-05,
      "loss": 0.0014,
      "step": 18700
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.0010110210860148072,
      "learning_rate": 1.5861495831884942e-05,
      "loss": 0.0008,
      "step": 18800
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.11147645115852356,
      "learning_rate": 1.582049957982099e-05,
      "loss": 0.0004,
      "step": 18900
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.002563257934525609,
      "learning_rate": 1.5779354844216377e-05,
      "loss": 0.0003,
      "step": 19000
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.002472821157425642,
      "learning_rate": 1.5738062674691657e-05,
      "loss": 0.0003,
      "step": 19100
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.08172155171632767,
      "learning_rate": 1.5696624124628495e-05,
      "loss": 0.0005,
      "step": 19200
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.009228968061506748,
      "learning_rate": 1.5655040251142787e-05,
      "loss": 0.0008,
      "step": 19300
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.027170058339834213,
      "learning_rate": 1.5613312115057697e-05,
      "loss": 0.0002,
      "step": 19400
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.003953781444579363,
      "learning_rate": 1.5571440780876588e-05,
      "loss": 0.0009,
      "step": 19500
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.2073652297258377,
      "learning_rate": 1.5529427316755876e-05,
      "loss": 0.001,
      "step": 19600
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.000581290340051055,
      "learning_rate": 1.548727279447777e-05,
      "loss": 0.0007,
      "step": 19700
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.001082689268514514,
      "learning_rate": 1.5444978289422937e-05,
      "loss": 0.0011,
      "step": 19800
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.020316652953624725,
      "learning_rate": 1.540254488054307e-05,
      "loss": 0.0007,
      "step": 19900
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.0026674780528992414,
      "learning_rate": 1.5359973650333352e-05,
      "loss": 0.0006,
      "step": 20000
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.007332425098866224,
      "learning_rate": 1.5317265684804865e-05,
      "loss": 0.001,
      "step": 20100
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.0030178299639374018,
      "learning_rate": 1.5274422073456853e-05,
      "loss": 0.0002,
      "step": 20200
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.0009229824645444751,
      "learning_rate": 1.5231443909248956e-05,
      "loss": 0.0006,
      "step": 20300
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.000619510596152395,
      "learning_rate": 1.5188332288573313e-05,
      "loss": 0.0011,
      "step": 20400
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.0273954626172781,
      "learning_rate": 1.5145088311226599e-05,
      "loss": 0.0004,
      "step": 20500
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.009902927093207836,
      "learning_rate": 1.510171308038197e-05,
      "loss": 0.0006,
      "step": 20600
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.012104487977921963,
      "learning_rate": 1.5058207702560907e-05,
      "loss": 0.0004,
      "step": 20700
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.8639243245124817,
      "learning_rate": 1.501457328760501e-05,
      "loss": 0.0007,
      "step": 20800
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.07374394685029984,
      "learning_rate": 1.4970810948647664e-05,
      "loss": 0.0007,
      "step": 20900
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.0010848238598555326,
      "learning_rate": 1.4926921802085662e-05,
      "loss": 0.0008,
      "step": 21000
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.03615230694413185,
      "learning_rate": 1.4882906967550708e-05,
      "loss": 0.0002,
      "step": 21100
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.00893151480704546,
      "learning_rate": 1.4838767567880865e-05,
      "loss": 0.0012,
      "step": 21200
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.0023355227895081043,
      "learning_rate": 1.479450472909191e-05,
      "loss": 0.0023,
      "step": 21300
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.7856515645980835,
      "learning_rate": 1.4750119580348601e-05,
      "loss": 0.0008,
      "step": 21400
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.0007943073869682848,
      "learning_rate": 1.4705613253935886e-05,
      "loss": 0.0002,
      "step": 21500
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.000991353183053434,
      "learning_rate": 1.4660986885230002e-05,
      "loss": 0.0005,
      "step": 21600
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.0023457545321434736,
      "learning_rate": 1.4616241612669523e-05,
      "loss": 0.0035,
      "step": 21700
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.0007867334061302245,
      "learning_rate": 1.4571378577726317e-05,
      "loss": 0.0006,
      "step": 21800
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.0013897059252485633,
      "learning_rate": 1.4526398924876407e-05,
      "loss": 0.0005,
      "step": 21900
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.0020729138050228357,
      "learning_rate": 1.4481303801570805e-05,
      "loss": 0.0009,
      "step": 22000
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.002158819232136011,
      "learning_rate": 1.4436094358206224e-05,
      "loss": 0.0005,
      "step": 22100
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.0038599702529609203,
      "learning_rate": 1.4390771748095735e-05,
      "loss": 0.0007,
      "step": 22200
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.0022886607330292463,
      "learning_rate": 1.4345337127439333e-05,
      "loss": 0.0002,
      "step": 22300
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.011882874183356762,
      "learning_rate": 1.4299791655294461e-05,
      "loss": 0.0021,
      "step": 22400
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.011625411920249462,
      "learning_rate": 1.4254136493546432e-05,
      "loss": 0.0003,
      "step": 22500
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.004132369067519903,
      "learning_rate": 1.4208372806878782e-05,
      "loss": 0.0004,
      "step": 22600
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.0008625888731330633,
      "learning_rate": 1.4162501762743579e-05,
      "loss": 0.0013,
      "step": 22700
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.02038603462278843,
      "learning_rate": 1.4116524531331616e-05,
      "loss": 0.001,
      "step": 22800
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.021469444036483765,
      "learning_rate": 1.4070442285542579e-05,
      "loss": 0.0004,
      "step": 22900
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.7221536040306091,
      "learning_rate": 1.402425620095511e-05,
      "loss": 0.0007,
      "step": 23000
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.0006885859766043723,
      "learning_rate": 1.3977967455796828e-05,
      "loss": 0.0009,
      "step": 23100
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.0010358589934185147,
      "learning_rate": 1.393157723091428e-05,
      "loss": 0.0003,
      "step": 23200
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.0008139715064316988,
      "learning_rate": 1.3885086709742788e-05,
      "loss": 0.0005,
      "step": 23300
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.6740529537200928,
      "learning_rate": 1.3838497078276288e-05,
      "loss": 0.0018,
      "step": 23400
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.0022279045078903437,
      "learning_rate": 1.3791809525037057e-05,
      "loss": 0.0005,
      "step": 23500
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.0020128132309764624,
      "learning_rate": 1.3745025241045414e-05,
      "loss": 0.0002,
      "step": 23600
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.000531968311406672,
      "learning_rate": 1.3698145419789302e-05,
      "loss": 0.0007,
      "step": 23700
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.00039000410470180213,
      "learning_rate": 1.3651171257193883e-05,
      "loss": 0.0006,
      "step": 23800
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.041364409029483795,
      "learning_rate": 1.3604103951590993e-05,
      "loss": 0.0003,
      "step": 23900
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.0011203879257664084,
      "learning_rate": 1.3556944703688592e-05,
      "loss": 0.0003,
      "step": 24000
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.001194770447909832,
      "learning_rate": 1.3509694716540135e-05,
      "loss": 0.0002,
      "step": 24100
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.002378718461841345,
      "learning_rate": 1.3462355195513868e-05,
      "loss": 0.0006,
      "step": 24200
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.32328903675079346,
      "learning_rate": 1.341492734826209e-05,
      "loss": 0.001,
      "step": 24300
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.002817463595420122,
      "learning_rate": 1.3367412384690346e-05,
      "loss": 0.0017,
      "step": 24400
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.0024801292456686497,
      "learning_rate": 1.3319811516926541e-05,
      "loss": 0.0005,
      "step": 24500
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.004569421522319317,
      "learning_rate": 1.3272125959290059e-05,
      "loss": 0.0008,
      "step": 24600
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.1658785045146942,
      "learning_rate": 1.3224356928260735e-05,
      "loss": 0.0005,
      "step": 24700
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.001219902653247118,
      "learning_rate": 1.317650564244787e-05,
      "loss": 0.0005,
      "step": 24800
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.002142696175724268,
      "learning_rate": 1.3128573322559097e-05,
      "loss": 0.0022,
      "step": 24900
    }
  ],
  "logging_steps": 100,
  "max_steps": 62300,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 1.706415322300416e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}