{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.2886961298148445,
  "eval_steps": 500000,
  "global_step": 100000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011443480649074222,
      "grad_norm": 603.6091918945312,
      "learning_rate": 1.1666666666666667e-07,
      "loss": 10.731,
      "step": 500
    },
    {
      "epoch": 0.022886961298148444,
      "grad_norm": 718.9038696289062,
      "learning_rate": 2.3333333333333333e-07,
      "loss": 9.2432,
      "step": 1000
    },
    {
      "epoch": 0.034330441947222665,
      "grad_norm": 38.804962158203125,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 9.1395,
      "step": 1500
    },
    {
      "epoch": 0.04577392259629689,
      "grad_norm": 619.4507446289062,
      "learning_rate": 4.6666666666666666e-07,
      "loss": 11.2638,
      "step": 2000
    },
    {
      "epoch": 0.05721740324537111,
      "grad_norm": 0.11708039790391922,
      "learning_rate": 5.833333333333333e-07,
      "loss": 9.1618,
      "step": 2500
    },
    {
      "epoch": 0.06866088389444533,
      "grad_norm": 206.2231903076172,
      "learning_rate": 7.000000000000001e-07,
      "loss": 12.4357,
      "step": 3000
    },
    {
      "epoch": 0.08010436454351956,
      "grad_norm": 26.35240936279297,
      "learning_rate": 8.166666666666666e-07,
      "loss": 9.4019,
      "step": 3500
    },
    {
      "epoch": 0.09154784519259378,
      "grad_norm": 390.7612609863281,
      "learning_rate": 9.333333333333333e-07,
      "loss": 10.9883,
      "step": 4000
    },
    {
      "epoch": 0.10299132584166801,
      "grad_norm": 560.8733520507812,
      "learning_rate": 1.05e-06,
      "loss": 11.4073,
      "step": 4500
    },
    {
      "epoch": 0.11443480649074222,
      "grad_norm": 733.6063232421875,
      "learning_rate": 1.1666666666666666e-06,
      "loss": 9.4747,
      "step": 5000
    },
    {
      "epoch": 0.12587828713981644,
      "grad_norm": 176.73806762695312,
      "learning_rate": 1.2833333333333333e-06,
      "loss": 11.2038,
      "step": 5500
    },
    {
      "epoch": 0.13732176778889066,
      "grad_norm": 40.61717224121094,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 11.3759,
      "step": 6000
    },
    {
      "epoch": 0.1487652484379649,
      "grad_norm": 2191.57470703125,
      "learning_rate": 1.5166666666666666e-06,
      "loss": 10.7893,
      "step": 6500
    },
    {
      "epoch": 0.16020872908703912,
      "grad_norm": 10.336786270141602,
      "learning_rate": 1.6333333333333333e-06,
      "loss": 11.6208,
      "step": 7000
    },
    {
      "epoch": 0.17165220973611334,
      "grad_norm": 0.11516555398702621,
      "learning_rate": 1.75e-06,
      "loss": 10.8105,
      "step": 7500
    },
    {
      "epoch": 0.18309569038518755,
      "grad_norm": 0.2883591651916504,
      "learning_rate": 1.8666666666666667e-06,
      "loss": 11.147,
      "step": 8000
    },
    {
      "epoch": 0.19453917103426177,
      "grad_norm": 862.2401123046875,
      "learning_rate": 1.983333333333333e-06,
      "loss": 10.8087,
      "step": 8500
    },
    {
      "epoch": 0.20598265168333602,
      "grad_norm": 0.004448544699698687,
      "learning_rate": 2.1e-06,
      "loss": 11.3934,
      "step": 9000
    },
    {
      "epoch": 0.21742613233241023,
      "grad_norm": 0.008581358008086681,
      "learning_rate": 2.2166666666666665e-06,
      "loss": 13.5201,
      "step": 9500
    },
    {
      "epoch": 0.22886961298148445,
      "grad_norm": 0.170095756649971,
      "learning_rate": 2.333333333333333e-06,
      "loss": 10.7812,
      "step": 10000
    },
    {
      "epoch": 0.24031309363055867,
      "grad_norm": 0.8723887801170349,
      "learning_rate": 2.45e-06,
      "loss": 10.8569,
      "step": 10500
    },
    {
      "epoch": 0.2517565742796329,
      "grad_norm": 1424.5357666015625,
      "learning_rate": 2.5666666666666665e-06,
      "loss": 13.1824,
      "step": 11000
    },
    {
      "epoch": 0.26320005492870713,
      "grad_norm": 2.006991386413574,
      "learning_rate": 2.6833333333333336e-06,
      "loss": 12.9087,
      "step": 11500
    },
    {
      "epoch": 0.2746435355777813,
      "grad_norm": 492.7056884765625,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 11.9267,
      "step": 12000
    },
    {
      "epoch": 0.28608701622685556,
      "grad_norm": 6.009059906005859,
      "learning_rate": 2.9166666666666666e-06,
      "loss": 12.8387,
      "step": 12500
    },
    {
      "epoch": 0.2975304968759298,
      "grad_norm": 1738.22900390625,
      "learning_rate": 3.0333333333333332e-06,
      "loss": 13.1134,
      "step": 13000
    },
    {
      "epoch": 0.308973977525004,
      "grad_norm": 1644.960205078125,
      "learning_rate": 3.15e-06,
      "loss": 10.5768,
      "step": 13500
    },
    {
      "epoch": 0.32041745817407824,
      "grad_norm": 2049.489501953125,
      "learning_rate": 3.2666666666666666e-06,
      "loss": 12.9949,
      "step": 14000
    },
    {
      "epoch": 0.33186093882315243,
      "grad_norm": 2.0113472938537598,
      "learning_rate": 3.3833333333333333e-06,
      "loss": 12.5091,
      "step": 14500
    },
    {
      "epoch": 0.3433044194722267,
      "grad_norm": 767.9404296875,
      "learning_rate": 3.5e-06,
      "loss": 12.5176,
      "step": 15000
    },
    {
      "epoch": 0.3547479001213009,
      "grad_norm": 342.45513916015625,
      "learning_rate": 3.616666666666667e-06,
      "loss": 12.6625,
      "step": 15500
    },
    {
      "epoch": 0.3661913807703751,
      "grad_norm": 771.4798583984375,
      "learning_rate": 3.7333333333333333e-06,
      "loss": 14.2435,
      "step": 16000
    },
    {
      "epoch": 0.37763486141944935,
      "grad_norm": 4853.9267578125,
      "learning_rate": 3.85e-06,
      "loss": 15.1053,
      "step": 16500
    },
    {
      "epoch": 0.38907834206852354,
      "grad_norm": 0.0017908277222886682,
      "learning_rate": 3.966666666666666e-06,
      "loss": 12.6706,
      "step": 17000
    },
    {
      "epoch": 0.4005218227175978,
      "grad_norm": 0.2733666002750397,
      "learning_rate": 4.083333333333334e-06,
      "loss": 12.4815,
      "step": 17500
    },
    {
      "epoch": 0.41196530336667203,
      "grad_norm": 0.0,
      "learning_rate": 4.2e-06,
      "loss": 12.5318,
      "step": 18000
    },
    {
      "epoch": 0.4234087840157462,
      "grad_norm": 1851.682861328125,
      "learning_rate": 4.316666666666667e-06,
      "loss": 13.7826,
      "step": 18500
    },
    {
      "epoch": 0.43485226466482046,
      "grad_norm": 157.1131591796875,
      "learning_rate": 4.433333333333333e-06,
      "loss": 12.2737,
      "step": 19000
    },
    {
      "epoch": 0.44629574531389465,
      "grad_norm": 1271.1109619140625,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 10.5671,
      "step": 19500
    },
    {
      "epoch": 0.4577392259629689,
      "grad_norm": 480.1098937988281,
      "learning_rate": 4.666666666666666e-06,
      "loss": 13.5673,
      "step": 20000
    },
    {
      "epoch": 0.46918270661204314,
      "grad_norm": 1566.4591064453125,
      "learning_rate": 4.783333333333333e-06,
      "loss": 14.4705,
      "step": 20500
    },
    {
      "epoch": 0.48062618726111733,
      "grad_norm": 0.0007110056467354298,
      "learning_rate": 4.9e-06,
      "loss": 11.6491,
      "step": 21000
    },
    {
      "epoch": 0.4920696679101916,
      "grad_norm": 4.593759059906006,
      "learning_rate": 5.016666666666666e-06,
      "loss": 12.8944,
      "step": 21500
    },
    {
      "epoch": 0.5035131485592658,
      "grad_norm": 0.0,
      "learning_rate": 5.133333333333333e-06,
      "loss": 13.6318,
      "step": 22000
    },
    {
      "epoch": 0.51495662920834,
      "grad_norm": 132.595458984375,
      "learning_rate": 5.25e-06,
      "loss": 12.8382,
      "step": 22500
    },
    {
      "epoch": 0.5264001098574143,
      "grad_norm": 347.5989074707031,
      "learning_rate": 5.366666666666667e-06,
      "loss": 11.8936,
      "step": 23000
    },
    {
      "epoch": 0.5378435905064884,
      "grad_norm": 0.009264913387596607,
      "learning_rate": 5.483333333333333e-06,
      "loss": 14.0151,
      "step": 23500
    },
    {
      "epoch": 0.5492870711555626,
      "grad_norm": 245.79705810546875,
      "learning_rate": 5.600000000000001e-06,
      "loss": 11.9611,
      "step": 24000
    },
    {
      "epoch": 0.5607305518046369,
      "grad_norm": 1345.8895263671875,
      "learning_rate": 5.7166666666666664e-06,
      "loss": 13.2515,
      "step": 24500
    },
    {
      "epoch": 0.5721740324537111,
      "grad_norm": 0.26148831844329834,
      "learning_rate": 5.833333333333333e-06,
      "loss": 13.5492,
      "step": 25000
    },
    {
      "epoch": 0.5836175131027853,
      "grad_norm": 4638.9169921875,
      "learning_rate": 5.95e-06,
      "loss": 14.1596,
      "step": 25500
    },
    {
      "epoch": 0.5950609937518596,
      "grad_norm": 184.8077392578125,
      "learning_rate": 6.0666666666666665e-06,
      "loss": 11.8104,
      "step": 26000
    },
    {
      "epoch": 0.6065044744009338,
      "grad_norm": 140.49444580078125,
      "learning_rate": 6.183333333333333e-06,
      "loss": 11.2624,
      "step": 26500
    },
    {
      "epoch": 0.617947955050008,
      "grad_norm": 0.0011962183052673936,
      "learning_rate": 6.3e-06,
      "loss": 12.7681,
      "step": 27000
    },
    {
      "epoch": 0.6293914356990822,
      "grad_norm": 0.057224027812480927,
      "learning_rate": 6.4166666666666665e-06,
      "loss": 15.7726,
      "step": 27500
    },
    {
      "epoch": 0.6408349163481565,
      "grad_norm": 0.04104612395167351,
      "learning_rate": 6.533333333333333e-06,
      "loss": 15.2471,
      "step": 28000
    },
    {
      "epoch": 0.6522783969972307,
      "grad_norm": 0.003121469169855118,
      "learning_rate": 6.65e-06,
      "loss": 16.5911,
      "step": 28500
    },
    {
      "epoch": 0.6637218776463049,
      "grad_norm": 0.0,
      "learning_rate": 6.7666666666666665e-06,
      "loss": 11.7935,
      "step": 29000
    },
    {
      "epoch": 0.6751653582953792,
      "grad_norm": 1068.4923095703125,
      "learning_rate": 6.883333333333333e-06,
      "loss": 12.6445,
      "step": 29500
    },
    {
      "epoch": 0.6866088389444533,
      "grad_norm": 49.82635498046875,
      "learning_rate": 7e-06,
      "loss": 13.4921,
      "step": 30000
    },
    {
      "epoch": 0.6980523195935275,
      "grad_norm": 0.021331172436475754,
      "learning_rate": 6.987037037037037e-06,
      "loss": 13.8913,
      "step": 30500
    },
    {
      "epoch": 0.7094958002426018,
      "grad_norm": 390.8102111816406,
      "learning_rate": 6.974074074074074e-06,
      "loss": 14.1043,
      "step": 31000
    },
    {
      "epoch": 0.720939280891676,
      "grad_norm": 464.5948791503906,
      "learning_rate": 6.9611111111111116e-06,
      "loss": 14.1976,
      "step": 31500
    },
    {
      "epoch": 0.7323827615407502,
      "grad_norm": 383.39373779296875,
      "learning_rate": 6.948148148148148e-06,
      "loss": 13.338,
      "step": 32000
    },
    {
      "epoch": 0.7438262421898244,
      "grad_norm": 0.005027694161981344,
      "learning_rate": 6.935185185185185e-06,
      "loss": 13.6854,
      "step": 32500
    },
    {
      "epoch": 0.7552697228388987,
      "grad_norm": 1620.7523193359375,
      "learning_rate": 6.922222222222222e-06,
      "loss": 11.7295,
      "step": 33000
    },
    {
      "epoch": 0.7667132034879729,
      "grad_norm": 959.1514282226562,
      "learning_rate": 6.90925925925926e-06,
      "loss": 14.5347,
      "step": 33500
    },
    {
      "epoch": 0.7781566841370471,
      "grad_norm": 778.1695556640625,
      "learning_rate": 6.896296296296296e-06,
      "loss": 12.9294,
      "step": 34000
    },
    {
      "epoch": 0.7896001647861214,
      "grad_norm": 18.240928649902344,
      "learning_rate": 6.883333333333333e-06,
      "loss": 12.5578,
      "step": 34500
    },
    {
      "epoch": 0.8010436454351956,
      "grad_norm": 0.060547858476638794,
      "learning_rate": 6.8703703703703704e-06,
      "loss": 11.7427,
      "step": 35000
    },
    {
      "epoch": 0.8124871260842698,
      "grad_norm": 0.03452019393444061,
      "learning_rate": 6.857407407407408e-06,
      "loss": 13.3464,
      "step": 35500
    },
    {
      "epoch": 0.8239306067333441,
      "grad_norm": 0.0,
      "learning_rate": 6.844444444444444e-06,
      "loss": 12.3813,
      "step": 36000
    },
    {
      "epoch": 0.8353740873824183,
      "grad_norm": 0.01644955575466156,
      "learning_rate": 6.831481481481482e-06,
      "loss": 15.1999,
      "step": 36500
    },
    {
      "epoch": 0.8468175680314924,
      "grad_norm": 0.041852571070194244,
      "learning_rate": 6.8185185185185185e-06,
      "loss": 14.611,
      "step": 37000
    },
    {
      "epoch": 0.8582610486805666,
      "grad_norm": 0.03489886596798897,
      "learning_rate": 6.805555555555556e-06,
      "loss": 12.7151,
      "step": 37500
    },
    {
      "epoch": 0.8697045293296409,
      "grad_norm": 1784.8714599609375,
      "learning_rate": 6.792592592592592e-06,
      "loss": 13.324,
      "step": 38000
    },
    {
      "epoch": 0.8811480099787151,
      "grad_norm": 325.86151123046875,
      "learning_rate": 6.77962962962963e-06,
      "loss": 13.8547,
      "step": 38500
    },
    {
      "epoch": 0.8925914906277893,
      "grad_norm": 209.41114807128906,
      "learning_rate": 6.7666666666666665e-06,
      "loss": 13.9472,
      "step": 39000
    },
    {
      "epoch": 0.9040349712768636,
      "grad_norm": 0.0,
      "learning_rate": 6.753703703703704e-06,
      "loss": 14.3618,
      "step": 39500
    },
    {
      "epoch": 0.9154784519259378,
      "grad_norm": 1665.116943359375,
      "learning_rate": 6.74074074074074e-06,
      "loss": 15.8669,
      "step": 40000
    },
    {
      "epoch": 0.926921932575012,
      "grad_norm": 0.0066319056786596775,
      "learning_rate": 6.727777777777778e-06,
      "loss": 14.8973,
      "step": 40500
    },
    {
      "epoch": 0.9383654132240863,
      "grad_norm": 738.2325439453125,
      "learning_rate": 6.714814814814815e-06,
      "loss": 13.7844,
      "step": 41000
    },
    {
      "epoch": 0.9498088938731605,
      "grad_norm": 276.5779113769531,
      "learning_rate": 6.701851851851852e-06,
      "loss": 12.9653,
      "step": 41500
    },
    {
      "epoch": 0.9612523745222347,
      "grad_norm": 1572.8726806640625,
      "learning_rate": 6.688888888888889e-06,
      "loss": 12.8457,
      "step": 42000
    },
    {
      "epoch": 0.9726958551713089,
      "grad_norm": 1.0475573539733887,
      "learning_rate": 6.675925925925926e-06,
      "loss": 12.2727,
      "step": 42500
    },
    {
      "epoch": 0.9841393358203832,
      "grad_norm": 12.392683029174805,
      "learning_rate": 6.662962962962963e-06,
      "loss": 14.6495,
      "step": 43000
    },
    {
      "epoch": 0.9955828164694573,
      "grad_norm": 0.0,
      "learning_rate": 6.65e-06,
      "loss": 13.6766,
      "step": 43500
    },
    {
      "epoch": 1.0070262971185315,
      "grad_norm": 178.05191040039062,
      "learning_rate": 6.637037037037037e-06,
      "loss": 12.0881,
      "step": 44000
    },
    {
      "epoch": 1.0184697777676057,
      "grad_norm": 287.11749267578125,
      "learning_rate": 6.624074074074074e-06,
      "loss": 10.3373,
      "step": 44500
    },
    {
      "epoch": 1.02991325841668,
      "grad_norm": 229.68655395507812,
      "learning_rate": 6.611111111111111e-06,
      "loss": 12.0917,
      "step": 45000
    },
    {
      "epoch": 1.0413567390657543,
      "grad_norm": 0.005943186115473509,
      "learning_rate": 6.598148148148148e-06,
      "loss": 11.5574,
      "step": 45500
    },
    {
      "epoch": 1.0528002197148285,
      "grad_norm": 0.00034078938188031316,
      "learning_rate": 6.585185185185185e-06,
      "loss": 13.1375,
      "step": 46000
    },
    {
      "epoch": 1.0642437003639027,
      "grad_norm": 0.3885180354118347,
      "learning_rate": 6.572222222222222e-06,
      "loss": 11.2181,
      "step": 46500
    },
    {
      "epoch": 1.075687181012977,
      "grad_norm": 1127.3404541015625,
      "learning_rate": 6.559259259259259e-06,
      "loss": 12.5768,
      "step": 47000
    },
    {
      "epoch": 1.087130661662051,
      "grad_norm": 1220.1749267578125,
      "learning_rate": 6.546296296296297e-06,
      "loss": 13.4144,
      "step": 47500
    },
    {
      "epoch": 1.0985741423111253,
      "grad_norm": 0.0005302530480548739,
      "learning_rate": 6.533333333333333e-06,
      "loss": 11.0923,
      "step": 48000
    },
    {
      "epoch": 1.1100176229601995,
      "grad_norm": 661.6603393554688,
      "learning_rate": 6.52037037037037e-06,
      "loss": 11.0951,
      "step": 48500
    },
    {
      "epoch": 1.1214611036092739,
      "grad_norm": 0.08705360442399979,
      "learning_rate": 6.507407407407407e-06,
      "loss": 11.5883,
      "step": 49000
    },
    {
      "epoch": 1.132904584258348,
      "grad_norm": 171.4398651123047,
      "learning_rate": 6.494444444444445e-06,
      "loss": 13.5062,
      "step": 49500
    },
    {
      "epoch": 1.1443480649074222,
      "grad_norm": 0.04586634412407875,
      "learning_rate": 6.481481481481481e-06,
      "loss": 12.4203,
      "step": 50000
    },
    {
      "epoch": 1.1557915455564964,
      "grad_norm": 0.0418708510696888,
      "learning_rate": 6.4685185185185185e-06,
      "loss": 12.8418,
      "step": 50500
    },
    {
      "epoch": 1.1672350262055706,
      "grad_norm": 635.2694091796875,
      "learning_rate": 6.455555555555556e-06,
      "loss": 11.7328,
      "step": 51000
    },
    {
      "epoch": 1.1786785068546448,
      "grad_norm": 216.6066131591797,
      "learning_rate": 6.442592592592593e-06,
      "loss": 13.6954,
      "step": 51500
    },
    {
      "epoch": 1.1901219875037192,
      "grad_norm": 120.59862518310547,
      "learning_rate": 6.429629629629629e-06,
      "loss": 11.0234,
      "step": 52000
    },
    {
      "epoch": 1.2015654681527934,
      "grad_norm": 647.171875,
      "learning_rate": 6.4166666666666665e-06,
      "loss": 14.4181,
      "step": 52500
    },
    {
      "epoch": 1.2130089488018676,
      "grad_norm": 0.8411855697631836,
      "learning_rate": 6.403703703703704e-06,
      "loss": 13.5984,
      "step": 53000
    },
    {
      "epoch": 1.2244524294509418,
      "grad_norm": 0.0012393624056130648,
      "learning_rate": 6.390740740740741e-06,
      "loss": 12.1035,
      "step": 53500
    },
    {
      "epoch": 1.235895910100016,
      "grad_norm": 0.0,
      "learning_rate": 6.377777777777777e-06,
      "loss": 12.8851,
      "step": 54000
    },
    {
      "epoch": 1.2473393907490902,
      "grad_norm": 0.0,
      "learning_rate": 6.3648148148148145e-06,
      "loss": 15.1293,
      "step": 54500
    },
    {
      "epoch": 1.2587828713981644,
      "grad_norm": 2496.68505859375,
      "learning_rate": 6.351851851851852e-06,
      "loss": 11.7267,
      "step": 55000
    },
    {
      "epoch": 1.2702263520472385,
      "grad_norm": 603.0614013671875,
      "learning_rate": 6.338888888888889e-06,
      "loss": 12.6656,
      "step": 55500
    },
    {
      "epoch": 1.281669832696313,
      "grad_norm": 0.1624222695827484,
      "learning_rate": 6.325925925925925e-06,
      "loss": 12.3149,
      "step": 56000
    },
    {
      "epoch": 1.2931133133453871,
      "grad_norm": 0.0846194475889206,
      "learning_rate": 6.3129629629629634e-06,
      "loss": 14.1863,
      "step": 56500
    },
    {
      "epoch": 1.3045567939944613,
      "grad_norm": 0.0,
      "learning_rate": 6.3e-06,
      "loss": 13.3858,
      "step": 57000
    },
    {
      "epoch": 1.3160002746435355,
      "grad_norm": 0.0017423485405743122,
      "learning_rate": 6.287037037037037e-06,
      "loss": 10.2554,
      "step": 57500
    },
    {
      "epoch": 1.3274437552926097,
      "grad_norm": 1450.389892578125,
      "learning_rate": 6.274074074074073e-06,
      "loss": 12.7777,
      "step": 58000
    },
    {
      "epoch": 1.3388872359416841,
      "grad_norm": 0.043471284210681915,
      "learning_rate": 6.2611111111111115e-06,
      "loss": 11.1225,
      "step": 58500
    },
    {
      "epoch": 1.3503307165907583,
      "grad_norm": 0.0038315041456371546,
      "learning_rate": 6.248148148148148e-06,
      "loss": 10.2773,
      "step": 59000
    },
    {
      "epoch": 1.3617741972398325,
      "grad_norm": 2073.392578125,
      "learning_rate": 6.235185185185185e-06,
      "loss": 13.5407,
      "step": 59500
    },
    {
      "epoch": 1.3732176778889067,
      "grad_norm": 4.363699913024902,
      "learning_rate": 6.2222222222222215e-06,
      "loss": 14.8187,
      "step": 60000
    },
    {
      "epoch": 1.3846611585379809,
      "grad_norm": 0.010874895378947258,
      "learning_rate": 6.2092592592592595e-06,
      "loss": 11.7766,
      "step": 60500
    },
    {
      "epoch": 1.396104639187055,
      "grad_norm": 5.799459457397461,
      "learning_rate": 6.196296296296296e-06,
      "loss": 12.6413,
      "step": 61000
    },
    {
      "epoch": 1.4075481198361293,
      "grad_norm": 0.0036441339179873466,
      "learning_rate": 6.183333333333333e-06,
      "loss": 11.2557,
      "step": 61500
    },
    {
      "epoch": 1.4189916004852035,
      "grad_norm": 1275.374755859375,
      "learning_rate": 6.17037037037037e-06,
      "loss": 11.7272,
      "step": 62000
    },
    {
      "epoch": 1.4304350811342779,
      "grad_norm": 4855.68212890625,
      "learning_rate": 6.157407407407408e-06,
      "loss": 12.1518,
      "step": 62500
    },
    {
      "epoch": 1.441878561783352,
      "grad_norm": 523.7196655273438,
      "learning_rate": 6.144444444444444e-06,
      "loss": 12.6978,
      "step": 63000
    },
    {
      "epoch": 1.4533220424324262,
      "grad_norm": 1098.2850341796875,
      "learning_rate": 6.131481481481481e-06,
      "loss": 11.8028,
      "step": 63500
    },
    {
      "epoch": 1.4647655230815004,
      "grad_norm": 1153.9100341796875,
      "learning_rate": 6.118518518518518e-06,
      "loss": 11.3937,
      "step": 64000
    },
    {
      "epoch": 1.4762090037305746,
      "grad_norm": 1573.587890625,
      "learning_rate": 6.105555555555556e-06,
      "loss": 13.9497,
      "step": 64500
    },
    {
      "epoch": 1.487652484379649,
      "grad_norm": 908.9359130859375,
      "learning_rate": 6.092592592592592e-06,
      "loss": 11.6579,
      "step": 65000
    },
    {
      "epoch": 1.4990959650287232,
      "grad_norm": 1250.6552734375,
      "learning_rate": 6.079629629629629e-06,
      "loss": 12.2497,
      "step": 65500
    },
    {
      "epoch": 1.5105394456777974,
      "grad_norm": 0.030488021671772003,
      "learning_rate": 6.0666666666666665e-06,
      "loss": 13.4229,
      "step": 66000
    },
    {
      "epoch": 1.5219829263268716,
      "grad_norm": 3663.417724609375,
      "learning_rate": 6.053703703703704e-06,
      "loss": 11.9166,
      "step": 66500
    },
    {
      "epoch": 1.5334264069759458,
      "grad_norm": 0.10368392616510391,
      "learning_rate": 6.04074074074074e-06,
      "loss": 11.7133,
      "step": 67000
    },
    {
      "epoch": 1.54486988762502,
      "grad_norm": 23.950149536132812,
      "learning_rate": 6.027777777777778e-06,
      "loss": 10.2979,
      "step": 67500
    },
    {
      "epoch": 1.5563133682740942,
      "grad_norm": 1314.4285888671875,
      "learning_rate": 6.0148148148148145e-06,
      "loss": 11.5517,
      "step": 68000
    },
    {
      "epoch": 1.5677568489231684,
      "grad_norm": 0.0,
      "learning_rate": 6.001851851851852e-06,
      "loss": 12.2986,
      "step": 68500
    },
    {
      "epoch": 1.5792003295722425,
      "grad_norm": 558.1203002929688,
      "learning_rate": 5.988888888888888e-06,
      "loss": 11.5298,
      "step": 69000
    },
    {
      "epoch": 1.590643810221317,
      "grad_norm": 0.5483641624450684,
      "learning_rate": 5.975925925925926e-06,
      "loss": 11.2464,
      "step": 69500
    },
    {
      "epoch": 1.6020872908703911,
      "grad_norm": 0.009755097329616547,
      "learning_rate": 5.9629629629629626e-06,
      "loss": 13.0815,
      "step": 70000
    },
    {
      "epoch": 1.6135307715194653,
      "grad_norm": 0.007829127833247185,
      "learning_rate": 5.95e-06,
      "loss": 10.083,
      "step": 70500
    },
    {
      "epoch": 1.6249742521685397,
      "grad_norm": 435.7135314941406,
      "learning_rate": 5.937037037037036e-06,
      "loss": 11.4477,
      "step": 71000
    },
    {
      "epoch": 1.636417732817614,
      "grad_norm": 5.702239513397217,
      "learning_rate": 5.924074074074074e-06,
      "loss": 11.0763,
      "step": 71500
    },
    {
      "epoch": 1.6478612134666881,
      "grad_norm": 0.10179603844881058,
      "learning_rate": 5.911111111111111e-06,
      "loss": 12.8192,
      "step": 72000
    },
    {
      "epoch": 1.6593046941157623,
      "grad_norm": 681.2998046875,
      "learning_rate": 5.898148148148148e-06,
      "loss": 11.8389,
      "step": 72500
    },
    {
      "epoch": 1.6707481747648365,
      "grad_norm": 168.10958862304688,
      "learning_rate": 5.885185185185185e-06,
      "loss": 9.9504,
      "step": 73000
    },
    {
      "epoch": 1.6821916554139107,
      "grad_norm": 0.004441286437213421,
      "learning_rate": 5.872222222222222e-06,
      "loss": 14.6829,
      "step": 73500
    },
    {
      "epoch": 1.6936351360629849,
      "grad_norm": 0.020898401737213135,
      "learning_rate": 5.859259259259259e-06,
      "loss": 11.1156,
      "step": 74000
    },
    {
      "epoch": 1.705078616712059,
      "grad_norm": 0.021843064576387405,
      "learning_rate": 5.846296296296296e-06,
      "loss": 10.6854,
      "step": 74500
    },
    {
      "epoch": 1.7165220973611333,
      "grad_norm": 33.806251525878906,
      "learning_rate": 5.833333333333333e-06,
      "loss": 13.1055,
      "step": 75000
    },
    {
      "epoch": 1.7279655780102074,
      "grad_norm": 29.790348052978516,
      "learning_rate": 5.82037037037037e-06,
      "loss": 12.1967,
      "step": 75500
    },
    {
      "epoch": 1.7394090586592819,
      "grad_norm": 2041.1212158203125,
      "learning_rate": 5.8074074074074076e-06,
      "loss": 12.7192,
      "step": 76000
    },
    {
      "epoch": 1.750852539308356,
      "grad_norm": 1129.54833984375,
      "learning_rate": 5.794444444444444e-06,
      "loss": 12.4452,
      "step": 76500
    },
    {
      "epoch": 1.7622960199574302,
      "grad_norm": 0.3305358290672302,
      "learning_rate": 5.781481481481482e-06,
      "loss": 12.6818,
      "step": 77000
    },
    {
      "epoch": 1.7737395006065044,
      "grad_norm": 5318.0498046875,
      "learning_rate": 5.768518518518518e-06,
      "loss": 11.3843,
      "step": 77500
    },
    {
      "epoch": 1.7851829812555788,
      "grad_norm": 1354.53466796875,
      "learning_rate": 5.755555555555556e-06,
      "loss": 11.9361,
      "step": 78000
    },
    {
      "epoch": 1.796626461904653,
      "grad_norm": 0.01847067102789879,
      "learning_rate": 5.742592592592593e-06,
      "loss": 14.4414,
      "step": 78500
    },
    {
      "epoch": 1.8080699425537272,
      "grad_norm": 0.0,
      "learning_rate": 5.72962962962963e-06,
      "loss": 12.2809,
      "step": 79000
    },
    {
      "epoch": 1.8195134232028014,
      "grad_norm": 0.0010120772058144212,
      "learning_rate": 5.7166666666666664e-06,
      "loss": 11.9991,
      "step": 79500
    },
    {
      "epoch": 1.8309569038518756,
      "grad_norm": 0.0,
      "learning_rate": 5.703703703703704e-06,
      "loss": 10.9662,
      "step": 80000
    },
    {
      "epoch": 1.8424003845009498,
      "grad_norm": 279.47137451171875,
      "learning_rate": 5.690740740740741e-06,
      "loss": 11.1481,
      "step": 80500
    },
    {
      "epoch": 1.853843865150024,
      "grad_norm": 0.0,
      "learning_rate": 5.677777777777778e-06,
      "loss": 12.123,
      "step": 81000
    },
    {
      "epoch": 1.8652873457990982,
      "grad_norm": 807.4469604492188,
      "learning_rate": 5.6648148148148145e-06,
      "loss": 13.1075,
      "step": 81500
    },
    {
      "epoch": 1.8767308264481724,
      "grad_norm": 173.2933349609375,
      "learning_rate": 5.6518518518518525e-06,
      "loss": 12.5405,
      "step": 82000
    },
    {
      "epoch": 1.8881743070972465,
      "grad_norm": 0.0,
      "learning_rate": 5.638888888888889e-06,
      "loss": 13.9821,
      "step": 82500
    },
    {
      "epoch": 1.899617787746321,
      "grad_norm": 2014.615478515625,
      "learning_rate": 5.625925925925926e-06,
      "loss": 9.4421,
      "step": 83000
    },
    {
      "epoch": 1.9110612683953951,
      "grad_norm": 0.0,
      "learning_rate": 5.6129629629629625e-06,
      "loss": 10.8762,
      "step": 83500
    },
    {
      "epoch": 1.9225047490444693,
      "grad_norm": 70.37690734863281,
      "learning_rate": 5.600000000000001e-06,
      "loss": 12.3405,
      "step": 84000
    },
    {
      "epoch": 1.9339482296935437,
      "grad_norm": 0.5092810392379761,
      "learning_rate": 5.587037037037037e-06,
      "loss": 10.5346,
      "step": 84500
    },
    {
      "epoch": 1.945391710342618,
      "grad_norm": 1005.1751098632812,
      "learning_rate": 5.574074074074074e-06,
      "loss": 13.8286,
      "step": 85000
    },
    {
      "epoch": 1.9568351909916921,
      "grad_norm": 0.003633589716628194,
      "learning_rate": 5.5611111111111106e-06,
      "loss": 12.5279,
      "step": 85500
    },
    {
      "epoch": 1.9682786716407663,
      "grad_norm": 398.6470642089844,
      "learning_rate": 5.548148148148149e-06,
      "loss": 10.5745,
      "step": 86000
    },
    {
      "epoch": 1.9797221522898405,
      "grad_norm": 538.2572021484375,
      "learning_rate": 5.535185185185185e-06,
      "loss": 14.3954,
      "step": 86500
    },
    {
      "epoch": 1.9911656329389147,
      "grad_norm": 656.5657348632812,
      "learning_rate": 5.522222222222222e-06,
      "loss": 13.2243,
      "step": 87000
    },
    {
      "epoch": 2.002609113587989,
      "grad_norm": 195.96044921875,
      "learning_rate": 5.5092592592592595e-06,
      "loss": 13.1255,
      "step": 87500
    },
    {
      "epoch": 2.014052594237063,
      "grad_norm": 1494.8990478515625,
      "learning_rate": 5.496296296296297e-06,
      "loss": 8.6693,
      "step": 88000
    },
    {
      "epoch": 2.0254960748861373,
      "grad_norm": 0.0,
      "learning_rate": 5.483333333333333e-06,
      "loss": 10.1512,
      "step": 88500
    },
    {
      "epoch": 2.0369395555352114,
      "grad_norm": 3263.656982421875,
      "learning_rate": 5.47037037037037e-06,
      "loss": 11.7968,
      "step": 89000
    },
    {
      "epoch": 2.0483830361842856,
      "grad_norm": 0.5306602716445923,
      "learning_rate": 5.4574074074074075e-06,
      "loss": 10.1574,
      "step": 89500
    },
    {
      "epoch": 2.05982651683336,
      "grad_norm": 955.903076171875,
      "learning_rate": 5.444444444444445e-06,
      "loss": 8.7719,
      "step": 90000
    },
    {
      "epoch": 2.0712699974824345,
      "grad_norm": 28.916826248168945,
      "learning_rate": 5.431481481481481e-06,
      "loss": 7.6421,
      "step": 90500
    },
    {
      "epoch": 2.0827134781315086,
      "grad_norm": 0.018332993611693382,
      "learning_rate": 5.418518518518518e-06,
      "loss": 9.0135,
      "step": 91000
    },
    {
      "epoch": 2.094156958780583,
      "grad_norm": 240.7179412841797,
      "learning_rate": 5.4055555555555556e-06,
      "loss": 8.8454,
      "step": 91500
    },
    {
      "epoch": 2.105600439429657,
      "grad_norm": 237.4253692626953,
      "learning_rate": 5.392592592592593e-06,
      "loss": 9.7903,
      "step": 92000
    },
    {
      "epoch": 2.117043920078731,
      "grad_norm": 2.053232192993164,
      "learning_rate": 5.379629629629629e-06,
      "loss": 11.0608,
      "step": 92500
    },
    {
      "epoch": 2.1284874007278054,
      "grad_norm": 862.31201171875,
      "learning_rate": 5.366666666666667e-06,
      "loss": 10.3125,
      "step": 93000
    },
    {
      "epoch": 2.1399308813768796,
      "grad_norm": 0.03503242880105972,
      "learning_rate": 5.353703703703704e-06,
      "loss": 12.3845,
      "step": 93500
    },
    {
      "epoch": 2.151374362025954,
      "grad_norm": 0.035824213176965714,
      "learning_rate": 5.340740740740741e-06,
      "loss": 9.1687,
      "step": 94000
    },
    {
      "epoch": 2.162817842675028,
      "grad_norm": 0.0,
      "learning_rate": 5.327777777777777e-06,
      "loss": 9.9659,
      "step": 94500
    },
    {
      "epoch": 2.174261323324102,
      "grad_norm": 87.5394515991211,
      "learning_rate": 5.314814814814815e-06,
      "loss": 9.6173,
      "step": 95000
    },
    {
      "epoch": 2.1857048039731763,
      "grad_norm": 0.0,
      "learning_rate": 5.301851851851852e-06,
      "loss": 9.4475,
      "step": 95500
    },
    {
      "epoch": 2.1971482846222505,
      "grad_norm": 1670.8548583984375,
      "learning_rate": 5.288888888888889e-06,
      "loss": 9.944,
      "step": 96000
    },
    {
      "epoch": 2.2085917652713247,
      "grad_norm": 0.0,
      "learning_rate": 5.275925925925925e-06,
      "loss": 9.0547,
      "step": 96500
    },
    {
      "epoch": 2.220035245920399,
      "grad_norm": 0.06572415679693222,
      "learning_rate": 5.262962962962963e-06,
      "loss": 8.8515,
      "step": 97000
    },
    {
      "epoch": 2.2314787265694735,
      "grad_norm": 0.009067896753549576,
      "learning_rate": 5.25e-06,
      "loss": 10.3228,
      "step": 97500
    },
    {
      "epoch": 2.2429222072185477,
      "grad_norm": 0.0,
      "learning_rate": 5.237037037037037e-06,
      "loss": 10.091,
      "step": 98000
    },
    {
      "epoch": 2.254365687867622,
      "grad_norm": 12197.4453125,
      "learning_rate": 5.224074074074074e-06,
      "loss": 10.6507,
      "step": 98500
    },
    {
      "epoch": 2.265809168516696,
      "grad_norm": 0.006908933632075787,
      "learning_rate": 5.211111111111111e-06,
      "loss": 9.9997,
      "step": 99000
    },
    {
      "epoch": 2.2772526491657703,
      "grad_norm": 786.2550659179688,
      "learning_rate": 5.198148148148148e-06,
      "loss": 10.4626,
      "step": 99500
    },
    {
      "epoch": 2.2886961298148445,
      "grad_norm": 206.81629943847656,
      "learning_rate": 5.185185185185185e-06,
      "loss": 9.6626,
      "step": 100000
    }
  ],
  "logging_steps": 500,
  "max_steps": 300000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}