{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.6753246753246753,
  "eval_steps": 1000,
  "global_step": 52000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0012987012987012987,
      "grad_norm": 0.8535653352737427,
      "learning_rate": 9.42857142857143e-05,
      "loss": 8.6239,
      "step": 100
    },
    {
      "epoch": 0.0025974025974025974,
      "grad_norm": 0.9884688258171082,
      "learning_rate": 0.0001895238095238095,
      "loss": 6.7685,
      "step": 200
    },
    {
      "epoch": 0.003896103896103896,
      "grad_norm": 0.5704336762428284,
      "learning_rate": 0.0002847619047619048,
      "loss": 5.9666,
      "step": 300
    },
    {
      "epoch": 0.005194805194805195,
      "grad_norm": 0.6232308149337769,
      "learning_rate": 0.00038,
      "loss": 5.518,
      "step": 400
    },
    {
      "epoch": 0.006493506493506494,
      "grad_norm": 0.6595708131790161,
      "learning_rate": 0.00047523809523809525,
      "loss": 5.205,
      "step": 500
    },
    {
      "epoch": 0.007792207792207792,
      "grad_norm": 0.5402975678443909,
      "learning_rate": 0.0005704761904761905,
      "loss": 4.9366,
      "step": 600
    },
    {
      "epoch": 0.00909090909090909,
      "grad_norm": 0.5196501016616821,
      "learning_rate": 0.0006657142857142858,
      "loss": 4.7613,
      "step": 700
    },
    {
      "epoch": 0.01038961038961039,
      "grad_norm": 0.6473610401153564,
      "learning_rate": 0.0007609523809523809,
      "loss": 4.6361,
      "step": 800
    },
    {
      "epoch": 0.011688311688311689,
      "grad_norm": 0.4435962736606598,
      "learning_rate": 0.0008561904761904762,
      "loss": 4.5062,
      "step": 900
    },
    {
      "epoch": 0.012987012987012988,
      "grad_norm": 0.46511825919151306,
      "learning_rate": 0.0009514285714285714,
      "loss": 4.3812,
      "step": 1000
    },
    {
      "epoch": 0.012987012987012988,
      "eval_loss": 4.666484355926514,
      "eval_runtime": 15.6105,
      "eval_samples_per_second": 36.898,
      "eval_steps_per_second": 9.225,
      "step": 1000
    },
    {
      "epoch": 0.014285714285714285,
      "grad_norm": 0.38127702474594116,
      "learning_rate": 0.0009999995009117187,
      "loss": 4.2976,
      "step": 1100
    },
    {
      "epoch": 0.015584415584415584,
      "grad_norm": 0.4769284129142761,
      "learning_rate": 0.0009999953851546302,
      "loss": 4.1888,
      "step": 1200
    },
    {
      "epoch": 0.016883116883116882,
      "grad_norm": 0.4059095084667206,
      "learning_rate": 0.000999987112101314,
      "loss": 4.0874,
      "step": 1300
    },
    {
      "epoch": 0.01818181818181818,
      "grad_norm": 0.34086140990257263,
      "learning_rate": 0.0009999746818205575,
      "loss": 3.974,
      "step": 1400
    },
    {
      "epoch": 0.01948051948051948,
      "grad_norm": 0.32273247838020325,
      "learning_rate": 0.0009999580944157148,
      "loss": 3.9202,
      "step": 1500
    },
    {
      "epoch": 0.02077922077922078,
      "grad_norm": 0.36565664410591125,
      "learning_rate": 0.000999937350024705,
      "loss": 3.8821,
      "step": 1600
    },
    {
      "epoch": 0.02207792207792208,
      "grad_norm": 0.3269343376159668,
      "learning_rate": 0.0009999124488200097,
      "loss": 3.7686,
      "step": 1700
    },
    {
      "epoch": 0.023376623376623377,
      "grad_norm": 0.3859221339225769,
      "learning_rate": 0.0009998833910086751,
      "loss": 3.7079,
      "step": 1800
    },
    {
      "epoch": 0.024675324675324677,
      "grad_norm": 0.3663485646247864,
      "learning_rate": 0.000999850176832307,
      "loss": 3.6803,
      "step": 1900
    },
    {
      "epoch": 0.025974025974025976,
      "grad_norm": 0.46907544136047363,
      "learning_rate": 0.0009998128065670706,
      "loss": 3.6502,
      "step": 2000
    },
    {
      "epoch": 0.025974025974025976,
      "eval_loss": 3.9888415336608887,
      "eval_runtime": 15.6284,
      "eval_samples_per_second": 36.856,
      "eval_steps_per_second": 9.214,
      "step": 2000
    },
    {
      "epoch": 0.02727272727272727,
      "grad_norm": 0.3356938362121582,
      "learning_rate": 0.0009997712805236867,
      "loss": 3.6339,
      "step": 2100
    },
    {
      "epoch": 0.02857142857142857,
      "grad_norm": 0.2554614543914795,
      "learning_rate": 0.0009997255990474312,
      "loss": 3.601,
      "step": 2200
    },
    {
      "epoch": 0.02987012987012987,
      "grad_norm": 0.2642477750778198,
      "learning_rate": 0.00099967576251813,
      "loss": 3.5622,
      "step": 2300
    },
    {
      "epoch": 0.03116883116883117,
      "grad_norm": 0.2671547532081604,
      "learning_rate": 0.0009996217713501576,
      "loss": 3.517,
      "step": 2400
    },
    {
      "epoch": 0.032467532467532464,
      "grad_norm": 0.26076218485832214,
      "learning_rate": 0.0009995636259924328,
      "loss": 3.5058,
      "step": 2500
    },
    {
      "epoch": 0.033766233766233764,
      "grad_norm": 0.39065253734588623,
      "learning_rate": 0.0009995013269284145,
      "loss": 3.5262,
      "step": 2600
    },
    {
      "epoch": 0.03506493506493506,
      "grad_norm": 0.23748748004436493,
      "learning_rate": 0.0009994348746760996,
      "loss": 3.4818,
      "step": 2700
    },
    {
      "epoch": 0.03636363636363636,
      "grad_norm": 0.23590387403964996,
      "learning_rate": 0.000999364269788016,
      "loss": 3.4639,
      "step": 2800
    },
    {
      "epoch": 0.03766233766233766,
      "grad_norm": 0.2318197786808014,
      "learning_rate": 0.00099928951285122,
      "loss": 3.4444,
      "step": 2900
    },
    {
      "epoch": 0.03896103896103896,
      "grad_norm": 0.2852722406387329,
      "learning_rate": 0.0009992106044872913,
      "loss": 3.4227,
      "step": 3000
    },
    {
      "epoch": 0.03896103896103896,
      "eval_loss": 3.778900623321533,
      "eval_runtime": 15.1585,
      "eval_samples_per_second": 37.998,
      "eval_steps_per_second": 9.5,
      "step": 3000
    },
    {
      "epoch": 0.04025974025974026,
      "grad_norm": 0.23955008387565613,
      "learning_rate": 0.0009991275453523263,
      "loss": 3.3974,
      "step": 3100
    },
    {
      "epoch": 0.04155844155844156,
      "grad_norm": 0.2509465217590332,
      "learning_rate": 0.0009990403361369346,
      "loss": 3.4157,
      "step": 3200
    },
    {
      "epoch": 0.04285714285714286,
      "grad_norm": 0.24335195124149323,
      "learning_rate": 0.0009989489775662315,
      "loss": 3.3972,
      "step": 3300
    },
    {
      "epoch": 0.04415584415584416,
      "grad_norm": 0.7076203227043152,
      "learning_rate": 0.000998853470399834,
      "loss": 3.4082,
      "step": 3400
    },
    {
      "epoch": 0.045454545454545456,
      "grad_norm": 0.25543642044067383,
      "learning_rate": 0.0009987538154318518,
      "loss": 3.373,
      "step": 3500
    },
    {
      "epoch": 0.046753246753246755,
      "grad_norm": 0.7122920155525208,
      "learning_rate": 0.0009986500134908838,
      "loss": 3.3898,
      "step": 3600
    },
    {
      "epoch": 0.048051948051948054,
      "grad_norm": 0.2252849042415619,
      "learning_rate": 0.000998542065440008,
      "loss": 3.3739,
      "step": 3700
    },
    {
      "epoch": 0.04935064935064935,
      "grad_norm": 0.20722781121730804,
      "learning_rate": 0.0009984299721767773,
      "loss": 3.353,
      "step": 3800
    },
    {
      "epoch": 0.05064935064935065,
      "grad_norm": 0.2006598860025406,
      "learning_rate": 0.00099831373463321,
      "loss": 3.3124,
      "step": 3900
    },
    {
      "epoch": 0.05194805194805195,
      "grad_norm": 0.2024812549352646,
      "learning_rate": 0.0009981933537757828,
      "loss": 3.3367,
      "step": 4000
    },
    {
      "epoch": 0.05194805194805195,
      "eval_loss": 3.681532144546509,
      "eval_runtime": 16.9008,
      "eval_samples_per_second": 34.081,
      "eval_steps_per_second": 8.52,
      "step": 4000
    },
    {
      "epoch": 0.053246753246753244,
      "grad_norm": 0.24371765553951263,
      "learning_rate": 0.0009980688306054227,
      "loss": 3.2775,
      "step": 4100
    },
    {
      "epoch": 0.05454545454545454,
      "grad_norm": 0.20401820540428162,
      "learning_rate": 0.0009979401661574985,
      "loss": 3.2952,
      "step": 4200
    },
    {
      "epoch": 0.05584415584415584,
      "grad_norm": 0.2212119996547699,
      "learning_rate": 0.0009978073615018127,
      "loss": 3.257,
      "step": 4300
    },
    {
      "epoch": 0.05714285714285714,
      "grad_norm": 0.2384602129459381,
      "learning_rate": 0.000997670417742592,
      "loss": 3.2859,
      "step": 4400
    },
    {
      "epoch": 0.05844155844155844,
      "grad_norm": 0.21143975853919983,
      "learning_rate": 0.0009975293360184785,
      "loss": 3.2475,
      "step": 4500
    },
    {
      "epoch": 0.05974025974025974,
      "grad_norm": 0.20411203801631927,
      "learning_rate": 0.00099738411750252,
      "loss": 3.2443,
      "step": 4600
    },
    {
      "epoch": 0.06103896103896104,
      "grad_norm": 0.2097695916891098,
      "learning_rate": 0.0009972347634021606,
      "loss": 3.2703,
      "step": 4700
    },
    {
      "epoch": 0.06233766233766234,
      "grad_norm": 0.2140115350484848,
      "learning_rate": 0.00099708127495923,
      "loss": 3.2569,
      "step": 4800
    },
    {
      "epoch": 0.06363636363636363,
      "grad_norm": 0.2163541465997696,
      "learning_rate": 0.000996923653449934,
      "loss": 3.2646,
      "step": 4900
    },
    {
      "epoch": 0.06493506493506493,
      "grad_norm": 0.1931372582912445,
      "learning_rate": 0.0009967619001848434,
      "loss": 3.2224,
      "step": 5000
    },
    {
      "epoch": 0.06493506493506493,
      "eval_loss": 3.59517240524292,
      "eval_runtime": 15.382,
      "eval_samples_per_second": 37.446,
      "eval_steps_per_second": 9.362,
      "step": 5000
    },
    {
      "epoch": 0.06623376623376623,
      "grad_norm": 0.2017965465784073,
      "learning_rate": 0.0009965960165088828,
      "loss": 3.2159,
      "step": 5100
    },
    {
      "epoch": 0.06753246753246753,
      "grad_norm": 0.2096720039844513,
      "learning_rate": 0.0009964260038013203,
      "loss": 3.1967,
      "step": 5200
    },
    {
      "epoch": 0.06883116883116883,
      "grad_norm": 0.2119290977716446,
      "learning_rate": 0.000996251863475755,
      "loss": 3.1962,
      "step": 5300
    },
    {
      "epoch": 0.07012987012987013,
      "grad_norm": 0.27017727494239807,
      "learning_rate": 0.0009960735969801067,
      "loss": 3.1672,
      "step": 5400
    },
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 0.21126972138881683,
      "learning_rate": 0.0009958912057966018,
      "loss": 3.2301,
      "step": 5500
    },
    {
      "epoch": 0.07272727272727272,
      "grad_norm": 0.238760307431221,
      "learning_rate": 0.0009957046914417628,
      "loss": 3.2367,
      "step": 5600
    },
    {
      "epoch": 0.07402597402597402,
      "grad_norm": 0.19038313627243042,
      "learning_rate": 0.000995514055466395,
      "loss": 3.2026,
      "step": 5700
    },
    {
      "epoch": 0.07532467532467532,
      "grad_norm": 0.20789292454719543,
      "learning_rate": 0.0009953192994555736,
      "loss": 3.1718,
      "step": 5800
    },
    {
      "epoch": 0.07662337662337662,
      "grad_norm": 0.18676762282848358,
      "learning_rate": 0.00099512042502863,
      "loss": 3.167,
      "step": 5900
    },
    {
      "epoch": 0.07792207792207792,
      "grad_norm": 0.19430765509605408,
      "learning_rate": 0.0009949174338391394,
      "loss": 3.1721,
      "step": 6000
    },
    {
      "epoch": 0.07792207792207792,
      "eval_loss": 3.5258212089538574,
      "eval_runtime": 15.3635,
      "eval_samples_per_second": 37.491,
      "eval_steps_per_second": 9.373,
      "step": 6000
    },
    {
      "epoch": 0.07922077922077922,
      "grad_norm": 0.20157617330551147,
      "learning_rate": 0.0009947103275749064,
      "loss": 3.1696,
      "step": 6100
    },
    {
      "epoch": 0.08051948051948052,
      "grad_norm": 0.18047969043254852,
      "learning_rate": 0.0009944991079579514,
      "loss": 3.1601,
      "step": 6200
    },
    {
      "epoch": 0.08181818181818182,
      "grad_norm": 0.2030571699142456,
      "learning_rate": 0.0009942837767444952,
      "loss": 3.1709,
      "step": 6300
    },
    {
      "epoch": 0.08311688311688312,
      "grad_norm": 0.18791615962982178,
      "learning_rate": 0.000994064335724946,
      "loss": 3.1808,
      "step": 6400
    },
    {
      "epoch": 0.08441558441558442,
      "grad_norm": 0.1927482783794403,
      "learning_rate": 0.0009938407867238828,
      "loss": 3.1272,
      "step": 6500
    },
    {
      "epoch": 0.08571428571428572,
      "grad_norm": 0.1764034628868103,
      "learning_rate": 0.0009936131316000418,
      "loss": 3.1407,
      "step": 6600
    },
    {
      "epoch": 0.08701298701298701,
      "grad_norm": 0.1786874383687973,
      "learning_rate": 0.0009933813722463004,
      "loss": 3.107,
      "step": 6700
    },
    {
      "epoch": 0.08831168831168831,
      "grad_norm": 0.2685558497905731,
      "learning_rate": 0.0009931455105896606,
      "loss": 3.1609,
      "step": 6800
    },
    {
      "epoch": 0.08961038961038961,
      "grad_norm": 0.19291436672210693,
      "learning_rate": 0.000992905548591234,
      "loss": 3.1444,
      "step": 6900
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 0.1896260678768158,
      "learning_rate": 0.0009926614882462253,
      "loss": 3.1329,
      "step": 7000
    },
    {
      "epoch": 0.09090909090909091,
      "eval_loss": 3.4966745376586914,
      "eval_runtime": 14.4834,
      "eval_samples_per_second": 39.77,
      "eval_steps_per_second": 9.942,
      "step": 7000
    },
    {
      "epoch": 0.09220779220779221,
      "grad_norm": 0.17750093340873718,
      "learning_rate": 0.0009924133315839156,
      "loss": 3.1602,
      "step": 7100
    },
    {
      "epoch": 0.09350649350649351,
      "grad_norm": 0.18574649095535278,
      "learning_rate": 0.0009921610806676456,
      "loss": 3.1451,
      "step": 7200
    },
    {
      "epoch": 0.09480519480519481,
      "grad_norm": 0.18256109952926636,
      "learning_rate": 0.000991904737594798,
      "loss": 3.1305,
      "step": 7300
    },
    {
      "epoch": 0.09610389610389611,
      "grad_norm": 0.17488889396190643,
      "learning_rate": 0.0009916443044967807,
      "loss": 3.1211,
      "step": 7400
    },
    {
      "epoch": 0.09740259740259741,
      "grad_norm": 0.205893412232399,
      "learning_rate": 0.0009913797835390088,
      "loss": 3.1039,
      "step": 7500
    },
    {
      "epoch": 0.0987012987012987,
      "grad_norm": 0.20423896610736847,
      "learning_rate": 0.0009911111769208864,
      "loss": 3.1159,
      "step": 7600
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.18055465817451477,
      "learning_rate": 0.000990838486875789,
      "loss": 3.0872,
      "step": 7700
    },
    {
      "epoch": 0.1012987012987013,
      "grad_norm": 0.18993014097213745,
      "learning_rate": 0.0009905617156710437,
      "loss": 3.0976,
      "step": 7800
    },
    {
      "epoch": 0.1025974025974026,
      "grad_norm": 0.1896851807832718,
      "learning_rate": 0.000990280865607912,
      "loss": 3.0932,
      "step": 7900
    },
    {
      "epoch": 0.1038961038961039,
      "grad_norm": 0.18168048560619354,
      "learning_rate": 0.0009899959390215689,
      "loss": 3.0862,
      "step": 8000
    },
    {
      "epoch": 0.1038961038961039,
      "eval_loss": 3.4424216747283936,
      "eval_runtime": 12.6555,
      "eval_samples_per_second": 45.514,
      "eval_steps_per_second": 11.378,
      "step": 8000
    },
    {
      "epoch": 0.10519480519480519,
      "grad_norm": 0.19026321172714233,
      "learning_rate": 0.000989706938281085,
      "loss": 3.0792,
      "step": 8100
    },
    {
      "epoch": 0.10649350649350649,
      "grad_norm": 0.22636525332927704,
      "learning_rate": 0.0009894138657894054,
      "loss": 3.0808,
      "step": 8200
    },
    {
      "epoch": 0.10779220779220779,
      "grad_norm": 0.1759577989578247,
      "learning_rate": 0.0009891167239833311,
      "loss": 3.0651,
      "step": 8300
    },
    {
      "epoch": 0.10909090909090909,
      "grad_norm": 0.19504410028457642,
      "learning_rate": 0.0009888155153334984,
      "loss": 3.0701,
      "step": 8400
    },
    {
      "epoch": 0.11038961038961038,
      "grad_norm": 0.20704291760921478,
      "learning_rate": 0.000988510242344357,
      "loss": 3.072,
      "step": 8500
    },
    {
      "epoch": 0.11168831168831168,
      "grad_norm": 0.19917143881320953,
      "learning_rate": 0.000988200907554151,
      "loss": 3.0716,
      "step": 8600
    },
    {
      "epoch": 0.11298701298701298,
      "grad_norm": 0.1709250509738922,
      "learning_rate": 0.000987887513534897,
      "loss": 3.0622,
      "step": 8700
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 0.2797200381755829,
      "learning_rate": 0.0009875700628923622,
      "loss": 3.0743,
      "step": 8800
    },
    {
      "epoch": 0.11558441558441558,
      "grad_norm": 0.172173872590065,
      "learning_rate": 0.000987248558266044,
      "loss": 3.0676,
      "step": 8900
    },
    {
      "epoch": 0.11688311688311688,
      "grad_norm": 0.1997879594564438,
      "learning_rate": 0.000986923002329147,
      "loss": 3.0591,
      "step": 9000
    },
    {
      "epoch": 0.11688311688311688,
      "eval_loss": 3.424792766571045,
      "eval_runtime": 15.3808,
      "eval_samples_per_second": 37.449,
      "eval_steps_per_second": 9.362,
      "step": 9000
    },
    {
      "epoch": 0.11818181818181818,
      "grad_norm": 0.16804350912570953,
      "learning_rate": 0.0009865933977885612,
      "loss": 3.0642,
      "step": 9100
    },
    {
      "epoch": 0.11948051948051948,
      "grad_norm": 0.1866862177848816,
      "learning_rate": 0.0009862597473848393,
      "loss": 3.0378,
      "step": 9200
    },
    {
      "epoch": 0.12077922077922078,
      "grad_norm": 0.2142857313156128,
      "learning_rate": 0.000985922053892174,
      "loss": 3.057,
      "step": 9300
    },
    {
      "epoch": 0.12207792207792208,
      "grad_norm": 0.1772162914276123,
      "learning_rate": 0.0009855803201183743,
      "loss": 3.0585,
      "step": 9400
    },
    {
      "epoch": 0.12337662337662338,
      "grad_norm": 0.1796869933605194,
      "learning_rate": 0.0009852345489048447,
      "loss": 3.0228,
      "step": 9500
    },
    {
      "epoch": 0.12467532467532468,
      "grad_norm": 0.19519853591918945,
      "learning_rate": 0.0009848847431265576,
      "loss": 3.0493,
      "step": 9600
    },
    {
      "epoch": 0.12597402597402596,
      "grad_norm": 0.1866455227136612,
      "learning_rate": 0.0009845309056920326,
      "loss": 3.0369,
      "step": 9700
    },
    {
      "epoch": 0.12727272727272726,
      "grad_norm": 0.1888640969991684,
      "learning_rate": 0.000984173039543311,
      "loss": 3.0388,
      "step": 9800
    },
    {
      "epoch": 0.12857142857142856,
      "grad_norm": 0.1753017008304596,
      "learning_rate": 0.0009838111476559313,
      "loss": 3.0533,
      "step": 9900
    },
    {
      "epoch": 0.12987012987012986,
      "grad_norm": 0.17457301914691925,
      "learning_rate": 0.000983445233038905,
      "loss": 3.0569,
      "step": 10000
    },
    {
      "epoch": 0.12987012987012986,
      "eval_loss": 3.401273012161255,
      "eval_runtime": 14.6757,
      "eval_samples_per_second": 39.248,
      "eval_steps_per_second": 9.812,
      "step": 10000
    },
    {
      "epoch": 0.13116883116883116,
      "grad_norm": 0.17421157658100128,
      "learning_rate": 0.0009830752987346908,
      "loss": 3.0596,
      "step": 10100
    },
    {
      "epoch": 0.13246753246753246,
      "grad_norm": 0.1772899180650711,
      "learning_rate": 0.0009827013478191703,
      "loss": 3.016,
      "step": 10200
    },
    {
      "epoch": 0.13376623376623376,
      "grad_norm": 0.169020414352417,
      "learning_rate": 0.0009823233834016214,
      "loss": 3.0173,
      "step": 10300
    },
    {
      "epoch": 0.13506493506493505,
      "grad_norm": 0.34023940563201904,
      "learning_rate": 0.0009819414086246938,
      "loss": 3.0004,
      "step": 10400
    },
    {
      "epoch": 0.13636363636363635,
      "grad_norm": 0.19924494624137878,
      "learning_rate": 0.0009815554266643808,
      "loss": 3.003,
      "step": 10500
    },
    {
      "epoch": 0.13766233766233765,
      "grad_norm": 0.1713639199733734,
      "learning_rate": 0.0009811654407299948,
      "loss": 2.9895,
      "step": 10600
    },
    {
      "epoch": 0.13896103896103895,
      "grad_norm": 0.16809602081775665,
      "learning_rate": 0.00098077145406414,
      "loss": 3.0078,
      "step": 10700
    },
    {
      "epoch": 0.14025974025974025,
      "grad_norm": 0.18249379098415375,
      "learning_rate": 0.0009803734699426853,
      "loss": 3.0379,
      "step": 10800
    },
    {
      "epoch": 0.14155844155844155,
      "grad_norm": 0.19280089437961578,
      "learning_rate": 0.0009799714916747368,
      "loss": 2.9917,
      "step": 10900
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 0.18399550020694733,
      "learning_rate": 0.000979565522602611,
      "loss": 3.0597,
      "step": 11000
    },
    {
      "epoch": 0.14285714285714285,
      "eval_loss": 3.4125261306762695,
      "eval_runtime": 15.5875,
      "eval_samples_per_second": 36.953,
      "eval_steps_per_second": 9.238,
      "step": 11000
    },
    {
      "epoch": 0.14415584415584415,
      "grad_norm": 0.17574363946914673,
      "learning_rate": 0.000979155566101806,
      "loss": 3.0168,
      "step": 11100
    },
    {
      "epoch": 0.14545454545454545,
      "grad_norm": 0.17517109215259552,
      "learning_rate": 0.0009787416255809752,
      "loss": 2.9817,
      "step": 11200
    },
    {
      "epoch": 0.14675324675324675,
      "grad_norm": 0.18359977006912231,
      "learning_rate": 0.0009783237044818968,
      "loss": 2.9916,
      "step": 11300
    },
    {
      "epoch": 0.14805194805194805,
      "grad_norm": 0.17518875002861023,
      "learning_rate": 0.000977901806279446,
      "loss": 2.9567,
      "step": 11400
    },
    {
      "epoch": 0.14935064935064934,
      "grad_norm": 0.19247221946716309,
      "learning_rate": 0.0009774759344815674,
      "loss": 2.9901,
      "step": 11500
    },
    {
      "epoch": 0.15064935064935064,
      "grad_norm": 0.18513117730617523,
      "learning_rate": 0.000977046092629244,
      "loss": 2.992,
      "step": 11600
    },
    {
      "epoch": 0.15194805194805194,
      "grad_norm": 0.18412470817565918,
      "learning_rate": 0.0009766122842964683,
      "loss": 2.9985,
      "step": 11700
    },
    {
      "epoch": 0.15324675324675324,
      "grad_norm": 0.19853200018405914,
      "learning_rate": 0.0009761745130902134,
      "loss": 2.978,
      "step": 11800
    },
    {
      "epoch": 0.15454545454545454,
      "grad_norm": 0.19611628353595734,
      "learning_rate": 0.0009757327826504022,
      "loss": 2.9771,
      "step": 11900
    },
    {
      "epoch": 0.15584415584415584,
      "grad_norm": 0.17522920668125153,
      "learning_rate": 0.0009752870966498766,
      "loss": 2.9641,
      "step": 12000
    },
    {
      "epoch": 0.15584415584415584,
      "eval_loss": 3.3487231731414795,
      "eval_runtime": 13.9045,
      "eval_samples_per_second": 41.425,
      "eval_steps_per_second": 10.356,
      "step": 12000
    },
    {
      "epoch": 0.15714285714285714,
      "grad_norm": 0.18391777575016022,
      "learning_rate": 0.0009748374587943688,
      "loss": 2.9694,
      "step": 12100
    },
    {
      "epoch": 0.15844155844155844,
      "grad_norm": 0.20561572909355164,
      "learning_rate": 0.0009743838728224687,
      "loss": 3.0187,
      "step": 12200
    },
    {
      "epoch": 0.15974025974025974,
      "grad_norm": 0.17854483425617218,
      "learning_rate": 0.0009739263425055934,
      "loss": 2.9687,
      "step": 12300
    },
    {
      "epoch": 0.16103896103896104,
      "grad_norm": 0.18647781014442444,
      "learning_rate": 0.0009734648716479563,
      "loss": 2.9574,
      "step": 12400
    },
    {
      "epoch": 0.16233766233766234,
      "grad_norm": 0.1802452653646469,
      "learning_rate": 0.0009729994640865349,
      "loss": 2.9812,
      "step": 12500
    },
    {
      "epoch": 0.16363636363636364,
      "grad_norm": 0.18787842988967896,
      "learning_rate": 0.0009725301236910393,
      "loss": 2.9771,
      "step": 12600
    },
    {
      "epoch": 0.16493506493506493,
      "grad_norm": 0.18528838455677032,
      "learning_rate": 0.0009720568543638793,
      "loss": 2.9525,
      "step": 12700
    },
    {
      "epoch": 0.16623376623376623,
      "grad_norm": 0.17883966863155365,
      "learning_rate": 0.000971579660040133,
      "loss": 2.9624,
      "step": 12800
    },
    {
      "epoch": 0.16753246753246753,
      "grad_norm": 0.17459270358085632,
      "learning_rate": 0.0009710985446875134,
      "loss": 2.9696,
      "step": 12900
    },
    {
      "epoch": 0.16883116883116883,
      "grad_norm": 0.1830231249332428,
      "learning_rate": 0.0009706135123063352,
      "loss": 2.9609,
      "step": 13000
    },
    {
      "epoch": 0.16883116883116883,
      "eval_loss": 3.337019920349121,
      "eval_runtime": 15.8318,
      "eval_samples_per_second": 36.383,
      "eval_steps_per_second": 9.096,
      "step": 13000
    },
    {
      "epoch": 0.17012987012987013,
      "grad_norm": 0.20791789889335632,
      "learning_rate": 0.0009701245669294825,
      "loss": 2.9953,
      "step": 13100
    },
    {
      "epoch": 0.17142857142857143,
      "grad_norm": 0.20302486419677734,
      "learning_rate": 0.0009696317126223742,
      "loss": 3.0519,
      "step": 13200
    },
    {
      "epoch": 0.17272727272727273,
      "grad_norm": 0.17939983308315277,
      "learning_rate": 0.000969134953482931,
      "loss": 2.9618,
      "step": 13300
    },
    {
      "epoch": 0.17402597402597403,
      "grad_norm": 0.2437734603881836,
      "learning_rate": 0.0009686342936415407,
      "loss": 2.955,
      "step": 13400
    },
    {
      "epoch": 0.17532467532467533,
      "grad_norm": 0.19909310340881348,
      "learning_rate": 0.000968129737261024,
      "loss": 2.9898,
      "step": 13500
    },
    {
      "epoch": 0.17662337662337663,
      "grad_norm": 0.1696212738752365,
      "learning_rate": 0.000967621288536601,
      "loss": 2.9793,
      "step": 13600
    },
    {
      "epoch": 0.17792207792207793,
      "grad_norm": 0.171333447098732,
      "learning_rate": 0.0009671089516958538,
      "loss": 2.942,
      "step": 13700
    },
    {
      "epoch": 0.17922077922077922,
      "grad_norm": 0.17436903715133667,
      "learning_rate": 0.0009665927309986944,
      "loss": 2.9574,
      "step": 13800
    },
    {
      "epoch": 0.18051948051948052,
      "grad_norm": 0.20607399940490723,
      "learning_rate": 0.0009660726307373266,
      "loss": 2.965,
      "step": 13900
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 0.17206591367721558,
      "learning_rate": 0.0009655486552362127,
      "loss": 2.9532,
      "step": 14000
    },
    {
      "epoch": 0.18181818181818182,
      "eval_loss": 3.327735424041748,
      "eval_runtime": 13.5364,
      "eval_samples_per_second": 42.552,
      "eval_steps_per_second": 10.638,
      "step": 14000
    },
    {
      "epoch": 0.18311688311688312,
      "grad_norm": 0.20287004113197327,
      "learning_rate": 0.000965020808852035,
      "loss": 2.9153,
      "step": 14100
    },
    {
      "epoch": 0.18441558441558442,
      "grad_norm": 0.18036919832229614,
      "learning_rate": 0.0009644890959736619,
      "loss": 2.933,
      "step": 14200
    },
    {
      "epoch": 0.18571428571428572,
      "grad_norm": 0.21330875158309937,
      "learning_rate": 0.0009639535210221099,
      "loss": 2.9773,
      "step": 14300
    },
    {
      "epoch": 0.18701298701298702,
      "grad_norm": 0.20942148566246033,
      "learning_rate": 0.000963414088450508,
      "loss": 2.9225,
      "step": 14400
    },
    {
      "epoch": 0.18831168831168832,
      "grad_norm": 0.188313826918602,
      "learning_rate": 0.0009628708027440592,
      "loss": 2.9409,
      "step": 14500
    },
    {
      "epoch": 0.18961038961038962,
      "grad_norm": 0.17884506285190582,
      "learning_rate": 0.0009623236684200043,
      "loss": 2.9635,
      "step": 14600
    },
    {
      "epoch": 0.19090909090909092,
      "grad_norm": 0.17130698263645172,
      "learning_rate": 0.0009617726900275845,
      "loss": 2.9447,
      "step": 14700
    },
    {
      "epoch": 0.19220779220779222,
      "grad_norm": 0.19943124055862427,
      "learning_rate": 0.0009612178721480027,
      "loss": 2.9525,
      "step": 14800
    },
    {
      "epoch": 0.19350649350649352,
      "grad_norm": 0.18249951303005219,
      "learning_rate": 0.0009606592193943861,
      "loss": 2.9405,
      "step": 14900
    },
    {
      "epoch": 0.19480519480519481,
      "grad_norm": 0.17710338532924652,
      "learning_rate": 0.0009600967364117478,
      "loss": 2.921,
      "step": 15000
    },
    {
      "epoch": 0.19480519480519481,
      "eval_loss": 3.307467460632324,
      "eval_runtime": 15.2964,
      "eval_samples_per_second": 37.656,
      "eval_steps_per_second": 9.414,
      "step": 15000
    },
    {
      "epoch": 0.1961038961038961,
      "grad_norm": 0.2033987194299698,
      "learning_rate": 0.0009595304278769472,
      "loss": 2.9253,
      "step": 15100
    },
    {
      "epoch": 0.1974025974025974,
      "grad_norm": 0.17457351088523865,
      "learning_rate": 0.000958960298498653,
      "loss": 2.9183,
      "step": 15200
    },
    {
      "epoch": 0.1987012987012987,
      "grad_norm": 0.2011554092168808,
      "learning_rate": 0.0009583863530173018,
      "loss": 2.9254,
      "step": 15300
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.18333646655082703,
      "learning_rate": 0.0009578085962050609,
      "loss": 2.9383,
      "step": 15400
    },
    {
      "epoch": 0.2012987012987013,
      "grad_norm": 0.167351633310318,
      "learning_rate": 0.0009572270328657869,
      "loss": 2.9435,
      "step": 15500
    },
    {
      "epoch": 0.2025974025974026,
      "grad_norm": 0.17198020219802856,
      "learning_rate": 0.0009566416678349864,
      "loss": 2.9336,
      "step": 15600
    },
    {
      "epoch": 0.2038961038961039,
      "grad_norm": 0.19012530148029327,
      "learning_rate": 0.0009560525059797762,
      "loss": 2.9353,
      "step": 15700
    },
    {
      "epoch": 0.2051948051948052,
      "grad_norm": 0.18222008645534515,
      "learning_rate": 0.0009554595521988423,
      "loss": 2.934,
      "step": 15800
    },
    {
      "epoch": 0.2064935064935065,
      "grad_norm": 0.18934360146522522,
      "learning_rate": 0.0009548628114223989,
      "loss": 2.902,
      "step": 15900
    },
    {
      "epoch": 0.2077922077922078,
      "grad_norm": 0.17503845691680908,
      "learning_rate": 0.0009542622886121486,
      "loss": 2.9485,
      "step": 16000
    },
    {
      "epoch": 0.2077922077922078,
      "eval_loss": 3.305332660675049,
      "eval_runtime": 14.5112,
      "eval_samples_per_second": 39.693,
      "eval_steps_per_second": 9.923,
      "step": 16000
    },
    {
      "epoch": 0.20909090909090908,
      "grad_norm": 0.19923605024814606,
      "learning_rate": 0.0009536579887612396,
      "loss": 2.9049,
      "step": 16100
    },
    {
      "epoch": 0.21038961038961038,
      "grad_norm": 0.1969510018825531,
      "learning_rate": 0.0009530499168942252,
      "loss": 2.9263,
      "step": 16200
    },
    {
      "epoch": 0.21168831168831168,
      "grad_norm": 0.3835534155368805,
      "learning_rate": 0.0009524380780670223,
      "loss": 2.9207,
      "step": 16300
    },
    {
      "epoch": 0.21298701298701297,
      "grad_norm": 0.18829414248466492,
      "learning_rate": 0.0009518224773668678,
      "loss": 2.9247,
      "step": 16400
    },
    {
      "epoch": 0.21428571428571427,
      "grad_norm": 0.18950419127941132,
      "learning_rate": 0.0009512031199122779,
      "loss": 2.914,
      "step": 16500
    },
    {
      "epoch": 0.21558441558441557,
      "grad_norm": 0.23597006499767303,
      "learning_rate": 0.0009505800108530054,
      "loss": 2.9064,
      "step": 16600
    },
    {
      "epoch": 0.21688311688311687,
      "grad_norm": 0.18245990574359894,
      "learning_rate": 0.0009499531553699957,
      "loss": 2.9219,
      "step": 16700
    },
    {
      "epoch": 0.21818181818181817,
      "grad_norm": 0.20525243878364563,
      "learning_rate": 0.0009493225586753449,
      "loss": 2.9373,
      "step": 16800
    },
    {
      "epoch": 0.21948051948051947,
      "grad_norm": 0.17929033935070038,
      "learning_rate": 0.0009486882260122557,
      "loss": 2.8947,
      "step": 16900
    },
    {
      "epoch": 0.22077922077922077,
      "grad_norm": 0.18927592039108276,
      "learning_rate": 0.0009480501626549945,
      "loss": 2.903,
      "step": 17000
    },
    {
      "epoch": 0.22077922077922077,
      "eval_loss": 3.2817673683166504,
      "eval_runtime": 16.3954,
      "eval_samples_per_second": 35.132,
      "eval_steps_per_second": 8.783,
      "step": 17000
    },
    {
      "epoch": 0.22207792207792207,
      "grad_norm": 0.22612017393112183,
      "learning_rate": 0.0009474083739088471,
      "loss": 2.91,
      "step": 17100
    },
    {
      "epoch": 0.22337662337662337,
      "grad_norm": 0.19862578809261322,
      "learning_rate": 0.0009467628651100745,
      "loss": 2.9319,
      "step": 17200
    },
    {
      "epoch": 0.22467532467532467,
      "grad_norm": 0.19449101388454437,
      "learning_rate": 0.0009461136416258689,
      "loss": 2.9244,
      "step": 17300
    },
    {
      "epoch": 0.22597402597402597,
      "grad_norm": 0.18568117916584015,
      "learning_rate": 0.0009454607088543088,
      "loss": 2.9118,
      "step": 17400
    },
    {
      "epoch": 0.22727272727272727,
      "grad_norm": 0.174198180437088,
      "learning_rate": 0.000944804072224314,
      "loss": 2.9134,
      "step": 17500
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 0.18078738451004028,
      "learning_rate": 0.0009441437371956013,
      "loss": 2.9039,
      "step": 17600
    },
    {
      "epoch": 0.22987012987012986,
      "grad_norm": 0.20317985117435455,
      "learning_rate": 0.0009434797092586377,
      "loss": 2.8848,
      "step": 17700
    },
    {
      "epoch": 0.23116883116883116,
      "grad_norm": 0.17165842652320862,
      "learning_rate": 0.0009428119939345959,
      "loss": 2.8846,
      "step": 17800
    },
    {
      "epoch": 0.23246753246753246,
      "grad_norm": 0.21153512597084045,
      "learning_rate": 0.0009421405967753078,
      "loss": 2.8944,
      "step": 17900
    },
    {
      "epoch": 0.23376623376623376,
      "grad_norm": 0.22300300002098083,
      "learning_rate": 0.000941465523363219,
      "loss": 2.8981,
      "step": 18000
    },
    {
      "epoch": 0.23376623376623376,
      "eval_loss": 3.2792882919311523,
      "eval_runtime": 15.157,
      "eval_samples_per_second": 38.002,
      "eval_steps_per_second": 9.501,
      "step": 18000
    },
    {
      "epoch": 0.23506493506493506,
      "grad_norm": 0.17719519138336182,
      "learning_rate": 0.0009407867793113414,
      "loss": 2.9096,
      "step": 18100
    },
    {
      "epoch": 0.23636363636363636,
      "grad_norm": 0.17834420502185822,
      "learning_rate": 0.0009401043702632071,
      "loss": 2.8922,
      "step": 18200
    },
    {
      "epoch": 0.23766233766233766,
      "grad_norm": 0.20472975075244904,
      "learning_rate": 0.000939418301892822,
      "loss": 2.8927,
      "step": 18300
    },
    {
      "epoch": 0.23896103896103896,
      "grad_norm": 0.1972654014825821,
      "learning_rate": 0.0009387285799046174,
      "loss": 2.904,
      "step": 18400
    },
    {
      "epoch": 0.24025974025974026,
      "grad_norm": 0.1871407926082611,
      "learning_rate": 0.0009380352100334035,
      "loss": 2.8829,
      "step": 18500
    },
    {
      "epoch": 0.24155844155844156,
      "grad_norm": 0.1807365119457245,
      "learning_rate": 0.0009373381980443212,
      "loss": 2.8977,
      "step": 18600
    },
    {
      "epoch": 0.24285714285714285,
      "grad_norm": 0.19222930073738098,
      "learning_rate": 0.000936637549732795,
      "loss": 2.8991,
      "step": 18700
    },
    {
      "epoch": 0.24415584415584415,
      "grad_norm": 0.2098565548658371,
      "learning_rate": 0.0009359332709244837,
      "loss": 2.8827,
      "step": 18800
    },
    {
      "epoch": 0.24545454545454545,
      "grad_norm": 0.2327863723039627,
      "learning_rate": 0.0009352253674752325,
      "loss": 2.8909,
      "step": 18900
    },
    {
      "epoch": 0.24675324675324675,
      "grad_norm": 0.1913413554430008,
      "learning_rate": 0.0009345138452710245,
      "loss": 2.8913,
      "step": 19000
    },
    {
      "epoch": 0.24675324675324675,
      "eval_loss": 3.2709906101226807,
      "eval_runtime": 14.4105,
      "eval_samples_per_second": 39.971,
      "eval_steps_per_second": 9.993,
      "step": 19000
    },
    {
      "epoch": 0.24805194805194805,
      "grad_norm": 0.17969286441802979,
      "learning_rate": 0.0009337987102279313,
      "loss": 2.8998,
      "step": 19100
    },
    {
      "epoch": 0.24935064935064935,
      "grad_norm": 0.18576957285404205,
      "learning_rate": 0.0009330799682920645,
      "loss": 2.8886,
      "step": 19200
    },
    {
      "epoch": 0.2506493506493506,
      "grad_norm": 0.17968888580799103,
      "learning_rate": 0.0009323576254395253,
      "loss": 2.8998,
      "step": 19300
    },
    {
      "epoch": 0.2519480519480519,
      "grad_norm": 0.1743050366640091,
      "learning_rate": 0.0009316316876763557,
      "loss": 2.8837,
      "step": 19400
    },
    {
      "epoch": 0.2532467532467532,
      "grad_norm": 0.17573101818561554,
      "learning_rate": 0.0009309021610384879,
      "loss": 2.8681,
      "step": 19500
    },
    {
      "epoch": 0.2545454545454545,
      "grad_norm": 0.199496790766716,
      "learning_rate": 0.0009301690515916948,
      "loss": 2.8717,
      "step": 19600
    },
    {
      "epoch": 0.2558441558441558,
      "grad_norm": 0.1846366673707962,
      "learning_rate": 0.0009294323654315387,
      "loss": 2.9166,
      "step": 19700
    },
    {
      "epoch": 0.2571428571428571,
      "grad_norm": 0.18776021897792816,
      "learning_rate": 0.0009286921086833214,
      "loss": 2.8916,
      "step": 19800
    },
    {
      "epoch": 0.2584415584415584,
      "grad_norm": 0.1786223202943802,
      "learning_rate": 0.000927948287502033,
      "loss": 2.9013,
      "step": 19900
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 0.18665452301502228,
      "learning_rate": 0.0009272009080723005,
      "loss": 2.9108,
      "step": 20000
    },
    {
      "epoch": 0.2597402597402597,
      "eval_loss": 3.2676138877868652,
      "eval_runtime": 15.0403,
      "eval_samples_per_second": 38.297,
      "eval_steps_per_second": 9.574,
      "step": 20000
    },
    {
      "epoch": 0.261038961038961,
      "grad_norm": 0.1809331625699997,
      "learning_rate": 0.0009264499766083365,
      "loss": 2.8836,
      "step": 20100
    },
    {
      "epoch": 0.2623376623376623,
      "grad_norm": 0.19160114228725433,
      "learning_rate": 0.0009256954993538877,
      "loss": 2.8881,
      "step": 20200
    },
    {
      "epoch": 0.2636363636363636,
      "grad_norm": 0.1818685233592987,
      "learning_rate": 0.0009249374825821832,
      "loss": 2.8543,
      "step": 20300
    },
    {
      "epoch": 0.2649350649350649,
      "grad_norm": 0.1778269112110138,
      "learning_rate": 0.0009241759325958815,
      "loss": 2.8741,
      "step": 20400
    },
    {
      "epoch": 0.2662337662337662,
      "grad_norm": 0.1888059824705124,
      "learning_rate": 0.0009234108557270187,
      "loss": 2.8772,
      "step": 20500
    },
    {
      "epoch": 0.2675324675324675,
      "grad_norm": 0.22666007280349731,
      "learning_rate": 0.000922642258336956,
      "loss": 2.9021,
      "step": 20600
    },
    {
      "epoch": 0.2688311688311688,
      "grad_norm": 0.18350407481193542,
      "learning_rate": 0.0009218701468163267,
      "loss": 2.8845,
      "step": 20700
    },
    {
      "epoch": 0.2701298701298701,
      "grad_norm": 0.17511753737926483,
      "learning_rate": 0.000921094527584982,
      "loss": 2.9005,
      "step": 20800
    },
    {
      "epoch": 0.2714285714285714,
      "grad_norm": 0.18808013200759888,
      "learning_rate": 0.0009203154070919398,
      "loss": 2.8658,
      "step": 20900
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 0.18752652406692505,
      "learning_rate": 0.0009195327918153292,
      "loss": 2.8649,
      "step": 21000
    },
    {
      "epoch": 0.2727272727272727,
      "eval_loss": 3.2498960494995117,
      "eval_runtime": 15.6914,
      "eval_samples_per_second": 36.708,
      "eval_steps_per_second": 9.177,
      "step": 21000
    },
    {
      "epoch": 0.274025974025974,
      "grad_norm": 0.19540126621723175,
      "learning_rate": 0.0009187466882623372,
      "loss": 2.8755,
      "step": 21100
    },
    {
      "epoch": 0.2753246753246753,
      "grad_norm": 0.18758288025856018,
      "learning_rate": 0.0009179571029691546,
      "loss": 2.8737,
      "step": 21200
    },
    {
      "epoch": 0.2766233766233766,
      "grad_norm": 0.18555940687656403,
      "learning_rate": 0.0009171640425009224,
      "loss": 2.8626,
      "step": 21300
    },
    {
      "epoch": 0.2779220779220779,
      "grad_norm": 0.18831279873847961,
      "learning_rate": 0.0009163675134516758,
      "loss": 2.883,
      "step": 21400
    },
    {
      "epoch": 0.2792207792207792,
      "grad_norm": 0.19656601548194885,
      "learning_rate": 0.0009155675224442904,
      "loss": 2.8959,
      "step": 21500
    },
    {
      "epoch": 0.2805194805194805,
      "grad_norm": 0.18778946995735168,
      "learning_rate": 0.0009147640761304266,
      "loss": 2.8723,
      "step": 21600
    },
    {
      "epoch": 0.2818181818181818,
      "grad_norm": 0.1975911557674408,
      "learning_rate": 0.000913957181190475,
      "loss": 2.8724,
      "step": 21700
    },
    {
      "epoch": 0.2831168831168831,
      "grad_norm": 0.18618452548980713,
      "learning_rate": 0.0009131468443334998,
      "loss": 2.8597,
      "step": 21800
    },
    {
      "epoch": 0.2844155844155844,
      "grad_norm": 0.19419926404953003,
      "learning_rate": 0.0009123330722971841,
      "loss": 2.8398,
      "step": 21900
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.1885763257741928,
      "learning_rate": 0.0009115158718477732,
      "loss": 2.875,
      "step": 22000
    },
    {
      "epoch": 0.2857142857142857,
      "eval_loss": 3.2455015182495117,
      "eval_runtime": 14.7497,
      "eval_samples_per_second": 39.052,
      "eval_steps_per_second": 9.763,
      "step": 22000
    },
    {
      "epoch": 0.287012987012987,
      "grad_norm": 0.17821335792541504,
      "learning_rate": 0.0009106952497800183,
      "loss": 2.8378,
      "step": 22100
    },
    {
      "epoch": 0.2883116883116883,
      "grad_norm": 0.20162396132946014,
      "learning_rate": 0.0009098712129171207,
      "loss": 2.8512,
      "step": 22200
    },
    {
      "epoch": 0.2896103896103896,
      "grad_norm": 0.21083980798721313,
      "learning_rate": 0.0009090437681106742,
      "loss": 2.8587,
      "step": 22300
    },
    {
      "epoch": 0.2909090909090909,
      "grad_norm": 0.1923542320728302,
      "learning_rate": 0.0009082129222406087,
      "loss": 2.843,
      "step": 22400
    },
    {
      "epoch": 0.2922077922077922,
      "grad_norm": 0.19237105548381805,
      "learning_rate": 0.0009073786822151326,
      "loss": 2.8554,
      "step": 22500
    },
    {
      "epoch": 0.2935064935064935,
      "grad_norm": 0.1831800788640976,
      "learning_rate": 0.000906541054970676,
      "loss": 2.8928,
      "step": 22600
    },
    {
      "epoch": 0.2948051948051948,
      "grad_norm": 0.20052975416183472,
      "learning_rate": 0.000905700047471832,
      "loss": 2.8462,
      "step": 22700
    },
    {
      "epoch": 0.2961038961038961,
      "grad_norm": 0.20518037676811218,
      "learning_rate": 0.0009048556667113002,
      "loss": 2.8624,
      "step": 22800
    },
    {
      "epoch": 0.2974025974025974,
      "grad_norm": 0.18524277210235596,
      "learning_rate": 0.0009040079197098268,
      "loss": 2.8683,
      "step": 22900
    },
    {
      "epoch": 0.2987012987012987,
      "grad_norm": 0.18116071820259094,
      "learning_rate": 0.000903156813516148,
      "loss": 2.8335,
      "step": 23000
    },
    {
      "epoch": 0.2987012987012987,
      "eval_loss": 3.236686944961548,
      "eval_runtime": 16.906,
      "eval_samples_per_second": 34.071,
      "eval_steps_per_second": 8.518,
      "step": 23000
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.2058299481868744,
      "learning_rate": 0.0009023023552069303,
      "loss": 2.8548,
      "step": 23100
    },
    {
      "epoch": 0.3012987012987013,
      "grad_norm": 0.21021772921085358,
      "learning_rate": 0.0009014445518867116,
      "loss": 2.8435,
      "step": 23200
    },
    {
      "epoch": 0.3025974025974026,
      "grad_norm": 0.18895122408866882,
      "learning_rate": 0.000900583410687843,
      "loss": 2.839,
      "step": 23300
    },
    {
      "epoch": 0.3038961038961039,
      "grad_norm": 0.18983018398284912,
      "learning_rate": 0.0008997189387704286,
      "loss": 2.843,
      "step": 23400
    },
    {
      "epoch": 0.3051948051948052,
      "grad_norm": 0.21453820168972015,
      "learning_rate": 0.0008988511433222664,
      "loss": 2.8443,
      "step": 23500
    },
    {
      "epoch": 0.3064935064935065,
      "grad_norm": 0.1873568445444107,
      "learning_rate": 0.0008979800315587885,
      "loss": 2.8473,
      "step": 23600
    },
    {
      "epoch": 0.3077922077922078,
      "grad_norm": 0.19516156613826752,
      "learning_rate": 0.000897105610723001,
      "loss": 2.8579,
      "step": 23700
    },
    {
      "epoch": 0.3090909090909091,
      "grad_norm": 0.1961786150932312,
      "learning_rate": 0.000896227888085424,
      "loss": 2.8282,
      "step": 23800
    },
    {
      "epoch": 0.3103896103896104,
      "grad_norm": 0.19147805869579315,
      "learning_rate": 0.0008953468709440309,
      "loss": 2.8734,
      "step": 23900
    },
    {
      "epoch": 0.3116883116883117,
      "grad_norm": 0.20040926337242126,
      "learning_rate": 0.0008944625666241877,
      "loss": 2.8491,
      "step": 24000
    },
    {
      "epoch": 0.3116883116883117,
      "eval_loss": 3.2374658584594727,
      "eval_runtime": 16.1905,
      "eval_samples_per_second": 35.576,
      "eval_steps_per_second": 8.894,
      "step": 24000
    },
    {
      "epoch": 0.312987012987013,
      "grad_norm": 0.21797960996627808,
      "learning_rate": 0.0008935749824785922,
      "loss": 2.8992,
      "step": 24100
    },
    {
      "epoch": 0.3142857142857143,
      "grad_norm": 0.19554930925369263,
      "learning_rate": 0.0008926841258872131,
      "loss": 2.8771,
      "step": 24200
    },
    {
      "epoch": 0.3155844155844156,
      "grad_norm": 0.18446357548236847,
      "learning_rate": 0.0008917900042572283,
      "loss": 2.8753,
      "step": 24300
    },
    {
      "epoch": 0.3168831168831169,
      "grad_norm": 0.18406012654304504,
      "learning_rate": 0.0008908926250229633,
      "loss": 2.8819,
      "step": 24400
    },
    {
      "epoch": 0.3181818181818182,
      "grad_norm": 0.21075041592121124,
      "learning_rate": 0.0008899919956458296,
      "loss": 2.8796,
      "step": 24500
    },
    {
      "epoch": 0.3194805194805195,
      "grad_norm": 0.20150282979011536,
      "learning_rate": 0.0008890881236142624,
      "loss": 2.8611,
      "step": 24600
    },
    {
      "epoch": 0.3207792207792208,
      "grad_norm": 0.43753939867019653,
      "learning_rate": 0.0008881810164436588,
      "loss": 2.8587,
      "step": 24700
    },
    {
      "epoch": 0.3220779220779221,
      "grad_norm": 0.1800195425748825,
      "learning_rate": 0.0008872706816763148,
      "loss": 2.8797,
      "step": 24800
    },
    {
      "epoch": 0.3233766233766234,
      "grad_norm": 0.2071843296289444,
      "learning_rate": 0.0008863571268813628,
      "loss": 2.8697,
      "step": 24900
    },
    {
      "epoch": 0.3246753246753247,
      "grad_norm": 0.19320254027843475,
      "learning_rate": 0.0008854403596547088,
      "loss": 2.8938,
      "step": 25000
    },
    {
      "epoch": 0.3246753246753247,
      "eval_loss": 3.232848882675171,
      "eval_runtime": 15.6644,
      "eval_samples_per_second": 36.771,
      "eval_steps_per_second": 9.193,
      "step": 25000
    },
    {
      "epoch": 0.32597402597402597,
      "grad_norm": 0.21449397504329681,
      "learning_rate": 0.000884520387618969,
      "loss": 2.8374,
      "step": 25100
    },
    {
      "epoch": 0.32727272727272727,
      "grad_norm": 0.2025729864835739,
      "learning_rate": 0.0008835972184234066,
      "loss": 2.845,
      "step": 25200
    },
    {
      "epoch": 0.32857142857142857,
      "grad_norm": 0.22090762853622437,
      "learning_rate": 0.000882670859743868,
      "loss": 2.8518,
      "step": 25300
    },
    {
      "epoch": 0.32987012987012987,
      "grad_norm": 0.20626656711101532,
      "learning_rate": 0.0008817413192827191,
      "loss": 2.8299,
      "step": 25400
    },
    {
      "epoch": 0.33116883116883117,
      "grad_norm": 0.18527238070964813,
      "learning_rate": 0.0008808086047687813,
      "loss": 2.8532,
      "step": 25500
    },
    {
      "epoch": 0.33246753246753247,
      "grad_norm": 0.20959258079528809,
      "learning_rate": 0.0008798727239572676,
      "loss": 2.8656,
      "step": 25600
    },
    {
      "epoch": 0.33376623376623377,
      "grad_norm": 0.2122378945350647,
      "learning_rate": 0.000878933684629717,
      "loss": 2.8532,
      "step": 25700
    },
    {
      "epoch": 0.33506493506493507,
      "grad_norm": 0.18408456444740295,
      "learning_rate": 0.0008779914945939311,
      "loss": 2.8639,
      "step": 25800
    },
    {
      "epoch": 0.33636363636363636,
      "grad_norm": 0.20426695048809052,
      "learning_rate": 0.0008770461616839082,
      "loss": 2.865,
      "step": 25900
    },
    {
      "epoch": 0.33766233766233766,
      "grad_norm": 0.1808163821697235,
      "learning_rate": 0.0008760976937597787,
      "loss": 2.8481,
      "step": 26000
    },
    {
      "epoch": 0.33766233766233766,
      "eval_loss": 3.214432716369629,
      "eval_runtime": 15.0454,
      "eval_samples_per_second": 38.284,
      "eval_steps_per_second": 9.571,
      "step": 26000
    },
    {
      "epoch": 0.33896103896103896,
      "grad_norm": 0.20782601833343506,
      "learning_rate": 0.0008751460987077398,
      "loss": 2.8405,
      "step": 26100
    },
    {
      "epoch": 0.34025974025974026,
      "grad_norm": 0.2266394942998886,
      "learning_rate": 0.0008741913844399896,
      "loss": 2.8304,
      "step": 26200
    },
    {
      "epoch": 0.34155844155844156,
      "grad_norm": 0.215355783700943,
      "learning_rate": 0.0008732335588946612,
      "loss": 2.8749,
      "step": 26300
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 0.17590366303920746,
      "learning_rate": 0.0008722726300357573,
      "loss": 2.8542,
      "step": 26400
    },
    {
      "epoch": 0.34415584415584416,
      "grad_norm": 0.19376814365386963,
      "learning_rate": 0.0008713086058530832,
      "loss": 2.8636,
      "step": 26500
    },
    {
      "epoch": 0.34545454545454546,
      "grad_norm": 0.19086168706417084,
      "learning_rate": 0.0008703414943621817,
      "loss": 2.8752,
      "step": 26600
    },
    {
      "epoch": 0.34675324675324676,
      "grad_norm": 0.19386428594589233,
      "learning_rate": 0.0008693713036042643,
      "loss": 2.8595,
      "step": 26700
    },
    {
      "epoch": 0.34805194805194806,
      "grad_norm": 0.20364010334014893,
      "learning_rate": 0.0008683980416461465,
      "loss": 2.8322,
      "step": 26800
    },
    {
      "epoch": 0.34935064935064936,
      "grad_norm": 0.20115472376346588,
      "learning_rate": 0.0008674217165801797,
      "loss": 2.8271,
      "step": 26900
    },
    {
      "epoch": 0.35064935064935066,
      "grad_norm": 0.19452853500843048,
      "learning_rate": 0.0008664423365241833,
      "loss": 2.8445,
      "step": 27000
    },
    {
      "epoch": 0.35064935064935066,
      "eval_loss": 3.206655263900757,
      "eval_runtime": 14.9037,
      "eval_samples_per_second": 38.648,
      "eval_steps_per_second": 9.662,
      "step": 27000
    },
    {
      "epoch": 0.35194805194805195,
      "grad_norm": 0.29245325922966003,
      "learning_rate": 0.0008654599096213792,
      "loss": 2.8584,
      "step": 27100
    },
    {
      "epoch": 0.35324675324675325,
      "grad_norm": 0.1941668838262558,
      "learning_rate": 0.0008644744440403214,
      "loss": 2.8146,
      "step": 27200
    },
    {
      "epoch": 0.35454545454545455,
      "grad_norm": 0.20218811929225922,
      "learning_rate": 0.0008634859479748306,
      "loss": 2.8623,
      "step": 27300
    },
    {
      "epoch": 0.35584415584415585,
      "grad_norm": 0.19882351160049438,
      "learning_rate": 0.0008624944296439246,
      "loss": 2.8491,
      "step": 27400
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.19472059607505798,
      "learning_rate": 0.0008614998972917503,
      "loss": 2.8372,
      "step": 27500
    },
    {
      "epoch": 0.35844155844155845,
      "grad_norm": 0.20453399419784546,
      "learning_rate": 0.000860502359187515,
      "loss": 2.8228,
      "step": 27600
    },
    {
      "epoch": 0.35974025974025975,
      "grad_norm": 0.20085613429546356,
      "learning_rate": 0.0008595018236254182,
      "loss": 2.8416,
      "step": 27700
    },
    {
      "epoch": 0.36103896103896105,
      "grad_norm": 0.1953079253435135,
      "learning_rate": 0.0008584982989245822,
      "loss": 2.844,
      "step": 27800
    },
    {
      "epoch": 0.36233766233766235,
      "grad_norm": 0.18861424922943115,
      "learning_rate": 0.0008574917934289829,
      "loss": 2.8586,
      "step": 27900
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.2032879889011383,
      "learning_rate": 0.0008564823155073804,
      "loss": 2.8676,
      "step": 28000
    },
    {
      "epoch": 0.36363636363636365,
      "eval_loss": 3.213535785675049,
      "eval_runtime": 15.9519,
      "eval_samples_per_second": 36.109,
      "eval_steps_per_second": 9.027,
      "step": 28000
    },
    {
      "epoch": 0.36493506493506495,
      "grad_norm": 0.20181582868099213,
      "learning_rate": 0.00085546987355325,
      "loss": 2.835,
      "step": 28100
    },
    {
      "epoch": 0.36623376623376624,
      "grad_norm": 0.1909325122833252,
      "learning_rate": 0.0008544544759847112,
      "loss": 2.8291,
      "step": 28200
    },
    {
      "epoch": 0.36753246753246754,
      "grad_norm": 0.21319599449634552,
      "learning_rate": 0.000853436131244459,
      "loss": 2.8536,
      "step": 28300
    },
    {
      "epoch": 0.36883116883116884,
      "grad_norm": 0.200112447142601,
      "learning_rate": 0.0008524148477996933,
      "loss": 2.8732,
      "step": 28400
    },
    {
      "epoch": 0.37012987012987014,
      "grad_norm": 0.19826772809028625,
      "learning_rate": 0.0008513906341420478,
      "loss": 2.8208,
      "step": 28500
    },
    {
      "epoch": 0.37142857142857144,
      "grad_norm": 0.21064135432243347,
      "learning_rate": 0.0008503634987875206,
      "loss": 2.805,
      "step": 28600
    },
    {
      "epoch": 0.37272727272727274,
      "grad_norm": 0.2014266848564148,
      "learning_rate": 0.0008493334502764021,
      "loss": 2.8257,
      "step": 28700
    },
    {
      "epoch": 0.37402597402597404,
      "grad_norm": 0.21088480949401855,
      "learning_rate": 0.0008483004971732049,
      "loss": 2.8261,
      "step": 28800
    },
    {
      "epoch": 0.37532467532467534,
      "grad_norm": 0.20050457119941711,
      "learning_rate": 0.0008472646480665924,
      "loss": 2.8694,
      "step": 28900
    },
    {
      "epoch": 0.37662337662337664,
      "grad_norm": 0.21420925855636597,
      "learning_rate": 0.0008462259115693076,
      "loss": 2.8473,
      "step": 29000
    },
    {
      "epoch": 0.37662337662337664,
      "eval_loss": 3.195821523666382,
      "eval_runtime": 15.2121,
      "eval_samples_per_second": 37.865,
      "eval_steps_per_second": 9.466,
      "step": 29000
    },
    {
      "epoch": 0.37792207792207794,
      "grad_norm": 0.2159246951341629,
      "learning_rate": 0.0008451842963181003,
      "loss": 2.8032,
      "step": 29100
    },
    {
      "epoch": 0.37922077922077924,
      "grad_norm": 0.19643986225128174,
      "learning_rate": 0.0008441398109736571,
      "loss": 2.847,
      "step": 29200
    },
    {
      "epoch": 0.38051948051948054,
      "grad_norm": 0.1965179294347763,
      "learning_rate": 0.0008430924642205279,
      "loss": 2.8529,
      "step": 29300
    },
    {
      "epoch": 0.38181818181818183,
      "grad_norm": 0.21244043111801147,
      "learning_rate": 0.0008420422647670548,
      "loss": 2.8067,
      "step": 29400
    },
    {
      "epoch": 0.38311688311688313,
      "grad_norm": 0.20399615168571472,
      "learning_rate": 0.0008409892213452987,
      "loss": 2.8176,
      "step": 29500
    },
    {
      "epoch": 0.38441558441558443,
      "grad_norm": 0.20329082012176514,
      "learning_rate": 0.0008399333427109672,
      "loss": 2.8316,
      "step": 29600
    },
    {
      "epoch": 0.38571428571428573,
      "grad_norm": 0.19244763255119324,
      "learning_rate": 0.0008388746376433419,
      "loss": 2.8367,
      "step": 29700
    },
    {
      "epoch": 0.38701298701298703,
      "grad_norm": 0.18999354541301727,
      "learning_rate": 0.0008378131149452053,
      "loss": 2.8212,
      "step": 29800
    },
    {
      "epoch": 0.38831168831168833,
      "grad_norm": 0.20788761973381042,
      "learning_rate": 0.0008367487834427674,
      "loss": 2.7999,
      "step": 29900
    },
    {
      "epoch": 0.38961038961038963,
      "grad_norm": 0.21678058803081512,
      "learning_rate": 0.0008356816519855926,
      "loss": 2.81,
      "step": 30000
    },
    {
      "epoch": 0.38961038961038963,
      "eval_loss": 3.1891965866088867,
      "eval_runtime": 15.5166,
      "eval_samples_per_second": 37.122,
      "eval_steps_per_second": 9.28,
      "step": 30000
    },
    {
      "epoch": 0.39090909090909093,
      "grad_norm": 0.25863537192344666,
      "learning_rate": 0.0008346117294465258,
      "loss": 2.7919,
      "step": 30100
    },
    {
      "epoch": 0.3922077922077922,
      "grad_norm": 0.2082231491804123,
      "learning_rate": 0.0008335390247216193,
      "loss": 2.8058,
      "step": 30200
    },
    {
      "epoch": 0.3935064935064935,
      "grad_norm": 0.2484658658504486,
      "learning_rate": 0.0008324635467300578,
      "loss": 2.7913,
      "step": 30300
    },
    {
      "epoch": 0.3948051948051948,
      "grad_norm": 0.19809049367904663,
      "learning_rate": 0.000831385304414085,
      "loss": 2.8392,
      "step": 30400
    },
    {
      "epoch": 0.3961038961038961,
      "grad_norm": 0.19220763444900513,
      "learning_rate": 0.0008303043067389293,
      "loss": 2.8326,
      "step": 30500
    },
    {
      "epoch": 0.3974025974025974,
      "grad_norm": 0.19120201468467712,
      "learning_rate": 0.0008292205626927285,
      "loss": 2.8008,
      "step": 30600
    },
    {
      "epoch": 0.3987012987012987,
      "grad_norm": 0.2093675136566162,
      "learning_rate": 0.000828134081286456,
      "loss": 2.8228,
      "step": 30700
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.2077227681875229,
      "learning_rate": 0.0008270448715538452,
      "loss": 2.8263,
      "step": 30800
    },
    {
      "epoch": 0.4012987012987013,
      "grad_norm": 0.2050541490316391,
      "learning_rate": 0.0008259529425513148,
      "loss": 2.8289,
      "step": 30900
    },
    {
      "epoch": 0.4025974025974026,
      "grad_norm": 0.19998504221439362,
      "learning_rate": 0.0008248583033578932,
      "loss": 2.8068,
      "step": 31000
    },
    {
      "epoch": 0.4025974025974026,
      "eval_loss": 3.1833224296569824,
      "eval_runtime": 11.0521,
      "eval_samples_per_second": 52.117,
      "eval_steps_per_second": 13.029,
      "step": 31000
    },
    {
      "epoch": 0.4038961038961039,
      "grad_norm": 0.20032985508441925,
      "learning_rate": 0.0008237609630751433,
      "loss": 2.8321,
      "step": 31100
    },
    {
      "epoch": 0.4051948051948052,
      "grad_norm": 0.2228243052959442,
      "learning_rate": 0.0008226609308270862,
      "loss": 2.8323,
      "step": 31200
    },
    {
      "epoch": 0.4064935064935065,
      "grad_norm": 0.22431324422359467,
      "learning_rate": 0.0008215582157601267,
      "loss": 2.8489,
      "step": 31300
    },
    {
      "epoch": 0.4077922077922078,
      "grad_norm": 0.20516513288021088,
      "learning_rate": 0.0008204528270429752,
      "loss": 2.8001,
      "step": 31400
    },
    {
      "epoch": 0.4090909090909091,
      "grad_norm": 0.20168229937553406,
      "learning_rate": 0.0008193447738665735,
      "loss": 2.8463,
      "step": 31500
    },
    {
      "epoch": 0.4103896103896104,
      "grad_norm": 0.20102174580097198,
      "learning_rate": 0.0008182340654440173,
      "loss": 2.8226,
      "step": 31600
    },
    {
      "epoch": 0.4116883116883117,
      "grad_norm": 0.2369510680437088,
      "learning_rate": 0.0008171207110104797,
      "loss": 2.8284,
      "step": 31700
    },
    {
      "epoch": 0.412987012987013,
      "grad_norm": 0.18998616933822632,
      "learning_rate": 0.0008160047198231344,
      "loss": 2.8506,
      "step": 31800
    },
    {
      "epoch": 0.4142857142857143,
      "grad_norm": 0.21406228840351105,
      "learning_rate": 0.000814886101161079,
      "loss": 2.81,
      "step": 31900
    },
    {
      "epoch": 0.4155844155844156,
      "grad_norm": 0.20903189480304718,
      "learning_rate": 0.0008137648643252575,
      "loss": 2.7976,
      "step": 32000
    },
    {
      "epoch": 0.4155844155844156,
      "eval_loss": 3.1785833835601807,
      "eval_runtime": 15.2674,
      "eval_samples_per_second": 37.727,
      "eval_steps_per_second": 9.432,
      "step": 32000
    },
    {
      "epoch": 0.41688311688311686,
      "grad_norm": 0.22132746875286102,
      "learning_rate": 0.0008126410186383836,
      "loss": 2.8195,
      "step": 32100
    },
    {
      "epoch": 0.41818181818181815,
      "grad_norm": 0.21235358715057373,
      "learning_rate": 0.0008115145734448622,
      "loss": 2.7824,
      "step": 32200
    },
    {
      "epoch": 0.41948051948051945,
      "grad_norm": 0.21031178534030914,
      "learning_rate": 0.0008103855381107122,
      "loss": 2.8135,
      "step": 32300
    },
    {
      "epoch": 0.42077922077922075,
      "grad_norm": 0.2186926305294037,
      "learning_rate": 0.0008092539220234894,
      "loss": 2.8166,
      "step": 32400
    },
    {
      "epoch": 0.42207792207792205,
      "grad_norm": 0.20855864882469177,
      "learning_rate": 0.0008081197345922069,
      "loss": 2.8595,
      "step": 32500
    },
    {
      "epoch": 0.42337662337662335,
      "grad_norm": 0.20601770281791687,
      "learning_rate": 0.0008069829852472583,
      "loss": 2.8201,
      "step": 32600
    },
    {
      "epoch": 0.42467532467532465,
      "grad_norm": 0.22806823253631592,
      "learning_rate": 0.000805843683440338,
      "loss": 2.8265,
      "step": 32700
    },
    {
      "epoch": 0.42597402597402595,
      "grad_norm": 0.21018485724925995,
      "learning_rate": 0.0008047018386443638,
      "loss": 2.8403,
      "step": 32800
    },
    {
      "epoch": 0.42727272727272725,
      "grad_norm": 0.218631774187088,
      "learning_rate": 0.0008035574603533975,
      "loss": 2.8052,
      "step": 32900
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.20622466504573822,
      "learning_rate": 0.000802410558082566,
      "loss": 2.824,
      "step": 33000
    },
    {
      "epoch": 0.42857142857142855,
      "eval_loss": 3.178421974182129,
      "eval_runtime": 15.6887,
      "eval_samples_per_second": 36.714,
      "eval_steps_per_second": 9.179,
      "step": 33000
    },
    {
      "epoch": 0.42987012987012985,
      "grad_norm": 0.24304358661174774,
      "learning_rate": 0.0008012611413679824,
      "loss": 2.8068,
      "step": 33100
    },
    {
      "epoch": 0.43116883116883115,
      "grad_norm": 0.20239757001399994,
      "learning_rate": 0.0008001092197666661,
      "loss": 2.7936,
      "step": 33200
    },
    {
      "epoch": 0.43246753246753245,
      "grad_norm": 0.21365991234779358,
      "learning_rate": 0.0007989548028564646,
      "loss": 2.8053,
      "step": 33300
    },
    {
      "epoch": 0.43376623376623374,
      "grad_norm": 0.19812121987342834,
      "learning_rate": 0.0007977979002359723,
      "loss": 2.8083,
      "step": 33400
    },
    {
      "epoch": 0.43506493506493504,
      "grad_norm": 0.20705953240394592,
      "learning_rate": 0.0007966385215244518,
      "loss": 2.7882,
      "step": 33500
    },
    {
      "epoch": 0.43636363636363634,
      "grad_norm": 0.23277799785137177,
      "learning_rate": 0.0007954766763617538,
      "loss": 2.8241,
      "step": 33600
    },
    {
      "epoch": 0.43766233766233764,
      "grad_norm": 0.2175544649362564,
      "learning_rate": 0.0007943123744082363,
      "loss": 2.816,
      "step": 33700
    },
    {
      "epoch": 0.43896103896103894,
      "grad_norm": 0.21744564175605774,
      "learning_rate": 0.000793145625344685,
      "loss": 2.7947,
      "step": 33800
    },
    {
      "epoch": 0.44025974025974024,
      "grad_norm": 0.2085115909576416,
      "learning_rate": 0.0007919764388722322,
      "loss": 2.8016,
      "step": 33900
    },
    {
      "epoch": 0.44155844155844154,
      "grad_norm": 0.21085582673549652,
      "learning_rate": 0.0007908048247122768,
      "loss": 2.7845,
      "step": 34000
    },
    {
      "epoch": 0.44155844155844154,
      "eval_loss": 3.1732594966888428,
      "eval_runtime": 14.9965,
      "eval_samples_per_second": 38.409,
      "eval_steps_per_second": 9.602,
      "step": 34000
    },
    {
      "epoch": 0.44285714285714284,
      "grad_norm": 0.1986636519432068,
      "learning_rate": 0.0007896307926064029,
      "loss": 2.8039,
      "step": 34100
    },
    {
      "epoch": 0.44415584415584414,
      "grad_norm": 0.2274903804063797,
      "learning_rate": 0.0007884543523162991,
      "loss": 2.7982,
      "step": 34200
    },
    {
      "epoch": 0.44545454545454544,
      "grad_norm": 0.23180291056632996,
      "learning_rate": 0.0007872755136236774,
      "loss": 2.8223,
      "step": 34300
    },
    {
      "epoch": 0.44675324675324674,
      "grad_norm": 0.21269173920154572,
      "learning_rate": 0.0007860942863301914,
      "loss": 2.8141,
      "step": 34400
    },
    {
      "epoch": 0.44805194805194803,
      "grad_norm": 0.20867326855659485,
      "learning_rate": 0.0007849106802573553,
      "loss": 2.7805,
      "step": 34500
    },
    {
      "epoch": 0.44935064935064933,
      "grad_norm": 1.6756339073181152,
      "learning_rate": 0.0007837247052464621,
      "loss": 2.7879,
      "step": 34600
    },
    {
      "epoch": 0.45064935064935063,
      "grad_norm": 0.24874623119831085,
      "learning_rate": 0.0007825363711585016,
      "loss": 2.8281,
      "step": 34700
    },
    {
      "epoch": 0.45194805194805193,
      "grad_norm": 0.19419661164283752,
      "learning_rate": 0.0007813456878740789,
      "loss": 2.7984,
      "step": 34800
    },
    {
      "epoch": 0.45324675324675323,
      "grad_norm": 0.21280324459075928,
      "learning_rate": 0.0007801526652933313,
      "loss": 2.834,
      "step": 34900
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.20888017117977142,
      "learning_rate": 0.0007789573133358471,
      "loss": 2.831,
      "step": 35000
    },
    {
      "epoch": 0.45454545454545453,
      "eval_loss": 3.1828813552856445,
      "eval_runtime": 14.929,
      "eval_samples_per_second": 38.583,
      "eval_steps_per_second": 9.646,
      "step": 35000
    },
    {
      "epoch": 0.45584415584415583,
      "grad_norm": 0.21340541541576385,
      "learning_rate": 0.0007777596419405823,
      "loss": 2.8065,
      "step": 35100
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 0.20886409282684326,
      "learning_rate": 0.0007765596610657783,
      "loss": 2.8423,
      "step": 35200
    },
    {
      "epoch": 0.4584415584415584,
      "grad_norm": 0.21619443595409393,
      "learning_rate": 0.0007753573806888795,
      "loss": 2.786,
      "step": 35300
    },
    {
      "epoch": 0.4597402597402597,
      "grad_norm": 0.21694345772266388,
      "learning_rate": 0.0007741528108064491,
      "loss": 2.7822,
      "step": 35400
    },
    {
      "epoch": 0.461038961038961,
      "grad_norm": 0.21181565523147583,
      "learning_rate": 0.0007729459614340872,
      "loss": 2.7599,
      "step": 35500
    },
    {
      "epoch": 0.4623376623376623,
      "grad_norm": 0.22892408072948456,
      "learning_rate": 0.0007717368426063475,
      "loss": 2.7934,
      "step": 35600
    },
    {
      "epoch": 0.4636363636363636,
      "grad_norm": 0.21864919364452362,
      "learning_rate": 0.0007705254643766527,
      "loss": 2.8059,
      "step": 35700
    },
    {
      "epoch": 0.4649350649350649,
      "grad_norm": 0.2145598828792572,
      "learning_rate": 0.000769311836817212,
      "loss": 2.8056,
      "step": 35800
    },
    {
      "epoch": 0.4662337662337662,
      "grad_norm": 0.21081195771694183,
      "learning_rate": 0.0007680959700189375,
      "loss": 2.7969,
      "step": 35900
    },
    {
      "epoch": 0.4675324675324675,
      "grad_norm": 0.22019855678081512,
      "learning_rate": 0.0007668778740913591,
      "loss": 2.7923,
      "step": 36000
    },
    {
      "epoch": 0.4675324675324675,
      "eval_loss": 3.161719560623169,
      "eval_runtime": 15.4038,
      "eval_samples_per_second": 37.393,
      "eval_steps_per_second": 9.348,
      "step": 36000
    },
    {
      "epoch": 0.4688311688311688,
      "grad_norm": 0.22442568838596344,
      "learning_rate": 0.0007656575591625415,
      "loss": 2.7954,
      "step": 36100
    },
    {
      "epoch": 0.4701298701298701,
      "grad_norm": 0.2147638201713562,
      "learning_rate": 0.0007644350353789997,
      "loss": 2.7897,
      "step": 36200
    },
    {
      "epoch": 0.4714285714285714,
      "grad_norm": 0.20484313368797302,
      "learning_rate": 0.0007632103129056145,
      "loss": 2.7859,
      "step": 36300
    },
    {
      "epoch": 0.4727272727272727,
      "grad_norm": 0.19691170752048492,
      "learning_rate": 0.0007619834019255482,
      "loss": 2.8251,
      "step": 36400
    },
    {
      "epoch": 0.474025974025974,
      "grad_norm": 0.2177974134683609,
      "learning_rate": 0.0007607543126401597,
      "loss": 2.7771,
      "step": 36500
    },
    {
      "epoch": 0.4753246753246753,
      "grad_norm": 0.24122121930122375,
      "learning_rate": 0.0007595230552689201,
      "loss": 2.7973,
      "step": 36600
    },
    {
      "epoch": 0.4766233766233766,
      "grad_norm": 0.2524343729019165,
      "learning_rate": 0.0007582896400493266,
      "loss": 2.82,
      "step": 36700
    },
    {
      "epoch": 0.4779220779220779,
      "grad_norm": 0.21695558726787567,
      "learning_rate": 0.000757054077236819,
      "loss": 2.802,
      "step": 36800
    },
    {
      "epoch": 0.4792207792207792,
      "grad_norm": 0.2205599695444107,
      "learning_rate": 0.0007558163771046934,
      "loss": 2.7509,
      "step": 36900
    },
    {
      "epoch": 0.4805194805194805,
      "grad_norm": 0.2173157036304474,
      "learning_rate": 0.0007545765499440169,
      "loss": 2.7978,
      "step": 37000
    },
    {
      "epoch": 0.4805194805194805,
      "eval_loss": 3.170949935913086,
      "eval_runtime": 16.3367,
      "eval_samples_per_second": 35.258,
      "eval_steps_per_second": 8.815,
      "step": 37000
    },
    {
      "epoch": 0.4818181818181818,
      "grad_norm": 0.21395540237426758,
      "learning_rate": 0.0007533346060635424,
      "loss": 2.775,
      "step": 37100
    },
    {
      "epoch": 0.4831168831168831,
      "grad_norm": 0.2262965440750122,
      "learning_rate": 0.0007520905557896221,
      "loss": 2.779,
      "step": 37200
    },
    {
      "epoch": 0.4844155844155844,
      "grad_norm": 0.2124248743057251,
      "learning_rate": 0.0007508444094661227,
      "loss": 2.8141,
      "step": 37300
    },
    {
      "epoch": 0.4857142857142857,
      "grad_norm": 0.24149860441684723,
      "learning_rate": 0.0007495961774543385,
      "loss": 2.8091,
      "step": 37400
    },
    {
      "epoch": 0.487012987012987,
      "grad_norm": 0.2356308251619339,
      "learning_rate": 0.0007483458701329059,
      "loss": 2.7898,
      "step": 37500
    },
    {
      "epoch": 0.4883116883116883,
      "grad_norm": 0.21705743670463562,
      "learning_rate": 0.0007470934978977167,
      "loss": 2.8022,
      "step": 37600
    },
    {
      "epoch": 0.4896103896103896,
      "grad_norm": 0.21993373334407806,
      "learning_rate": 0.0007458390711618315,
      "loss": 2.7881,
      "step": 37700
    },
    {
      "epoch": 0.4909090909090909,
      "grad_norm": 0.21296945214271545,
      "learning_rate": 0.0007445826003553939,
      "loss": 2.7987,
      "step": 37800
    },
    {
      "epoch": 0.4922077922077922,
      "grad_norm": 0.23273344337940216,
      "learning_rate": 0.000743324095925543,
      "loss": 2.8253,
      "step": 37900
    },
    {
      "epoch": 0.4935064935064935,
      "grad_norm": 0.2104436457157135,
      "learning_rate": 0.0007420635683363268,
      "loss": 2.7807,
      "step": 38000
    },
    {
      "epoch": 0.4935064935064935,
      "eval_loss": 3.151848316192627,
      "eval_runtime": 15.3437,
      "eval_samples_per_second": 37.54,
      "eval_steps_per_second": 9.385,
      "step": 38000
    },
    {
      "epoch": 0.4948051948051948,
      "grad_norm": 0.2078818380832672,
      "learning_rate": 0.0007408010280686151,
      "loss": 2.7504,
      "step": 38100
    },
    {
      "epoch": 0.4961038961038961,
      "grad_norm": 0.22127191722393036,
      "learning_rate": 0.0007395364856200127,
      "loss": 2.7698,
      "step": 38200
    },
    {
      "epoch": 0.4974025974025974,
      "grad_norm": 0.20785515010356903,
      "learning_rate": 0.0007382699515047715,
      "loss": 2.804,
      "step": 38300
    },
    {
      "epoch": 0.4987012987012987,
      "grad_norm": 0.267839640378952,
      "learning_rate": 0.0007370014362537041,
      "loss": 2.778,
      "step": 38400
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.22827385365962982,
      "learning_rate": 0.0007357309504140948,
      "loss": 2.7786,
      "step": 38500
    },
    {
      "epoch": 0.5012987012987012,
      "grad_norm": 0.22568035125732422,
      "learning_rate": 0.0007344585045496133,
      "loss": 2.7655,
      "step": 38600
    },
    {
      "epoch": 0.5025974025974026,
      "grad_norm": 0.2355123907327652,
      "learning_rate": 0.000733184109240226,
      "loss": 2.7714,
      "step": 38700
    },
    {
      "epoch": 0.5038961038961038,
      "grad_norm": 0.2227976769208908,
      "learning_rate": 0.0007319077750821084,
      "loss": 2.7534,
      "step": 38800
    },
    {
      "epoch": 0.5051948051948052,
      "grad_norm": 0.25165653228759766,
      "learning_rate": 0.0007306295126875566,
      "loss": 2.7832,
      "step": 38900
    },
    {
      "epoch": 0.5064935064935064,
      "grad_norm": 0.24306100606918335,
      "learning_rate": 0.0007293493326848997,
      "loss": 2.774,
      "step": 39000
    },
    {
      "epoch": 0.5064935064935064,
      "eval_loss": 3.1421139240264893,
      "eval_runtime": 15.2375,
      "eval_samples_per_second": 37.801,
      "eval_steps_per_second": 9.45,
      "step": 39000
    },
    {
      "epoch": 0.5077922077922078,
      "grad_norm": 0.23936377465724945,
      "learning_rate": 0.0007280672457184108,
      "loss": 2.7569,
      "step": 39100
    },
    {
      "epoch": 0.509090909090909,
      "grad_norm": 0.2100268453359604,
      "learning_rate": 0.0007267832624482191,
      "loss": 2.7524,
      "step": 39200
    },
    {
      "epoch": 0.5103896103896104,
      "grad_norm": 0.21765153110027313,
      "learning_rate": 0.0007254973935502207,
      "loss": 2.7636,
      "step": 39300
    },
    {
      "epoch": 0.5116883116883116,
      "grad_norm": 0.22430171072483063,
      "learning_rate": 0.0007242096497159902,
      "loss": 2.7875,
      "step": 39400
    },
    {
      "epoch": 0.512987012987013,
      "grad_norm": 0.23513422906398773,
      "learning_rate": 0.0007229200416526911,
      "loss": 2.7606,
      "step": 39500
    },
    {
      "epoch": 0.5142857142857142,
      "grad_norm": 0.24426428973674774,
      "learning_rate": 0.0007216285800829885,
      "loss": 2.7867,
      "step": 39600
    },
    {
      "epoch": 0.5155844155844156,
      "grad_norm": 0.313528835773468,
      "learning_rate": 0.0007203352757449577,
      "loss": 2.7645,
      "step": 39700
    },
    {
      "epoch": 0.5168831168831168,
      "grad_norm": 0.23338325321674347,
      "learning_rate": 0.0007190401393919968,
      "loss": 2.7358,
      "step": 39800
    },
    {
      "epoch": 0.5181818181818182,
      "grad_norm": 0.2229086011648178,
      "learning_rate": 0.0007177431817927358,
      "loss": 2.7574,
      "step": 39900
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 0.22037512063980103,
      "learning_rate": 0.000716444413730948,
      "loss": 2.7507,
      "step": 40000
    },
    {
      "epoch": 0.5194805194805194,
      "eval_loss": 3.1465635299682617,
      "eval_runtime": 15.78,
      "eval_samples_per_second": 36.502,
      "eval_steps_per_second": 9.125,
      "step": 40000
    },
    {
      "epoch": 0.5207792207792208,
      "grad_norm": 0.22620290517807007,
      "learning_rate": 0.0007151438460054607,
      "loss": 2.7895,
      "step": 40100
    },
    {
      "epoch": 0.522077922077922,
      "grad_norm": 0.21930238604545593,
      "learning_rate": 0.000713841489430064,
      "loss": 2.7593,
      "step": 40200
    },
    {
      "epoch": 0.5233766233766234,
      "grad_norm": 0.232050359249115,
      "learning_rate": 0.0007125373548334219,
      "loss": 2.7341,
      "step": 40300
    },
    {
      "epoch": 0.5246753246753246,
      "grad_norm": 0.22093582153320312,
      "learning_rate": 0.0007112314530589825,
      "loss": 2.7415,
      "step": 40400
    },
    {
      "epoch": 0.525974025974026,
      "grad_norm": 0.21834588050842285,
      "learning_rate": 0.0007099237949648867,
      "loss": 2.7734,
      "step": 40500
    },
    {
      "epoch": 0.5272727272727272,
      "grad_norm": 0.2412509173154831,
      "learning_rate": 0.0007086143914238792,
      "loss": 2.7344,
      "step": 40600
    },
    {
      "epoch": 0.5285714285714286,
      "grad_norm": 0.2703639268875122,
      "learning_rate": 0.0007073032533232172,
      "loss": 2.7758,
      "step": 40700
    },
    {
      "epoch": 0.5298701298701298,
      "grad_norm": 0.26110807061195374,
      "learning_rate": 0.0007059903915645802,
      "loss": 2.7708,
      "step": 40800
    },
    {
      "epoch": 0.5311688311688312,
      "grad_norm": 0.2253560870885849,
      "learning_rate": 0.0007046758170639795,
      "loss": 2.7728,
      "step": 40900
    },
    {
      "epoch": 0.5324675324675324,
      "grad_norm": 0.2106214314699173,
      "learning_rate": 0.0007033595407516674,
      "loss": 2.7487,
      "step": 41000
    },
    {
      "epoch": 0.5324675324675324,
      "eval_loss": 3.1333227157592773,
      "eval_runtime": 15.8871,
      "eval_samples_per_second": 36.256,
      "eval_steps_per_second": 9.064,
      "step": 41000
    },
    {
      "epoch": 0.5337662337662338,
      "grad_norm": 0.21964064240455627,
      "learning_rate": 0.0007020415735720458,
      "loss": 2.7644,
      "step": 41100
    },
    {
      "epoch": 0.535064935064935,
      "grad_norm": 0.2143610268831253,
      "learning_rate": 0.0007007219264835758,
      "loss": 2.7476,
      "step": 41200
    },
    {
      "epoch": 0.5363636363636364,
      "grad_norm": 0.3317408263683319,
      "learning_rate": 0.0006994006104586865,
      "loss": 2.7193,
      "step": 41300
    },
    {
      "epoch": 0.5376623376623376,
      "grad_norm": 0.2196430265903473,
      "learning_rate": 0.0006980776364836834,
      "loss": 2.7699,
      "step": 41400
    },
    {
      "epoch": 0.538961038961039,
      "grad_norm": 0.2349650114774704,
      "learning_rate": 0.0006967530155586577,
      "loss": 2.7812,
      "step": 41500
    },
    {
      "epoch": 0.5402597402597402,
      "grad_norm": 0.20911413431167603,
      "learning_rate": 0.0006954267586973939,
      "loss": 2.7562,
      "step": 41600
    },
    {
      "epoch": 0.5415584415584416,
      "grad_norm": 0.2115071713924408,
      "learning_rate": 0.0006940988769272794,
      "loss": 2.7577,
      "step": 41700
    },
    {
      "epoch": 0.5428571428571428,
      "grad_norm": 0.24407769739627838,
      "learning_rate": 0.0006927693812892116,
      "loss": 2.7342,
      "step": 41800
    },
    {
      "epoch": 0.5441558441558442,
      "grad_norm": 0.21947546303272247,
      "learning_rate": 0.0006914382828375068,
      "loss": 2.7554,
      "step": 41900
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 0.2191598266363144,
      "learning_rate": 0.000690105592639809,
      "loss": 2.7256,
      "step": 42000
    },
    {
      "epoch": 0.5454545454545454,
      "eval_loss": 3.1299455165863037,
      "eval_runtime": 15.1552,
      "eval_samples_per_second": 38.007,
      "eval_steps_per_second": 9.502,
      "step": 42000
    },
    {
      "epoch": 0.5467532467532468,
      "grad_norm": 0.22920465469360352,
      "learning_rate": 0.0006887713217769954,
      "loss": 2.7328,
      "step": 42100
    },
    {
      "epoch": 0.548051948051948,
      "grad_norm": 0.23324090242385864,
      "learning_rate": 0.0006874354813430874,
      "loss": 2.7545,
      "step": 42200
    },
    {
      "epoch": 0.5493506493506494,
      "grad_norm": 0.21898794174194336,
      "learning_rate": 0.0006860980824451563,
      "loss": 2.7611,
      "step": 42300
    },
    {
      "epoch": 0.5506493506493506,
      "grad_norm": 0.21969111263751984,
      "learning_rate": 0.0006847591362032313,
      "loss": 2.7413,
      "step": 42400
    },
    {
      "epoch": 0.551948051948052,
      "grad_norm": 0.24275629222393036,
      "learning_rate": 0.0006834186537502076,
      "loss": 2.7421,
      "step": 42500
    },
    {
      "epoch": 0.5532467532467532,
      "grad_norm": 0.20918132364749908,
      "learning_rate": 0.0006820766462317534,
      "loss": 2.73,
      "step": 42600
    },
    {
      "epoch": 0.5545454545454546,
      "grad_norm": 0.2445412427186966,
      "learning_rate": 0.0006807331248062174,
      "loss": 2.7811,
      "step": 42700
    },
    {
      "epoch": 0.5558441558441558,
      "grad_norm": 0.24175776541233063,
      "learning_rate": 0.0006793881006445354,
      "loss": 2.7529,
      "step": 42800
    },
    {
      "epoch": 0.5571428571428572,
      "grad_norm": 0.234651118516922,
      "learning_rate": 0.0006780415849301388,
      "loss": 2.7592,
      "step": 42900
    },
    {
      "epoch": 0.5584415584415584,
      "grad_norm": 0.24966347217559814,
      "learning_rate": 0.0006766935888588599,
      "loss": 2.7314,
      "step": 43000
    },
    {
      "epoch": 0.5584415584415584,
      "eval_loss": 3.1258981227874756,
      "eval_runtime": 15.1058,
      "eval_samples_per_second": 38.131,
      "eval_steps_per_second": 9.533,
      "step": 43000
    },
    {
      "epoch": 0.5597402597402598,
      "grad_norm": 0.30740296840667725,
      "learning_rate": 0.0006753441236388405,
      "loss": 2.7467,
      "step": 43100
    },
    {
      "epoch": 0.561038961038961,
      "grad_norm": 0.2325826734304428,
      "learning_rate": 0.0006739932004904373,
      "loss": 2.7428,
      "step": 43200
    },
    {
      "epoch": 0.5623376623376624,
      "grad_norm": 0.2326454520225525,
      "learning_rate": 0.0006726408306461294,
      "loss": 2.7515,
      "step": 43300
    },
    {
      "epoch": 0.5636363636363636,
      "grad_norm": 0.24041685461997986,
      "learning_rate": 0.000671287025350425,
      "loss": 2.7339,
      "step": 43400
    },
    {
      "epoch": 0.564935064935065,
      "grad_norm": 0.21853885054588318,
      "learning_rate": 0.0006699317958597668,
      "loss": 2.7188,
      "step": 43500
    },
    {
      "epoch": 0.5662337662337662,
      "grad_norm": 0.32151710987091064,
      "learning_rate": 0.00066857515344244,
      "loss": 2.7317,
      "step": 43600
    },
    {
      "epoch": 0.5675324675324676,
      "grad_norm": 0.2511979639530182,
      "learning_rate": 0.0006672171093784773,
      "loss": 2.7714,
      "step": 43700
    },
    {
      "epoch": 0.5688311688311688,
      "grad_norm": 0.2406754046678543,
      "learning_rate": 0.0006658576749595663,
      "loss": 2.7762,
      "step": 43800
    },
    {
      "epoch": 0.5701298701298702,
      "grad_norm": 0.23970568180084229,
      "learning_rate": 0.0006644968614889538,
      "loss": 2.7163,
      "step": 43900
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.23587392270565033,
      "learning_rate": 0.0006631346802813543,
      "loss": 2.7272,
      "step": 44000
    },
    {
      "epoch": 0.5714285714285714,
      "eval_loss": 3.120995044708252,
      "eval_runtime": 15.2013,
      "eval_samples_per_second": 37.892,
      "eval_steps_per_second": 9.473,
      "step": 44000
    },
    {
      "epoch": 0.5727272727272728,
      "grad_norm": 0.2565126419067383,
      "learning_rate": 0.0006617711426628536,
      "loss": 2.7398,
      "step": 44100
    },
    {
      "epoch": 0.574025974025974,
      "grad_norm": 0.22110450267791748,
      "learning_rate": 0.0006604062599708158,
      "loss": 2.7204,
      "step": 44200
    },
    {
      "epoch": 0.5753246753246753,
      "grad_norm": 0.2212057262659073,
      "learning_rate": 0.0006590400435537894,
      "loss": 2.7467,
      "step": 44300
    },
    {
      "epoch": 0.5766233766233766,
      "grad_norm": 0.24286267161369324,
      "learning_rate": 0.0006576725047714116,
      "loss": 2.7193,
      "step": 44400
    },
    {
      "epoch": 0.577922077922078,
      "grad_norm": 0.23778927326202393,
      "learning_rate": 0.0006563036549943152,
      "loss": 2.7388,
      "step": 44500
    },
    {
      "epoch": 0.5792207792207792,
      "grad_norm": 0.22259333729743958,
      "learning_rate": 0.000654933505604033,
      "loss": 2.7296,
      "step": 44600
    },
    {
      "epoch": 0.5805194805194805,
      "grad_norm": 0.23351649940013885,
      "learning_rate": 0.0006535620679929045,
      "loss": 2.753,
      "step": 44700
    },
    {
      "epoch": 0.5818181818181818,
      "grad_norm": 0.22059041261672974,
      "learning_rate": 0.0006521893535639792,
      "loss": 2.7281,
      "step": 44800
    },
    {
      "epoch": 0.5831168831168831,
      "grad_norm": 0.23977316915988922,
      "learning_rate": 0.0006508153737309235,
      "loss": 2.7003,
      "step": 44900
    },
    {
      "epoch": 0.5844155844155844,
      "grad_norm": 0.25528889894485474,
      "learning_rate": 0.0006494401399179255,
      "loss": 2.7206,
      "step": 45000
    },
    {
      "epoch": 0.5844155844155844,
      "eval_loss": 3.1150028705596924,
      "eval_runtime": 16.5286,
      "eval_samples_per_second": 34.849,
      "eval_steps_per_second": 8.712,
      "step": 45000
    },
    {
      "epoch": 0.5857142857142857,
      "grad_norm": 0.24675515294075012,
      "learning_rate": 0.0006480636635595993,
      "loss": 2.7617,
      "step": 45100
    },
    {
      "epoch": 0.587012987012987,
      "grad_norm": 0.23121733963489532,
      "learning_rate": 0.0006466859561008905,
      "loss": 2.721,
      "step": 45200
    },
    {
      "epoch": 0.5883116883116883,
      "grad_norm": 0.24000558257102966,
      "learning_rate": 0.0006453070289969807,
      "loss": 2.7091,
      "step": 45300
    },
    {
      "epoch": 0.5896103896103896,
      "grad_norm": 0.22867269814014435,
      "learning_rate": 0.0006439268937131929,
      "loss": 2.7395,
      "step": 45400
    },
    {
      "epoch": 0.5909090909090909,
      "grad_norm": 0.23441274464130402,
      "learning_rate": 0.0006425455617248952,
      "loss": 2.7507,
      "step": 45500
    },
    {
      "epoch": 0.5922077922077922,
      "grad_norm": 0.22815178334712982,
      "learning_rate": 0.0006411630445174063,
      "loss": 2.7257,
      "step": 45600
    },
    {
      "epoch": 0.5935064935064935,
      "grad_norm": 0.22590583562850952,
      "learning_rate": 0.0006397793535858992,
      "loss": 2.7303,
      "step": 45700
    },
    {
      "epoch": 0.5948051948051948,
      "grad_norm": 0.2853323221206665,
      "learning_rate": 0.0006383945004353064,
      "loss": 2.7414,
      "step": 45800
    },
    {
      "epoch": 0.5961038961038961,
      "grad_norm": 0.2671266794204712,
      "learning_rate": 0.000637008496580224,
      "loss": 2.7558,
      "step": 45900
    },
    {
      "epoch": 0.5974025974025974,
      "grad_norm": 0.23534780740737915,
      "learning_rate": 0.0006356213535448151,
      "loss": 2.7498,
      "step": 46000
    },
    {
      "epoch": 0.5974025974025974,
      "eval_loss": 3.1098833084106445,
      "eval_runtime": 14.8476,
      "eval_samples_per_second": 38.794,
      "eval_steps_per_second": 9.699,
      "step": 46000
    },
    {
      "epoch": 0.5987012987012987,
      "grad_norm": 0.25292590260505676,
      "learning_rate": 0.0006342330828627155,
      "loss": 2.7116,
      "step": 46100
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.24926777184009552,
      "learning_rate": 0.0006328436960769364,
      "loss": 2.694,
      "step": 46200
    },
    {
      "epoch": 0.6012987012987013,
      "grad_norm": 0.22545288503170013,
      "learning_rate": 0.0006314532047397697,
      "loss": 2.7382,
      "step": 46300
    },
    {
      "epoch": 0.6025974025974026,
      "grad_norm": 0.24442555010318756,
      "learning_rate": 0.0006300616204126905,
      "loss": 2.7408,
      "step": 46400
    },
    {
      "epoch": 0.6038961038961039,
      "grad_norm": 0.2365717589855194,
      "learning_rate": 0.0006286689546662625,
      "loss": 2.7386,
      "step": 46500
    },
    {
      "epoch": 0.6051948051948052,
      "grad_norm": 0.26666414737701416,
      "learning_rate": 0.0006272752190800404,
      "loss": 2.7408,
      "step": 46600
    },
    {
      "epoch": 0.6064935064935065,
      "grad_norm": 0.2522192597389221,
      "learning_rate": 0.0006258804252424745,
      "loss": 2.7487,
      "step": 46700
    },
    {
      "epoch": 0.6077922077922078,
      "grad_norm": 0.32666802406311035,
      "learning_rate": 0.0006244845847508144,
      "loss": 2.7319,
      "step": 46800
    },
    {
      "epoch": 0.6090909090909091,
      "grad_norm": 0.2485225796699524,
      "learning_rate": 0.0006230877092110119,
      "loss": 2.7767,
      "step": 46900
    },
    {
      "epoch": 0.6103896103896104,
      "grad_norm": 0.28847283124923706,
      "learning_rate": 0.0006216898102376251,
      "loss": 2.7883,
      "step": 47000
    },
    {
      "epoch": 0.6103896103896104,
      "eval_loss": 3.108853340148926,
      "eval_runtime": 15.0323,
      "eval_samples_per_second": 38.317,
      "eval_steps_per_second": 9.579,
      "step": 47000
    },
    {
      "epoch": 0.6116883116883117,
      "grad_norm": 0.2481118142604828,
      "learning_rate": 0.0006202908994537215,
      "loss": 2.7383,
      "step": 47100
    },
    {
      "epoch": 0.612987012987013,
      "grad_norm": 0.2495453953742981,
      "learning_rate": 0.0006188909884907814,
      "loss": 2.7935,
      "step": 47200
    },
    {
      "epoch": 0.6142857142857143,
      "grad_norm": 0.257658451795578,
      "learning_rate": 0.0006174900889886015,
      "loss": 2.7703,
      "step": 47300
    },
    {
      "epoch": 0.6155844155844156,
      "grad_norm": 0.23015721142292023,
      "learning_rate": 0.0006160882125951977,
      "loss": 2.7421,
      "step": 47400
    },
    {
      "epoch": 0.6168831168831169,
      "grad_norm": 0.2232840657234192,
      "learning_rate": 0.0006146853709667086,
      "loss": 2.7372,
      "step": 47500
    },
    {
      "epoch": 0.6181818181818182,
      "grad_norm": 0.24996255338191986,
      "learning_rate": 0.0006132815757672979,
      "loss": 2.7654,
      "step": 47600
    },
    {
      "epoch": 0.6194805194805195,
      "grad_norm": 0.2613258361816406,
      "learning_rate": 0.0006118768386690587,
      "loss": 2.7143,
      "step": 47700
    },
    {
      "epoch": 0.6207792207792208,
      "grad_norm": 0.2545129656791687,
      "learning_rate": 0.0006104711713519152,
      "loss": 2.7443,
      "step": 47800
    },
    {
      "epoch": 0.6220779220779221,
      "grad_norm": 0.24030067026615143,
      "learning_rate": 0.000609064585503526,
      "loss": 2.7386,
      "step": 47900
    },
    {
      "epoch": 0.6233766233766234,
      "grad_norm": 0.23413017392158508,
      "learning_rate": 0.0006076570928191872,
      "loss": 2.7276,
      "step": 48000
    },
    {
      "epoch": 0.6233766233766234,
      "eval_loss": 3.097559690475464,
      "eval_runtime": 15.7054,
      "eval_samples_per_second": 36.675,
      "eval_steps_per_second": 9.169,
      "step": 48000
    },
    {
      "epoch": 0.6246753246753247,
      "grad_norm": 0.2436034381389618,
      "learning_rate": 0.0006062487050017348,
      "loss": 2.7583,
      "step": 48100
    },
    {
      "epoch": 0.625974025974026,
      "grad_norm": 0.25060826539993286,
      "learning_rate": 0.0006048394337614477,
      "loss": 2.7889,
      "step": 48200
    },
    {
      "epoch": 0.6272727272727273,
      "grad_norm": 0.23102222383022308,
      "learning_rate": 0.0006034292908159502,
      "loss": 2.777,
      "step": 48300
    },
    {
      "epoch": 0.6285714285714286,
      "grad_norm": 0.2651401460170746,
      "learning_rate": 0.000602018287890114,
      "loss": 2.753,
      "step": 48400
    },
    {
      "epoch": 0.6298701298701299,
      "grad_norm": 0.2365243136882782,
      "learning_rate": 0.000600606436715962,
      "loss": 2.7541,
      "step": 48500
    },
    {
      "epoch": 0.6311688311688312,
      "grad_norm": 0.2493165135383606,
      "learning_rate": 0.0005991937490325696,
      "loss": 2.7529,
      "step": 48600
    },
    {
      "epoch": 0.6324675324675325,
      "grad_norm": 0.24916312098503113,
      "learning_rate": 0.0005977802365859677,
      "loss": 2.7335,
      "step": 48700
    },
    {
      "epoch": 0.6337662337662338,
      "grad_norm": 0.38456228375434875,
      "learning_rate": 0.0005963659111290444,
      "loss": 2.7492,
      "step": 48800
    },
    {
      "epoch": 0.6350649350649351,
      "grad_norm": 0.23968328535556793,
      "learning_rate": 0.0005949507844214482,
      "loss": 2.729,
      "step": 48900
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 0.27702993154525757,
      "learning_rate": 0.0005935348682294894,
      "loss": 2.7453,
      "step": 49000
    },
    {
      "epoch": 0.6363636363636364,
      "eval_loss": 3.0973825454711914,
      "eval_runtime": 15.2932,
      "eval_samples_per_second": 37.664,
      "eval_steps_per_second": 9.416,
      "step": 49000
    },
    {
      "epoch": 0.6376623376623377,
      "grad_norm": 0.2538464367389679,
      "learning_rate": 0.000592118174326043,
      "loss": 2.7291,
      "step": 49100
    },
    {
      "epoch": 0.638961038961039,
      "grad_norm": 0.27856433391571045,
      "learning_rate": 0.0005907007144904501,
      "loss": 2.7324,
      "step": 49200
    },
    {
      "epoch": 0.6402597402597403,
      "grad_norm": 0.26114940643310547,
      "learning_rate": 0.0005892825005084202,
      "loss": 2.7777,
      "step": 49300
    },
    {
      "epoch": 0.6415584415584416,
      "grad_norm": 0.25457167625427246,
      "learning_rate": 0.0005878635441719333,
      "loss": 2.7273,
      "step": 49400
    },
    {
      "epoch": 0.6428571428571429,
      "grad_norm": 0.24550481140613556,
      "learning_rate": 0.0005864438572791422,
      "loss": 2.7641,
      "step": 49500
    },
    {
      "epoch": 0.6441558441558441,
      "grad_norm": 0.2586402893066406,
      "learning_rate": 0.0005850234516342739,
      "loss": 2.7115,
      "step": 49600
    },
    {
      "epoch": 0.6454545454545455,
      "grad_norm": 0.25716432929039,
      "learning_rate": 0.000583602339047531,
      "loss": 2.7164,
      "step": 49700
    },
    {
      "epoch": 0.6467532467532467,
      "grad_norm": 0.26527953147888184,
      "learning_rate": 0.0005821805313349947,
      "loss": 2.752,
      "step": 49800
    },
    {
      "epoch": 0.6480519480519481,
      "grad_norm": 0.24985221028327942,
      "learning_rate": 0.000580758040318526,
      "loss": 2.7468,
      "step": 49900
    },
    {
      "epoch": 0.6493506493506493,
      "grad_norm": 0.23009921610355377,
      "learning_rate": 0.0005793348778256671,
      "loss": 2.7323,
      "step": 50000
    },
    {
      "epoch": 0.6493506493506493,
      "eval_loss": 3.0897603034973145,
      "eval_runtime": 14.9478,
      "eval_samples_per_second": 38.534,
      "eval_steps_per_second": 9.634,
      "step": 50000
    },
    {
      "epoch": 0.6506493506493507,
      "grad_norm": 0.227068230509758,
      "learning_rate": 0.0005779110556895433,
      "loss": 2.7248,
      "step": 50100
    },
    {
      "epoch": 0.6519480519480519,
      "grad_norm": 0.22547580301761627,
      "learning_rate": 0.0005764865857487646,
      "loss": 2.7372,
      "step": 50200
    },
    {
      "epoch": 0.6532467532467533,
      "grad_norm": 0.2620651423931122,
      "learning_rate": 0.0005750614798473273,
      "loss": 2.7249,
      "step": 50300
    },
    {
      "epoch": 0.6545454545454545,
      "grad_norm": 0.26259052753448486,
      "learning_rate": 0.0005736357498345156,
      "loss": 2.6864,
      "step": 50400
    },
    {
      "epoch": 0.6558441558441559,
      "grad_norm": 0.24948278069496155,
      "learning_rate": 0.0005722094075648032,
      "loss": 2.7276,
      "step": 50500
    },
    {
      "epoch": 0.6571428571428571,
      "grad_norm": 0.23260965943336487,
      "learning_rate": 0.0005707824648977536,
      "loss": 2.7114,
      "step": 50600
    },
    {
      "epoch": 0.6584415584415585,
      "grad_norm": 0.253491073846817,
      "learning_rate": 0.0005693549336979236,
      "loss": 2.7156,
      "step": 50700
    },
    {
      "epoch": 0.6597402597402597,
      "grad_norm": 0.24990208446979523,
      "learning_rate": 0.0005679268258347626,
      "loss": 2.7319,
      "step": 50800
    },
    {
      "epoch": 0.6610389610389611,
      "grad_norm": 0.24773485958576202,
      "learning_rate": 0.0005664981531825152,
      "loss": 2.7356,
      "step": 50900
    },
    {
      "epoch": 0.6623376623376623,
      "grad_norm": 0.23747920989990234,
      "learning_rate": 0.0005650689276201219,
      "loss": 2.7497,
      "step": 51000
    },
    {
      "epoch": 0.6623376623376623,
      "eval_loss": 3.0903031826019287,
      "eval_runtime": 14.4116,
      "eval_samples_per_second": 39.968,
      "eval_steps_per_second": 9.992,
      "step": 51000
    },
    {
      "epoch": 0.6636363636363637,
      "grad_norm": 0.24378643929958344,
      "learning_rate": 0.0005636391610311204,
      "loss": 2.7239,
      "step": 51100
    },
    {
      "epoch": 0.6649350649350649,
      "grad_norm": 0.2741662561893463,
      "learning_rate": 0.0005622088653035469,
      "loss": 2.7161,
      "step": 51200
    },
    {
      "epoch": 0.6662337662337663,
      "grad_norm": 0.25329113006591797,
      "learning_rate": 0.0005607780523298372,
      "loss": 2.7301,
      "step": 51300
    },
    {
      "epoch": 0.6675324675324675,
      "grad_norm": 0.2747024893760681,
      "learning_rate": 0.0005593467340067282,
      "loss": 2.7312,
      "step": 51400
    },
    {
      "epoch": 0.6688311688311688,
      "grad_norm": 0.24558883905410767,
      "learning_rate": 0.0005579149222351577,
      "loss": 2.75,
      "step": 51500
    },
    {
      "epoch": 0.6701298701298701,
      "grad_norm": 0.24378612637519836,
      "learning_rate": 0.0005564826289201674,
      "loss": 2.742,
      "step": 51600
    },
    {
      "epoch": 0.6714285714285714,
      "grad_norm": 0.24434533715248108,
      "learning_rate": 0.0005550498659708022,
      "loss": 2.7085,
      "step": 51700
    },
    {
      "epoch": 0.6727272727272727,
      "grad_norm": 0.26504042744636536,
      "learning_rate": 0.000553616645300012,
      "loss": 2.7147,
      "step": 51800
    },
    {
      "epoch": 0.674025974025974,
      "grad_norm": 0.24731750786304474,
      "learning_rate": 0.0005521829788245528,
      "loss": 2.7433,
      "step": 51900
    },
    {
      "epoch": 0.6753246753246753,
      "grad_norm": 0.23254041373729706,
      "learning_rate": 0.0005507488784648869,
      "loss": 2.7326,
      "step": 52000
    },
    {
      "epoch": 0.6753246753246753,
      "eval_loss": 3.080451011657715,
      "eval_runtime": 18.0502,
      "eval_samples_per_second": 31.911,
      "eval_steps_per_second": 7.978,
      "step": 52000
    }
  ],
  "logging_steps": 100,
  "max_steps": 77000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 4000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.54800431824896e+18,
  "train_batch_size": 22,
  "trial_name": null,
  "trial_params": null
}