{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.6871286537485446,
  "eval_steps": 1000,
  "global_step": 90000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.9856985041650495e-05,
      "grad_norm": 8.064935684204102,
      "learning_rate": 0.0,
      "loss": 3.4849,
      "step": 1
    },
    {
      "epoch": 0.0014928492520825246,
      "grad_norm": 3.3433420658111572,
      "learning_rate": 0.00019998742849959144,
      "loss": 1.9038,
      "step": 50
    },
    {
      "epoch": 0.0029856985041650493,
      "grad_norm": 2.907883405685425,
      "learning_rate": 0.0001999731427036726,
      "loss": 1.608,
      "step": 100
    },
    {
      "epoch": 0.004478547756247574,
      "grad_norm": 3.2614288330078125,
      "learning_rate": 0.00019995885690775376,
      "loss": 1.5505,
      "step": 150
    },
    {
      "epoch": 0.005971397008330099,
      "grad_norm": 3.8400654792785645,
      "learning_rate": 0.00019994457111183493,
      "loss": 1.5737,
      "step": 200
    },
    {
      "epoch": 0.007464246260412624,
      "grad_norm": 3.3358442783355713,
      "learning_rate": 0.00019993028531591612,
      "loss": 1.567,
      "step": 250
    },
    {
      "epoch": 0.008957095512495149,
      "grad_norm": 2.131911277770996,
      "learning_rate": 0.00019991599951999726,
      "loss": 1.5208,
      "step": 300
    },
    {
      "epoch": 0.010449944764577673,
      "grad_norm": 3.180992364883423,
      "learning_rate": 0.00019990171372407845,
      "loss": 1.5586,
      "step": 350
    },
    {
      "epoch": 0.011942794016660197,
      "grad_norm": 3.024989128112793,
      "learning_rate": 0.0001998874279281596,
      "loss": 1.5267,
      "step": 400
    },
    {
      "epoch": 0.013435643268742723,
      "grad_norm": 3.4935102462768555,
      "learning_rate": 0.00019987314213224078,
      "loss": 1.5609,
      "step": 450
    },
    {
      "epoch": 0.014928492520825247,
      "grad_norm": 3.565504550933838,
      "learning_rate": 0.00019985885633632194,
      "loss": 1.5112,
      "step": 500
    },
    {
      "epoch": 0.01642134177290777,
      "grad_norm": 2.0692882537841797,
      "learning_rate": 0.00019984457054040308,
      "loss": 1.5087,
      "step": 550
    },
    {
      "epoch": 0.017914191024990297,
      "grad_norm": 4.110323905944824,
      "learning_rate": 0.00019983028474448427,
      "loss": 1.5073,
      "step": 600
    },
    {
      "epoch": 0.01940704027707282,
      "grad_norm": 2.8632736206054688,
      "learning_rate": 0.0001998159989485654,
      "loss": 1.472,
      "step": 650
    },
    {
      "epoch": 0.020899889529155346,
      "grad_norm": 2.629347324371338,
      "learning_rate": 0.0001998017131526466,
      "loss": 1.5236,
      "step": 700
    },
    {
      "epoch": 0.022392738781237872,
      "grad_norm": 3.696873188018799,
      "learning_rate": 0.00019978742735672774,
      "loss": 1.5216,
      "step": 750
    },
    {
      "epoch": 0.023885588033320394,
      "grad_norm": 3.253305435180664,
      "learning_rate": 0.00019977314156080893,
      "loss": 1.5915,
      "step": 800
    },
    {
      "epoch": 0.02537843728540292,
      "grad_norm": 2.9587886333465576,
      "learning_rate": 0.0001997588557648901,
      "loss": 1.4862,
      "step": 850
    },
    {
      "epoch": 0.026871286537485446,
      "grad_norm": 3.3311073780059814,
      "learning_rate": 0.00019974456996897126,
      "loss": 1.4703,
      "step": 900
    },
    {
      "epoch": 0.02836413578956797,
      "grad_norm": 2.5960264205932617,
      "learning_rate": 0.00019973028417305243,
      "loss": 1.4818,
      "step": 950
    },
    {
      "epoch": 0.029856985041650495,
      "grad_norm": 3.3142144680023193,
      "learning_rate": 0.0001997159983771336,
      "loss": 1.4746,
      "step": 1000
    },
    {
      "epoch": 0.03134983429373302,
      "grad_norm": 3.5049827098846436,
      "learning_rate": 0.00019970171258121476,
      "loss": 1.442,
      "step": 1050
    },
    {
      "epoch": 0.03284268354581554,
      "grad_norm": 3.0218605995178223,
      "learning_rate": 0.00019968742678529592,
      "loss": 1.5265,
      "step": 1100
    },
    {
      "epoch": 0.03433553279789807,
      "grad_norm": 2.936182975769043,
      "learning_rate": 0.00019967314098937709,
      "loss": 1.5174,
      "step": 1150
    },
    {
      "epoch": 0.035828382049980595,
      "grad_norm": 2.877253293991089,
      "learning_rate": 0.00019965885519345825,
      "loss": 1.4499,
      "step": 1200
    },
    {
      "epoch": 0.03732123130206312,
      "grad_norm": 6.07016658782959,
      "learning_rate": 0.00019964456939753941,
      "loss": 1.4542,
      "step": 1250
    },
    {
      "epoch": 0.03881408055414564,
      "grad_norm": 2.1618189811706543,
      "learning_rate": 0.0001996302836016206,
      "loss": 1.4343,
      "step": 1300
    },
    {
      "epoch": 0.040306929806228166,
      "grad_norm": 2.8267719745635986,
      "learning_rate": 0.00019961599780570174,
      "loss": 1.463,
      "step": 1350
    },
    {
      "epoch": 0.04179977905831069,
      "grad_norm": 2.6036462783813477,
      "learning_rate": 0.00019960171200978294,
      "loss": 1.4557,
      "step": 1400
    },
    {
      "epoch": 0.04329262831039322,
      "grad_norm": 3.0187127590179443,
      "learning_rate": 0.00019958742621386407,
      "loss": 1.4472,
      "step": 1450
    },
    {
      "epoch": 0.044785477562475744,
      "grad_norm": 3.9822633266448975,
      "learning_rate": 0.00019957314041794526,
      "loss": 1.4384,
      "step": 1500
    },
    {
      "epoch": 0.04627832681455826,
      "grad_norm": 2.919654607772827,
      "learning_rate": 0.0001995588546220264,
      "loss": 1.4969,
      "step": 1550
    },
    {
      "epoch": 0.04777117606664079,
      "grad_norm": 2.922963857650757,
      "learning_rate": 0.0001995445688261076,
      "loss": 1.4987,
      "step": 1600
    },
    {
      "epoch": 0.049264025318723315,
      "grad_norm": 2.9638512134552,
      "learning_rate": 0.00019953028303018876,
      "loss": 1.4722,
      "step": 1650
    },
    {
      "epoch": 0.05075687457080584,
      "grad_norm": 3.408391237258911,
      "learning_rate": 0.00019951599723426992,
      "loss": 1.4723,
      "step": 1700
    },
    {
      "epoch": 0.052249723822888366,
      "grad_norm": 3.023597240447998,
      "learning_rate": 0.0001995017114383511,
      "loss": 1.4478,
      "step": 1750
    },
    {
      "epoch": 0.05374257307497089,
      "grad_norm": 2.1655213832855225,
      "learning_rate": 0.00019948742564243225,
      "loss": 1.4071,
      "step": 1800
    },
    {
      "epoch": 0.05523542232705341,
      "grad_norm": 3.796663999557495,
      "learning_rate": 0.00019947313984651342,
      "loss": 1.446,
      "step": 1850
    },
    {
      "epoch": 0.05672827157913594,
      "grad_norm": 3.0415594577789307,
      "learning_rate": 0.00019945885405059458,
      "loss": 1.4324,
      "step": 1900
    },
    {
      "epoch": 0.05822112083121846,
      "grad_norm": 2.524627685546875,
      "learning_rate": 0.00019944456825467575,
      "loss": 1.3963,
      "step": 1950
    },
    {
      "epoch": 0.05971397008330099,
      "grad_norm": 3.2881991863250732,
      "learning_rate": 0.0001994302824587569,
      "loss": 1.4522,
      "step": 2000
    },
    {
      "epoch": 0.061206819335383515,
      "grad_norm": 3.392430067062378,
      "learning_rate": 0.00019941599666283808,
      "loss": 1.4329,
      "step": 2050
    },
    {
      "epoch": 0.06269966858746603,
      "grad_norm": 3.9426393508911133,
      "learning_rate": 0.00019940171086691927,
      "loss": 1.5203,
      "step": 2100
    },
    {
      "epoch": 0.06419251783954856,
      "grad_norm": 3.3737235069274902,
      "learning_rate": 0.0001993874250710004,
      "loss": 1.3674,
      "step": 2150
    },
    {
      "epoch": 0.06568536709163109,
      "grad_norm": 3.783085346221924,
      "learning_rate": 0.0001993731392750816,
      "loss": 1.4339,
      "step": 2200
    },
    {
      "epoch": 0.06717821634371361,
      "grad_norm": 3.4819202423095703,
      "learning_rate": 0.00019935885347916273,
      "loss": 1.4436,
      "step": 2250
    },
    {
      "epoch": 0.06867106559579614,
      "grad_norm": 3.141775608062744,
      "learning_rate": 0.00019934456768324393,
      "loss": 1.4683,
      "step": 2300
    },
    {
      "epoch": 0.07016391484787866,
      "grad_norm": 3.2881035804748535,
      "learning_rate": 0.00019933028188732506,
      "loss": 1.4395,
      "step": 2350
    },
    {
      "epoch": 0.07165676409996119,
      "grad_norm": 3.718122959136963,
      "learning_rate": 0.00019931599609140626,
      "loss": 1.4396,
      "step": 2400
    },
    {
      "epoch": 0.07314961335204372,
      "grad_norm": 4.3829474449157715,
      "learning_rate": 0.00019930171029548742,
      "loss": 1.4477,
      "step": 2450
    },
    {
      "epoch": 0.07464246260412624,
      "grad_norm": 3.3698525428771973,
      "learning_rate": 0.00019928742449956858,
      "loss": 1.3529,
      "step": 2500
    },
    {
      "epoch": 0.07613531185620875,
      "grad_norm": 3.7569565773010254,
      "learning_rate": 0.00019927313870364975,
      "loss": 1.4246,
      "step": 2550
    },
    {
      "epoch": 0.07762816110829128,
      "grad_norm": 3.1486406326293945,
      "learning_rate": 0.00019925885290773091,
      "loss": 1.3813,
      "step": 2600
    },
    {
      "epoch": 0.0791210103603738,
      "grad_norm": 4.0635480880737305,
      "learning_rate": 0.00019924456711181208,
      "loss": 1.4259,
      "step": 2650
    },
    {
      "epoch": 0.08061385961245633,
      "grad_norm": 3.2710611820220947,
      "learning_rate": 0.00019923028131589324,
      "loss": 1.3747,
      "step": 2700
    },
    {
      "epoch": 0.08210670886453886,
      "grad_norm": 3.4968345165252686,
      "learning_rate": 0.0001992159955199744,
      "loss": 1.4721,
      "step": 2750
    },
    {
      "epoch": 0.08359955811662138,
      "grad_norm": 4.274214267730713,
      "learning_rate": 0.00019920170972405557,
      "loss": 1.437,
      "step": 2800
    },
    {
      "epoch": 0.08509240736870391,
      "grad_norm": 2.970602512359619,
      "learning_rate": 0.00019918742392813674,
      "loss": 1.3336,
      "step": 2850
    },
    {
      "epoch": 0.08658525662078644,
      "grad_norm": 4.143342971801758,
      "learning_rate": 0.00019917313813221793,
      "loss": 1.4264,
      "step": 2900
    },
    {
      "epoch": 0.08807810587286896,
      "grad_norm": 3.7546920776367188,
      "learning_rate": 0.00019915885233629907,
      "loss": 1.441,
      "step": 2950
    },
    {
      "epoch": 0.08957095512495149,
      "grad_norm": 3.9160516262054443,
      "learning_rate": 0.00019914456654038026,
      "loss": 1.4261,
      "step": 3000
    },
    {
      "epoch": 0.09106380437703401,
      "grad_norm": 3.842073917388916,
      "learning_rate": 0.0001991302807444614,
      "loss": 1.4076,
      "step": 3050
    },
    {
      "epoch": 0.09255665362911653,
      "grad_norm": 4.392395496368408,
      "learning_rate": 0.0001991159949485426,
      "loss": 1.3789,
      "step": 3100
    },
    {
      "epoch": 0.09404950288119905,
      "grad_norm": 3.822425603866577,
      "learning_rate": 0.00019910170915262373,
      "loss": 1.3877,
      "step": 3150
    },
    {
      "epoch": 0.09554235213328158,
      "grad_norm": 3.1348562240600586,
      "learning_rate": 0.0001990874233567049,
      "loss": 1.4081,
      "step": 3200
    },
    {
      "epoch": 0.0970352013853641,
      "grad_norm": 3.453887939453125,
      "learning_rate": 0.00019907313756078608,
      "loss": 1.4143,
      "step": 3250
    },
    {
      "epoch": 0.09852805063744663,
      "grad_norm": 3.5057384967803955,
      "learning_rate": 0.00019905885176486722,
      "loss": 1.4264,
      "step": 3300
    },
    {
      "epoch": 0.10002089988952916,
      "grad_norm": 3.145796060562134,
      "learning_rate": 0.0001990445659689484,
      "loss": 1.4368,
      "step": 3350
    },
    {
      "epoch": 0.10151374914161168,
      "grad_norm": 3.4077043533325195,
      "learning_rate": 0.00019903028017302955,
      "loss": 1.388,
      "step": 3400
    },
    {
      "epoch": 0.10300659839369421,
      "grad_norm": 3.65567946434021,
      "learning_rate": 0.00019901599437711074,
      "loss": 1.42,
      "step": 3450
    },
    {
      "epoch": 0.10449944764577673,
      "grad_norm": 4.460702419281006,
      "learning_rate": 0.0001990017085811919,
      "loss": 1.3991,
      "step": 3500
    },
    {
      "epoch": 0.10599229689785926,
      "grad_norm": 4.155653476715088,
      "learning_rate": 0.00019898742278527307,
      "loss": 1.371,
      "step": 3550
    },
    {
      "epoch": 0.10748514614994178,
      "grad_norm": 3.8904318809509277,
      "learning_rate": 0.00019897313698935423,
      "loss": 1.4378,
      "step": 3600
    },
    {
      "epoch": 0.1089779954020243,
      "grad_norm": 4.0509233474731445,
      "learning_rate": 0.0001989588511934354,
      "loss": 1.3945,
      "step": 3650
    },
    {
      "epoch": 0.11047084465410682,
      "grad_norm": 3.785123109817505,
      "learning_rate": 0.00019894456539751656,
      "loss": 1.436,
      "step": 3700
    },
    {
      "epoch": 0.11196369390618935,
      "grad_norm": 3.4556167125701904,
      "learning_rate": 0.00019893027960159773,
      "loss": 1.3794,
      "step": 3750
    },
    {
      "epoch": 0.11345654315827187,
      "grad_norm": 4.0479559898376465,
      "learning_rate": 0.0001989159938056789,
      "loss": 1.4734,
      "step": 3800
    },
    {
      "epoch": 0.1149493924103544,
      "grad_norm": 3.890805721282959,
      "learning_rate": 0.00019890170800976006,
      "loss": 1.4341,
      "step": 3850
    },
    {
      "epoch": 0.11644224166243693,
      "grad_norm": 3.8178727626800537,
      "learning_rate": 0.00019888742221384122,
      "loss": 1.4754,
      "step": 3900
    },
    {
      "epoch": 0.11793509091451945,
      "grad_norm": 2.456165075302124,
      "learning_rate": 0.00019887313641792241,
      "loss": 1.3887,
      "step": 3950
    },
    {
      "epoch": 0.11942794016660198,
      "grad_norm": 3.5763051509857178,
      "learning_rate": 0.00019885885062200355,
      "loss": 1.3901,
      "step": 4000
    },
    {
      "epoch": 0.1209207894186845,
      "grad_norm": 3.885662317276001,
      "learning_rate": 0.00019884456482608474,
      "loss": 1.3856,
      "step": 4050
    },
    {
      "epoch": 0.12241363867076703,
      "grad_norm": 3.6095409393310547,
      "learning_rate": 0.00019883027903016588,
      "loss": 1.448,
      "step": 4100
    },
    {
      "epoch": 0.12390648792284956,
      "grad_norm": 3.7112534046173096,
      "learning_rate": 0.00019881599323424707,
      "loss": 1.3537,
      "step": 4150
    },
    {
      "epoch": 0.12539933717493207,
      "grad_norm": 3.3566672801971436,
      "learning_rate": 0.0001988017074383282,
      "loss": 1.4389,
      "step": 4200
    },
    {
      "epoch": 0.1268921864270146,
      "grad_norm": 4.570401191711426,
      "learning_rate": 0.0001987874216424094,
      "loss": 1.4191,
      "step": 4250
    },
    {
      "epoch": 0.12838503567909712,
      "grad_norm": 4.455029010772705,
      "learning_rate": 0.00019877313584649057,
      "loss": 1.3677,
      "step": 4300
    },
    {
      "epoch": 0.12987788493117966,
      "grad_norm": 3.0861828327178955,
      "learning_rate": 0.00019875885005057173,
      "loss": 1.3677,
      "step": 4350
    },
    {
      "epoch": 0.13137073418326217,
      "grad_norm": 4.419896602630615,
      "learning_rate": 0.0001987445642546529,
      "loss": 1.4524,
      "step": 4400
    },
    {
      "epoch": 0.1328635834353447,
      "grad_norm": 5.187576770782471,
      "learning_rate": 0.00019873027845873406,
      "loss": 1.3868,
      "step": 4450
    },
    {
      "epoch": 0.13435643268742722,
      "grad_norm": 5.111696243286133,
      "learning_rate": 0.00019871599266281523,
      "loss": 1.4458,
      "step": 4500
    },
    {
      "epoch": 0.13584928193950974,
      "grad_norm": 3.2652997970581055,
      "learning_rate": 0.0001987017068668964,
      "loss": 1.4529,
      "step": 4550
    },
    {
      "epoch": 0.13734213119159228,
      "grad_norm": 4.190273761749268,
      "learning_rate": 0.00019868742107097755,
      "loss": 1.3991,
      "step": 4600
    },
    {
      "epoch": 0.1388349804436748,
      "grad_norm": 4.85620641708374,
      "learning_rate": 0.00019867313527505872,
      "loss": 1.3916,
      "step": 4650
    },
    {
      "epoch": 0.14032782969575733,
      "grad_norm": 3.030954360961914,
      "learning_rate": 0.00019865884947913988,
      "loss": 1.3805,
      "step": 4700
    },
    {
      "epoch": 0.14182067894783984,
      "grad_norm": 3.264406681060791,
      "learning_rate": 0.00019864456368322108,
      "loss": 1.4048,
      "step": 4750
    },
    {
      "epoch": 0.14331352819992238,
      "grad_norm": 3.2138588428497314,
      "learning_rate": 0.0001986302778873022,
      "loss": 1.4092,
      "step": 4800
    },
    {
      "epoch": 0.1448063774520049,
      "grad_norm": 3.847222328186035,
      "learning_rate": 0.0001986159920913834,
      "loss": 1.3871,
      "step": 4850
    },
    {
      "epoch": 0.14629922670408743,
      "grad_norm": 4.004987716674805,
      "learning_rate": 0.00019860170629546454,
      "loss": 1.3845,
      "step": 4900
    },
    {
      "epoch": 0.14779207595616994,
      "grad_norm": 3.5088725090026855,
      "learning_rate": 0.00019858742049954573,
      "loss": 1.379,
      "step": 4950
    },
    {
      "epoch": 0.14928492520825248,
      "grad_norm": 3.275099277496338,
      "learning_rate": 0.00019857313470362687,
      "loss": 1.3628,
      "step": 5000
    },
    {
      "epoch": 0.150777774460335,
      "grad_norm": 3.7903060913085938,
      "learning_rate": 0.00019855884890770806,
      "loss": 1.3804,
      "step": 5050
    },
    {
      "epoch": 0.1522706237124175,
      "grad_norm": 4.294798374176025,
      "learning_rate": 0.00019854456311178923,
      "loss": 1.3988,
      "step": 5100
    },
    {
      "epoch": 0.15376347296450005,
      "grad_norm": 3.2719295024871826,
      "learning_rate": 0.0001985302773158704,
      "loss": 1.387,
      "step": 5150
    },
    {
      "epoch": 0.15525632221658256,
      "grad_norm": 4.143224239349365,
      "learning_rate": 0.00019851599151995156,
      "loss": 1.393,
      "step": 5200
    },
    {
      "epoch": 0.1567491714686651,
      "grad_norm": 3.404754638671875,
      "learning_rate": 0.00019850170572403272,
      "loss": 1.4205,
      "step": 5250
    },
    {
      "epoch": 0.1582420207207476,
      "grad_norm": 3.607126474380493,
      "learning_rate": 0.0001984874199281139,
      "loss": 1.4239,
      "step": 5300
    },
    {
      "epoch": 0.15973486997283015,
      "grad_norm": 4.140823841094971,
      "learning_rate": 0.00019847313413219505,
      "loss": 1.4204,
      "step": 5350
    },
    {
      "epoch": 0.16122771922491266,
      "grad_norm": 3.893251419067383,
      "learning_rate": 0.00019845884833627622,
      "loss": 1.392,
      "step": 5400
    },
    {
      "epoch": 0.1627205684769952,
      "grad_norm": 4.304211139678955,
      "learning_rate": 0.00019844456254035738,
      "loss": 1.44,
      "step": 5450
    },
    {
      "epoch": 0.16421341772907772,
      "grad_norm": 5.273501873016357,
      "learning_rate": 0.00019843027674443855,
      "loss": 1.445,
      "step": 5500
    },
    {
      "epoch": 0.16570626698116026,
      "grad_norm": 4.787700176239014,
      "learning_rate": 0.00019841599094851974,
      "loss": 1.3668,
      "step": 5550
    },
    {
      "epoch": 0.16719911623324277,
      "grad_norm": 3.7984108924865723,
      "learning_rate": 0.00019840170515260087,
      "loss": 1.3523,
      "step": 5600
    },
    {
      "epoch": 0.16869196548532528,
      "grad_norm": 3.885608673095703,
      "learning_rate": 0.00019838741935668207,
      "loss": 1.3917,
      "step": 5650
    },
    {
      "epoch": 0.17018481473740782,
      "grad_norm": 3.459803342819214,
      "learning_rate": 0.0001983731335607632,
      "loss": 1.3833,
      "step": 5700
    },
    {
      "epoch": 0.17167766398949033,
      "grad_norm": 3.7103006839752197,
      "learning_rate": 0.0001983588477648444,
      "loss": 1.4473,
      "step": 5750
    },
    {
      "epoch": 0.17317051324157287,
      "grad_norm": 6.645928382873535,
      "learning_rate": 0.00019834456196892553,
      "loss": 1.3706,
      "step": 5800
    },
    {
      "epoch": 0.17466336249365538,
      "grad_norm": 3.7201037406921387,
      "learning_rate": 0.0001983302761730067,
      "loss": 1.3733,
      "step": 5850
    },
    {
      "epoch": 0.17615621174573792,
      "grad_norm": 4.050106048583984,
      "learning_rate": 0.0001983159903770879,
      "loss": 1.4096,
      "step": 5900
    },
    {
      "epoch": 0.17764906099782043,
      "grad_norm": 4.190842628479004,
      "learning_rate": 0.00019830170458116903,
      "loss": 1.4404,
      "step": 5950
    },
    {
      "epoch": 0.17914191024990297,
      "grad_norm": 4.393162727355957,
      "learning_rate": 0.00019828741878525022,
      "loss": 1.4443,
      "step": 6000
    },
    {
      "epoch": 0.1806347595019855,
      "grad_norm": 3.597520351409912,
      "learning_rate": 0.00019827313298933136,
      "loss": 1.4063,
      "step": 6050
    },
    {
      "epoch": 0.18212760875406803,
      "grad_norm": 3.608085870742798,
      "learning_rate": 0.00019825884719341255,
      "loss": 1.3857,
      "step": 6100
    },
    {
      "epoch": 0.18362045800615054,
      "grad_norm": 3.7055492401123047,
      "learning_rate": 0.0001982445613974937,
      "loss": 1.3997,
      "step": 6150
    },
    {
      "epoch": 0.18511330725823305,
      "grad_norm": 3.875457763671875,
      "learning_rate": 0.00019823027560157488,
      "loss": 1.4296,
      "step": 6200
    },
    {
      "epoch": 0.1866061565103156,
      "grad_norm": 5.074592590332031,
      "learning_rate": 0.00019821598980565604,
      "loss": 1.3785,
      "step": 6250
    },
    {
      "epoch": 0.1880990057623981,
      "grad_norm": 6.013392448425293,
      "learning_rate": 0.0001982017040097372,
      "loss": 1.4391,
      "step": 6300
    },
    {
      "epoch": 0.18959185501448064,
      "grad_norm": 5.679958820343018,
      "learning_rate": 0.00019818741821381837,
      "loss": 1.367,
      "step": 6350
    },
    {
      "epoch": 0.19108470426656315,
      "grad_norm": 3.6182546615600586,
      "learning_rate": 0.00019817313241789954,
      "loss": 1.4508,
      "step": 6400
    },
    {
      "epoch": 0.1925775535186457,
      "grad_norm": 5.209213733673096,
      "learning_rate": 0.0001981588466219807,
      "loss": 1.3878,
      "step": 6450
    },
    {
      "epoch": 0.1940704027707282,
      "grad_norm": 3.0043230056762695,
      "learning_rate": 0.00019814456082606187,
      "loss": 1.4248,
      "step": 6500
    },
    {
      "epoch": 0.19556325202281075,
      "grad_norm": 3.157851219177246,
      "learning_rate": 0.00019813027503014303,
      "loss": 1.3725,
      "step": 6550
    },
    {
      "epoch": 0.19705610127489326,
      "grad_norm": 3.5292418003082275,
      "learning_rate": 0.0001981159892342242,
      "loss": 1.3932,
      "step": 6600
    },
    {
      "epoch": 0.1985489505269758,
      "grad_norm": 3.2819600105285645,
      "learning_rate": 0.00019810170343830536,
      "loss": 1.3495,
      "step": 6650
    },
    {
      "epoch": 0.2000417997790583,
      "grad_norm": 3.0243399143218994,
      "learning_rate": 0.00019808741764238655,
      "loss": 1.3689,
      "step": 6700
    },
    {
      "epoch": 0.20153464903114082,
      "grad_norm": 3.4495368003845215,
      "learning_rate": 0.0001980731318464677,
      "loss": 1.3725,
      "step": 6750
    },
    {
      "epoch": 0.20302749828322336,
      "grad_norm": 3.538259744644165,
      "learning_rate": 0.00019805884605054888,
      "loss": 1.3905,
      "step": 6800
    },
    {
      "epoch": 0.20452034753530587,
      "grad_norm": 4.162181377410889,
      "learning_rate": 0.00019804456025463002,
      "loss": 1.4129,
      "step": 6850
    },
    {
      "epoch": 0.20601319678738841,
      "grad_norm": 4.592432022094727,
      "learning_rate": 0.0001980302744587112,
      "loss": 1.3634,
      "step": 6900
    },
    {
      "epoch": 0.20750604603947093,
      "grad_norm": 3.45967960357666,
      "learning_rate": 0.00019801598866279237,
      "loss": 1.416,
      "step": 6950
    },
    {
      "epoch": 0.20899889529155347,
      "grad_norm": 4.221930503845215,
      "learning_rate": 0.00019800170286687354,
      "loss": 1.4051,
      "step": 7000
    },
    {
      "epoch": 0.21049174454363598,
      "grad_norm": 4.144239902496338,
      "learning_rate": 0.0001979874170709547,
      "loss": 1.4219,
      "step": 7050
    },
    {
      "epoch": 0.21198459379571852,
      "grad_norm": 4.7492570877075195,
      "learning_rate": 0.00019797313127503587,
      "loss": 1.4028,
      "step": 7100
    },
    {
      "epoch": 0.21347744304780103,
      "grad_norm": 3.5841355323791504,
      "learning_rate": 0.00019795884547911703,
      "loss": 1.4361,
      "step": 7150
    },
    {
      "epoch": 0.21497029229988357,
      "grad_norm": 4.662593364715576,
      "learning_rate": 0.0001979445596831982,
      "loss": 1.3816,
      "step": 7200
    },
    {
      "epoch": 0.21646314155196608,
      "grad_norm": 4.700701713562012,
      "learning_rate": 0.00019793027388727936,
      "loss": 1.4226,
      "step": 7250
    },
    {
      "epoch": 0.2179559908040486,
      "grad_norm": 4.025181293487549,
      "learning_rate": 0.00019791598809136053,
      "loss": 1.4291,
      "step": 7300
    },
    {
      "epoch": 0.21944884005613113,
      "grad_norm": 3.064573049545288,
      "learning_rate": 0.0001979017022954417,
      "loss": 1.4293,
      "step": 7350
    },
    {
      "epoch": 0.22094168930821365,
      "grad_norm": 6.342152118682861,
      "learning_rate": 0.00019788741649952288,
      "loss": 1.4173,
      "step": 7400
    },
    {
      "epoch": 0.22243453856029619,
      "grad_norm": 5.89996337890625,
      "learning_rate": 0.00019787313070360402,
      "loss": 1.396,
      "step": 7450
    },
    {
      "epoch": 0.2239273878123787,
      "grad_norm": 4.462945938110352,
      "learning_rate": 0.0001978588449076852,
      "loss": 1.3868,
      "step": 7500
    },
    {
      "epoch": 0.22542023706446124,
      "grad_norm": 3.6449055671691895,
      "learning_rate": 0.00019784455911176635,
      "loss": 1.396,
      "step": 7550
    },
    {
      "epoch": 0.22691308631654375,
      "grad_norm": 4.674243927001953,
      "learning_rate": 0.00019783027331584754,
      "loss": 1.395,
      "step": 7600
    },
    {
      "epoch": 0.2284059355686263,
      "grad_norm": 3.6160385608673096,
      "learning_rate": 0.00019781598751992868,
      "loss": 1.3918,
      "step": 7650
    },
    {
      "epoch": 0.2298987848207088,
      "grad_norm": 4.326193332672119,
      "learning_rate": 0.00019780170172400987,
      "loss": 1.3947,
      "step": 7700
    },
    {
      "epoch": 0.23139163407279134,
      "grad_norm": 5.4003777503967285,
      "learning_rate": 0.00019778741592809104,
      "loss": 1.344,
      "step": 7750
    },
    {
      "epoch": 0.23288448332487385,
      "grad_norm": 4.711580753326416,
      "learning_rate": 0.0001977731301321722,
      "loss": 1.3959,
      "step": 7800
    },
    {
      "epoch": 0.23437733257695637,
      "grad_norm": 3.4752814769744873,
      "learning_rate": 0.00019775884433625337,
      "loss": 1.3722,
      "step": 7850
    },
    {
      "epoch": 0.2358701818290389,
      "grad_norm": 4.028527736663818,
      "learning_rate": 0.00019774455854033453,
      "loss": 1.3683,
      "step": 7900
    },
    {
      "epoch": 0.23736303108112142,
      "grad_norm": 4.094334602355957,
      "learning_rate": 0.0001977302727444157,
      "loss": 1.3607,
      "step": 7950
    },
    {
      "epoch": 0.23885588033320396,
      "grad_norm": 5.232580661773682,
      "learning_rate": 0.00019771598694849686,
      "loss": 1.4354,
      "step": 8000
    },
    {
      "epoch": 0.24034872958528647,
      "grad_norm": 4.269852161407471,
      "learning_rate": 0.00019770170115257802,
      "loss": 1.4372,
      "step": 8050
    },
    {
      "epoch": 0.241841578837369,
      "grad_norm": 3.312541961669922,
      "learning_rate": 0.0001976874153566592,
      "loss": 1.3826,
      "step": 8100
    },
    {
      "epoch": 0.24333442808945152,
      "grad_norm": 3.8900692462921143,
      "learning_rate": 0.00019767312956074035,
      "loss": 1.4189,
      "step": 8150
    },
    {
      "epoch": 0.24482727734153406,
      "grad_norm": 3.894512414932251,
      "learning_rate": 0.00019765884376482155,
      "loss": 1.3365,
      "step": 8200
    },
    {
      "epoch": 0.24632012659361657,
      "grad_norm": 4.644411563873291,
      "learning_rate": 0.00019764455796890268,
      "loss": 1.4311,
      "step": 8250
    },
    {
      "epoch": 0.2478129758456991,
      "grad_norm": 8.174029350280762,
      "learning_rate": 0.00019763027217298387,
      "loss": 1.361,
      "step": 8300
    },
    {
      "epoch": 0.24930582509778162,
      "grad_norm": 4.615732192993164,
      "learning_rate": 0.000197615986377065,
      "loss": 1.4552,
      "step": 8350
    },
    {
      "epoch": 0.25079867434986414,
      "grad_norm": 4.421249866485596,
      "learning_rate": 0.0001976017005811462,
      "loss": 1.3463,
      "step": 8400
    },
    {
      "epoch": 0.2522915236019467,
      "grad_norm": 2.8386716842651367,
      "learning_rate": 0.00019758741478522734,
      "loss": 1.348,
      "step": 8450
    },
    {
      "epoch": 0.2537843728540292,
      "grad_norm": 4.3141703605651855,
      "learning_rate": 0.0001975731289893085,
      "loss": 1.4306,
      "step": 8500
    },
    {
      "epoch": 0.2552772221061117,
      "grad_norm": 3.947331428527832,
      "learning_rate": 0.0001975588431933897,
      "loss": 1.3823,
      "step": 8550
    },
    {
      "epoch": 0.25677007135819424,
      "grad_norm": 3.2268636226654053,
      "learning_rate": 0.00019754455739747084,
      "loss": 1.4199,
      "step": 8600
    },
    {
      "epoch": 0.2582629206102768,
      "grad_norm": 4.0353102684021,
      "learning_rate": 0.00019753027160155203,
      "loss": 1.3927,
      "step": 8650
    },
    {
      "epoch": 0.2597557698623593,
      "grad_norm": 3.490560293197632,
      "learning_rate": 0.00019751598580563316,
      "loss": 1.401,
      "step": 8700
    },
    {
      "epoch": 0.2612486191144418,
      "grad_norm": 5.577207088470459,
      "learning_rate": 0.00019750170000971436,
      "loss": 1.3586,
      "step": 8750
    },
    {
      "epoch": 0.26274146836652434,
      "grad_norm": 4.168467998504639,
      "learning_rate": 0.0001974874142137955,
      "loss": 1.3303,
      "step": 8800
    },
    {
      "epoch": 0.2642343176186069,
      "grad_norm": 3.812627077102661,
      "learning_rate": 0.00019747312841787669,
      "loss": 1.3717,
      "step": 8850
    },
    {
      "epoch": 0.2657271668706894,
      "grad_norm": 4.875237464904785,
      "learning_rate": 0.00019745884262195785,
      "loss": 1.3873,
      "step": 8900
    },
    {
      "epoch": 0.2672200161227719,
      "grad_norm": 4.048189163208008,
      "learning_rate": 0.00019744455682603902,
      "loss": 1.3775,
      "step": 8950
    },
    {
      "epoch": 0.26871286537485445,
      "grad_norm": 3.9090261459350586,
      "learning_rate": 0.00019743027103012018,
      "loss": 1.4296,
      "step": 9000
    },
    {
      "epoch": 0.270205714626937,
      "grad_norm": 2.8476953506469727,
      "learning_rate": 0.00019741598523420134,
      "loss": 1.4175,
      "step": 9050
    },
    {
      "epoch": 0.2716985638790195,
      "grad_norm": 5.782102584838867,
      "learning_rate": 0.0001974016994382825,
      "loss": 1.3835,
      "step": 9100
    },
    {
      "epoch": 0.273191413131102,
      "grad_norm": 4.640264987945557,
      "learning_rate": 0.00019738741364236367,
      "loss": 1.4524,
      "step": 9150
    },
    {
      "epoch": 0.27468426238318455,
      "grad_norm": 4.81790828704834,
      "learning_rate": 0.00019737312784644484,
      "loss": 1.3183,
      "step": 9200
    },
    {
      "epoch": 0.2761771116352671,
      "grad_norm": 2.685009717941284,
      "learning_rate": 0.000197358842050526,
      "loss": 1.3243,
      "step": 9250
    },
    {
      "epoch": 0.2776699608873496,
      "grad_norm": 5.321321487426758,
      "learning_rate": 0.00019734455625460717,
      "loss": 1.4086,
      "step": 9300
    },
    {
      "epoch": 0.2791628101394321,
      "grad_norm": 3.065791368484497,
      "learning_rate": 0.00019733027045868836,
      "loss": 1.337,
      "step": 9350
    },
    {
      "epoch": 0.28065565939151466,
      "grad_norm": 4.3569817543029785,
      "learning_rate": 0.0001973159846627695,
      "loss": 1.4082,
      "step": 9400
    },
    {
      "epoch": 0.2821485086435972,
      "grad_norm": 4.67582368850708,
      "learning_rate": 0.0001973016988668507,
      "loss": 1.3832,
      "step": 9450
    },
    {
      "epoch": 0.2836413578956797,
      "grad_norm": 4.942144870758057,
      "learning_rate": 0.00019728741307093183,
      "loss": 1.3734,
      "step": 9500
    },
    {
      "epoch": 0.2851342071477622,
      "grad_norm": 4.853246688842773,
      "learning_rate": 0.00019727312727501302,
      "loss": 1.4111,
      "step": 9550
    },
    {
      "epoch": 0.28662705639984476,
      "grad_norm": 3.071237325668335,
      "learning_rate": 0.00019725884147909418,
      "loss": 1.3746,
      "step": 9600
    },
    {
      "epoch": 0.28811990565192724,
      "grad_norm": 4.844615459442139,
      "learning_rate": 0.00019724455568317535,
      "loss": 1.3051,
      "step": 9650
    },
    {
      "epoch": 0.2896127549040098,
      "grad_norm": 5.954223155975342,
      "learning_rate": 0.0001972302698872565,
      "loss": 1.4131,
      "step": 9700
    },
    {
      "epoch": 0.2911056041560923,
      "grad_norm": 3.6717801094055176,
      "learning_rate": 0.00019721598409133768,
      "loss": 1.4166,
      "step": 9750
    },
    {
      "epoch": 0.29259845340817486,
      "grad_norm": 3.6257095336914062,
      "learning_rate": 0.00019720169829541884,
      "loss": 1.3679,
      "step": 9800
    },
    {
      "epoch": 0.29409130266025735,
      "grad_norm": 4.245635032653809,
      "learning_rate": 0.0001971874124995,
      "loss": 1.3171,
      "step": 9850
    },
    {
      "epoch": 0.2955841519123399,
      "grad_norm": 5.362602710723877,
      "learning_rate": 0.00019717312670358117,
      "loss": 1.3932,
      "step": 9900
    },
    {
      "epoch": 0.2970770011644224,
      "grad_norm": 4.6283721923828125,
      "learning_rate": 0.00019715884090766234,
      "loss": 1.3757,
      "step": 9950
    },
    {
      "epoch": 0.29856985041650497,
      "grad_norm": 4.299574851989746,
      "learning_rate": 0.0001971445551117435,
      "loss": 1.4018,
      "step": 10000
    },
    {
      "epoch": 0.30006269966858745,
      "grad_norm": 4.7913641929626465,
      "learning_rate": 0.0001971302693158247,
      "loss": 1.4228,
      "step": 10050
    },
    {
      "epoch": 0.30155554892067,
      "grad_norm": 5.312823295593262,
      "learning_rate": 0.00019711598351990583,
      "loss": 1.3866,
      "step": 10100
    },
    {
      "epoch": 0.30304839817275253,
      "grad_norm": 4.099662780761719,
      "learning_rate": 0.00019710169772398702,
      "loss": 1.454,
      "step": 10150
    },
    {
      "epoch": 0.304541247424835,
      "grad_norm": 4.254878520965576,
      "learning_rate": 0.00019708741192806816,
      "loss": 1.3526,
      "step": 10200
    },
    {
      "epoch": 0.30603409667691756,
      "grad_norm": 4.056606292724609,
      "learning_rate": 0.00019707312613214935,
      "loss": 1.4167,
      "step": 10250
    },
    {
      "epoch": 0.3075269459290001,
      "grad_norm": 3.790809154510498,
      "learning_rate": 0.0001970588403362305,
      "loss": 1.3536,
      "step": 10300
    },
    {
      "epoch": 0.30901979518108263,
      "grad_norm": 4.46298360824585,
      "learning_rate": 0.00019704455454031168,
      "loss": 1.3613,
      "step": 10350
    },
    {
      "epoch": 0.3105126444331651,
      "grad_norm": 4.52452278137207,
      "learning_rate": 0.00019703026874439284,
      "loss": 1.4591,
      "step": 10400
    },
    {
      "epoch": 0.31200549368524766,
      "grad_norm": 4.735177040100098,
      "learning_rate": 0.000197015982948474,
      "loss": 1.4617,
      "step": 10450
    },
    {
      "epoch": 0.3134983429373302,
      "grad_norm": 4.48261833190918,
      "learning_rate": 0.00019700169715255517,
      "loss": 1.4072,
      "step": 10500
    },
    {
      "epoch": 0.31499119218941274,
      "grad_norm": 3.3441503047943115,
      "learning_rate": 0.00019698741135663634,
      "loss": 1.4082,
      "step": 10550
    },
    {
      "epoch": 0.3164840414414952,
      "grad_norm": 3.9771218299865723,
      "learning_rate": 0.0001969731255607175,
      "loss": 1.4184,
      "step": 10600
    },
    {
      "epoch": 0.31797689069357776,
      "grad_norm": 6.366194725036621,
      "learning_rate": 0.00019695883976479867,
      "loss": 1.387,
      "step": 10650
    },
    {
      "epoch": 0.3194697399456603,
      "grad_norm": 5.072678089141846,
      "learning_rate": 0.00019694455396887983,
      "loss": 1.3996,
      "step": 10700
    },
    {
      "epoch": 0.3209625891977428,
      "grad_norm": 3.7204978466033936,
      "learning_rate": 0.000196930268172961,
      "loss": 1.3774,
      "step": 10750
    },
    {
      "epoch": 0.3224554384498253,
      "grad_norm": 4.47731351852417,
      "learning_rate": 0.00019691598237704216,
      "loss": 1.3552,
      "step": 10800
    },
    {
      "epoch": 0.32394828770190787,
      "grad_norm": 3.4569220542907715,
      "learning_rate": 0.00019690169658112335,
      "loss": 1.3794,
      "step": 10850
    },
    {
      "epoch": 0.3254411369539904,
      "grad_norm": 4.344145774841309,
      "learning_rate": 0.0001968874107852045,
      "loss": 1.4419,
      "step": 10900
    },
    {
      "epoch": 0.3269339862060729,
      "grad_norm": 4.089848041534424,
      "learning_rate": 0.00019687312498928568,
      "loss": 1.4116,
      "step": 10950
    },
    {
      "epoch": 0.32842683545815543,
      "grad_norm": 3.995945930480957,
      "learning_rate": 0.00019685883919336682,
      "loss": 1.3532,
      "step": 11000
    },
    {
      "epoch": 0.32991968471023797,
      "grad_norm": 3.8309378623962402,
      "learning_rate": 0.00019684455339744798,
      "loss": 1.3957,
      "step": 11050
    },
    {
      "epoch": 0.3314125339623205,
      "grad_norm": 4.386235237121582,
      "learning_rate": 0.00019683026760152915,
      "loss": 1.3616,
      "step": 11100
    },
    {
      "epoch": 0.332905383214403,
      "grad_norm": 5.133239269256592,
      "learning_rate": 0.00019681598180561031,
      "loss": 1.3959,
      "step": 11150
    },
    {
      "epoch": 0.33439823246648553,
      "grad_norm": 4.216183662414551,
      "learning_rate": 0.0001968016960096915,
      "loss": 1.3446,
      "step": 11200
    },
    {
      "epoch": 0.3358910817185681,
      "grad_norm": 3.631131172180176,
      "learning_rate": 0.00019678741021377264,
      "loss": 1.3779,
      "step": 11250
    },
    {
      "epoch": 0.33738393097065056,
      "grad_norm": 4.603448390960693,
      "learning_rate": 0.00019677312441785384,
      "loss": 1.3962,
      "step": 11300
    },
    {
      "epoch": 0.3388767802227331,
      "grad_norm": 3.6482913494110107,
      "learning_rate": 0.00019675883862193497,
      "loss": 1.3931,
      "step": 11350
    },
    {
      "epoch": 0.34036962947481564,
      "grad_norm": 5.040388107299805,
      "learning_rate": 0.00019674455282601616,
      "loss": 1.3465,
      "step": 11400
    },
    {
      "epoch": 0.3418624787268982,
      "grad_norm": 5.762825012207031,
      "learning_rate": 0.0001967302670300973,
      "loss": 1.3577,
      "step": 11450
    },
    {
      "epoch": 0.34335532797898066,
      "grad_norm": 4.941501617431641,
      "learning_rate": 0.0001967159812341785,
      "loss": 1.3676,
      "step": 11500
    },
    {
      "epoch": 0.3448481772310632,
      "grad_norm": 5.368370532989502,
      "learning_rate": 0.00019670169543825966,
      "loss": 1.4265,
      "step": 11550
    },
    {
      "epoch": 0.34634102648314574,
      "grad_norm": 4.931522369384766,
      "learning_rate": 0.00019668740964234082,
      "loss": 1.3551,
      "step": 11600
    },
    {
      "epoch": 0.3478338757352283,
      "grad_norm": 3.9685990810394287,
      "learning_rate": 0.000196673123846422,
      "loss": 1.402,
      "step": 11650
    },
    {
      "epoch": 0.34932672498731077,
      "grad_norm": 5.771200656890869,
      "learning_rate": 0.00019665883805050315,
      "loss": 1.3596,
      "step": 11700
    },
    {
      "epoch": 0.3508195742393933,
      "grad_norm": 5.142852306365967,
      "learning_rate": 0.00019664455225458432,
      "loss": 1.385,
      "step": 11750
    },
    {
      "epoch": 0.35231242349147585,
      "grad_norm": 3.295628786087036,
      "learning_rate": 0.00019663026645866548,
      "loss": 1.3454,
      "step": 11800
    },
    {
      "epoch": 0.35380527274355833,
      "grad_norm": 4.34658145904541,
      "learning_rate": 0.00019661598066274665,
      "loss": 1.3976,
      "step": 11850
    },
    {
      "epoch": 0.35529812199564087,
      "grad_norm": 4.032591819763184,
      "learning_rate": 0.0001966016948668278,
      "loss": 1.3571,
      "step": 11900
    },
    {
      "epoch": 0.3567909712477234,
      "grad_norm": 3.9286158084869385,
      "learning_rate": 0.00019658740907090898,
      "loss": 1.398,
      "step": 11950
    },
    {
      "epoch": 0.35828382049980595,
      "grad_norm": 5.184597492218018,
      "learning_rate": 0.00019657312327499017,
      "loss": 1.3827,
      "step": 12000
    },
    {
      "epoch": 0.35977666975188843,
      "grad_norm": 4.4749226570129395,
      "learning_rate": 0.0001965588374790713,
      "loss": 1.373,
      "step": 12050
    },
    {
      "epoch": 0.361269519003971,
      "grad_norm": 3.5633764266967773,
      "learning_rate": 0.0001965445516831525,
      "loss": 1.3704,
      "step": 12100
    },
    {
      "epoch": 0.3627623682560535,
      "grad_norm": 7.570897102355957,
      "learning_rate": 0.00019653026588723363,
      "loss": 1.3515,
      "step": 12150
    },
    {
      "epoch": 0.36425521750813605,
      "grad_norm": 4.239411354064941,
      "learning_rate": 0.00019651598009131483,
      "loss": 1.3813,
      "step": 12200
    },
    {
      "epoch": 0.36574806676021854,
      "grad_norm": 3.8941049575805664,
      "learning_rate": 0.00019650169429539596,
      "loss": 1.3954,
      "step": 12250
    },
    {
      "epoch": 0.3672409160123011,
      "grad_norm": 4.8694586753845215,
      "learning_rate": 0.00019648740849947716,
      "loss": 1.3531,
      "step": 12300
    },
    {
      "epoch": 0.3687337652643836,
      "grad_norm": 3.914964437484741,
      "learning_rate": 0.00019647312270355832,
      "loss": 1.3954,
      "step": 12350
    },
    {
      "epoch": 0.3702266145164661,
      "grad_norm": 3.4050538539886475,
      "learning_rate": 0.00019645883690763948,
      "loss": 1.3498,
      "step": 12400
    },
    {
      "epoch": 0.37171946376854864,
      "grad_norm": 4.436797618865967,
      "learning_rate": 0.00019644455111172065,
      "loss": 1.3889,
      "step": 12450
    },
    {
      "epoch": 0.3732123130206312,
      "grad_norm": 2.7660670280456543,
      "learning_rate": 0.00019643026531580181,
      "loss": 1.3392,
      "step": 12500
    },
    {
      "epoch": 0.3747051622727137,
      "grad_norm": 5.364072799682617,
      "learning_rate": 0.00019641597951988298,
      "loss": 1.3157,
      "step": 12550
    },
    {
      "epoch": 0.3761980115247962,
      "grad_norm": 5.123339653015137,
      "learning_rate": 0.00019640169372396414,
      "loss": 1.3523,
      "step": 12600
    },
    {
      "epoch": 0.37769086077687875,
      "grad_norm": 3.4495439529418945,
      "learning_rate": 0.0001963874079280453,
      "loss": 1.3331,
      "step": 12650
    },
    {
      "epoch": 0.3791837100289613,
      "grad_norm": 4.613680362701416,
      "learning_rate": 0.00019637312213212647,
      "loss": 1.3707,
      "step": 12700
    },
    {
      "epoch": 0.3806765592810438,
      "grad_norm": 3.5819404125213623,
      "learning_rate": 0.00019635883633620764,
      "loss": 1.4023,
      "step": 12750
    },
    {
      "epoch": 0.3821694085331263,
      "grad_norm": 3.5075576305389404,
      "learning_rate": 0.00019634455054028883,
      "loss": 1.3815,
      "step": 12800
    },
    {
      "epoch": 0.38366225778520885,
      "grad_norm": 4.425256729125977,
      "learning_rate": 0.00019633026474436997,
      "loss": 1.3801,
      "step": 12850
    },
    {
      "epoch": 0.3851551070372914,
      "grad_norm": 3.5711112022399902,
      "learning_rate": 0.00019631597894845116,
      "loss": 1.3375,
      "step": 12900
    },
    {
      "epoch": 0.3866479562893739,
      "grad_norm": 5.728016376495361,
      "learning_rate": 0.0001963016931525323,
      "loss": 1.3914,
      "step": 12950
    },
    {
      "epoch": 0.3881408055414564,
      "grad_norm": 4.294504642486572,
      "learning_rate": 0.0001962874073566135,
      "loss": 1.4561,
      "step": 13000
    },
    {
      "epoch": 0.38963365479353895,
      "grad_norm": 5.329941749572754,
      "learning_rate": 0.00019627312156069465,
      "loss": 1.4018,
      "step": 13050
    },
    {
      "epoch": 0.3911265040456215,
      "grad_norm": 4.166362762451172,
      "learning_rate": 0.00019625883576477582,
      "loss": 1.4334,
      "step": 13100
    },
    {
      "epoch": 0.392619353297704,
      "grad_norm": 3.3922691345214844,
      "learning_rate": 0.00019624454996885698,
      "loss": 1.3956,
      "step": 13150
    },
    {
      "epoch": 0.3941122025497865,
      "grad_norm": 4.340898036956787,
      "learning_rate": 0.00019623026417293815,
      "loss": 1.3643,
      "step": 13200
    },
    {
      "epoch": 0.39560505180186906,
      "grad_norm": 6.623823165893555,
      "learning_rate": 0.0001962159783770193,
      "loss": 1.362,
      "step": 13250
    },
    {
      "epoch": 0.3970979010539516,
      "grad_norm": 4.490639686584473,
      "learning_rate": 0.00019620169258110048,
      "loss": 1.3742,
      "step": 13300
    },
    {
      "epoch": 0.3985907503060341,
      "grad_norm": 4.179808139801025,
      "learning_rate": 0.00019618740678518164,
      "loss": 1.4191,
      "step": 13350
    },
    {
      "epoch": 0.4000835995581166,
      "grad_norm": 5.623187065124512,
      "learning_rate": 0.0001961731209892628,
      "loss": 1.3793,
      "step": 13400
    },
    {
      "epoch": 0.40157644881019916,
      "grad_norm": 3.9650678634643555,
      "learning_rate": 0.00019615883519334397,
      "loss": 1.3917,
      "step": 13450
    },
    {
      "epoch": 0.40306929806228164,
      "grad_norm": 5.047702312469482,
      "learning_rate": 0.00019614454939742516,
      "loss": 1.4501,
      "step": 13500
    },
    {
      "epoch": 0.4045621473143642,
      "grad_norm": 3.46647310256958,
      "learning_rate": 0.0001961302636015063,
      "loss": 1.4091,
      "step": 13550
    },
    {
      "epoch": 0.4060549965664467,
      "grad_norm": 3.2203481197357178,
      "learning_rate": 0.0001961159778055875,
      "loss": 1.4003,
      "step": 13600
    },
    {
      "epoch": 0.40754784581852926,
      "grad_norm": 3.727679967880249,
      "learning_rate": 0.00019610169200966863,
      "loss": 1.4004,
      "step": 13650
    },
    {
      "epoch": 0.40904069507061175,
      "grad_norm": 4.469257831573486,
      "learning_rate": 0.0001960874062137498,
      "loss": 1.4061,
      "step": 13700
    },
    {
      "epoch": 0.4105335443226943,
      "grad_norm": 4.041538715362549,
      "learning_rate": 0.00019607312041783096,
      "loss": 1.411,
      "step": 13750
    },
    {
      "epoch": 0.41202639357477683,
      "grad_norm": 5.2691779136657715,
      "learning_rate": 0.00019605883462191212,
      "loss": 1.3396,
      "step": 13800
    },
    {
      "epoch": 0.41351924282685937,
      "grad_norm": 6.236726760864258,
      "learning_rate": 0.00019604454882599331,
      "loss": 1.4295,
      "step": 13850
    },
    {
      "epoch": 0.41501209207894185,
      "grad_norm": 6.703745365142822,
      "learning_rate": 0.00019603026303007445,
      "loss": 1.4105,
      "step": 13900
    },
    {
      "epoch": 0.4165049413310244,
      "grad_norm": 3.997664451599121,
      "learning_rate": 0.00019601597723415564,
      "loss": 1.347,
      "step": 13950
    },
    {
      "epoch": 0.41799779058310693,
      "grad_norm": 5.311407566070557,
      "learning_rate": 0.00019600169143823678,
      "loss": 1.4546,
      "step": 14000
    },
    {
      "epoch": 0.4194906398351894,
      "grad_norm": 4.0283098220825195,
      "learning_rate": 0.00019598740564231797,
      "loss": 1.4513,
      "step": 14050
    },
    {
      "epoch": 0.42098348908727196,
      "grad_norm": 7.345764636993408,
      "learning_rate": 0.0001959731198463991,
      "loss": 1.3832,
      "step": 14100
    },
    {
      "epoch": 0.4224763383393545,
      "grad_norm": 4.324542045593262,
      "learning_rate": 0.0001959588340504803,
      "loss": 1.3751,
      "step": 14150
    },
    {
      "epoch": 0.42396918759143704,
      "grad_norm": 3.8322675228118896,
      "learning_rate": 0.00019594454825456147,
      "loss": 1.344,
      "step": 14200
    },
    {
      "epoch": 0.4254620368435195,
      "grad_norm": 4.62548303604126,
      "learning_rate": 0.00019593026245864263,
      "loss": 1.4346,
      "step": 14250
    },
    {
      "epoch": 0.42695488609560206,
      "grad_norm": 4.585489273071289,
      "learning_rate": 0.0001959159766627238,
      "loss": 1.4145,
      "step": 14300
    },
    {
      "epoch": 0.4284477353476846,
      "grad_norm": 3.64227557182312,
      "learning_rate": 0.00019590169086680496,
      "loss": 1.367,
      "step": 14350
    },
    {
      "epoch": 0.42994058459976714,
      "grad_norm": 4.730580806732178,
      "learning_rate": 0.00019588740507088613,
      "loss": 1.3653,
      "step": 14400
    },
    {
      "epoch": 0.4314334338518496,
      "grad_norm": 4.4075398445129395,
      "learning_rate": 0.0001958731192749673,
      "loss": 1.4081,
      "step": 14450
    },
    {
      "epoch": 0.43292628310393216,
      "grad_norm": 3.0498785972595215,
      "learning_rate": 0.00019585883347904845,
      "loss": 1.343,
      "step": 14500
    },
    {
      "epoch": 0.4344191323560147,
      "grad_norm": 4.179199695587158,
      "learning_rate": 0.00019584454768312962,
      "loss": 1.3662,
      "step": 14550
    },
    {
      "epoch": 0.4359119816080972,
      "grad_norm": 4.1148786544799805,
      "learning_rate": 0.00019583026188721078,
      "loss": 1.3785,
      "step": 14600
    },
    {
      "epoch": 0.4374048308601797,
      "grad_norm": 4.012060165405273,
      "learning_rate": 0.00019581597609129198,
      "loss": 1.4056,
      "step": 14650
    },
    {
      "epoch": 0.43889768011226227,
      "grad_norm": 7.186342716217041,
      "learning_rate": 0.0001958016902953731,
      "loss": 1.3943,
      "step": 14700
    },
    {
      "epoch": 0.4403905293643448,
      "grad_norm": 3.951267957687378,
      "learning_rate": 0.0001957874044994543,
      "loss": 1.3976,
      "step": 14750
    },
    {
      "epoch": 0.4418833786164273,
      "grad_norm": 5.276801586151123,
      "learning_rate": 0.00019577311870353544,
      "loss": 1.4137,
      "step": 14800
    },
    {
      "epoch": 0.44337622786850983,
      "grad_norm": 4.107429504394531,
      "learning_rate": 0.00019575883290761663,
      "loss": 1.388,
      "step": 14850
    },
    {
      "epoch": 0.44486907712059237,
      "grad_norm": 4.274941444396973,
      "learning_rate": 0.00019574454711169777,
      "loss": 1.3446,
      "step": 14900
    },
    {
      "epoch": 0.4463619263726749,
      "grad_norm": 4.174200534820557,
      "learning_rate": 0.00019573026131577896,
      "loss": 1.3703,
      "step": 14950
    },
    {
      "epoch": 0.4478547756247574,
      "grad_norm": 3.867125988006592,
      "learning_rate": 0.00019571597551986013,
      "loss": 1.3745,
      "step": 15000
    },
    {
      "epoch": 0.44934762487683994,
      "grad_norm": 6.454402923583984,
      "learning_rate": 0.0001957016897239413,
      "loss": 1.3081,
      "step": 15050
    },
    {
      "epoch": 0.4508404741289225,
      "grad_norm": 3.8726885318756104,
      "learning_rate": 0.00019568740392802246,
      "loss": 1.3527,
      "step": 15100
    },
    {
      "epoch": 0.45233332338100496,
      "grad_norm": 3.6218361854553223,
      "learning_rate": 0.00019567311813210362,
      "loss": 1.4549,
      "step": 15150
    },
    {
      "epoch": 0.4538261726330875,
      "grad_norm": 5.2475361824035645,
      "learning_rate": 0.0001956588323361848,
      "loss": 1.361,
      "step": 15200
    },
    {
      "epoch": 0.45531902188517004,
      "grad_norm": 4.388748645782471,
      "learning_rate": 0.00019564454654026595,
      "loss": 1.3164,
      "step": 15250
    },
    {
      "epoch": 0.4568118711372526,
      "grad_norm": 4.97973108291626,
      "learning_rate": 0.00019563026074434712,
      "loss": 1.3755,
      "step": 15300
    },
    {
      "epoch": 0.45830472038933506,
      "grad_norm": 4.538138389587402,
      "learning_rate": 0.00019561597494842828,
      "loss": 1.4339,
      "step": 15350
    },
    {
      "epoch": 0.4597975696414176,
      "grad_norm": 4.389719009399414,
      "learning_rate": 0.00019560168915250945,
      "loss": 1.4163,
      "step": 15400
    },
    {
      "epoch": 0.46129041889350014,
      "grad_norm": 4.347919464111328,
      "learning_rate": 0.00019558740335659064,
      "loss": 1.3663,
      "step": 15450
    },
    {
      "epoch": 0.4627832681455827,
      "grad_norm": 4.820595741271973,
      "learning_rate": 0.00019557311756067177,
      "loss": 1.3441,
      "step": 15500
    },
    {
      "epoch": 0.46427611739766517,
      "grad_norm": 2.6965413093566895,
      "learning_rate": 0.00019555883176475297,
      "loss": 1.3657,
      "step": 15550
    },
    {
      "epoch": 0.4657689666497477,
      "grad_norm": 4.741116523742676,
      "learning_rate": 0.0001955445459688341,
      "loss": 1.4414,
      "step": 15600
    },
    {
      "epoch": 0.46726181590183025,
      "grad_norm": 3.9512829780578613,
      "learning_rate": 0.0001955302601729153,
      "loss": 1.354,
      "step": 15650
    },
    {
      "epoch": 0.46875466515391273,
      "grad_norm": 7.704863548278809,
      "learning_rate": 0.00019551597437699646,
      "loss": 1.4107,
      "step": 15700
    },
    {
      "epoch": 0.47024751440599527,
      "grad_norm": 3.502988338470459,
      "learning_rate": 0.00019550168858107763,
      "loss": 1.415,
      "step": 15750
    },
    {
      "epoch": 0.4717403636580778,
      "grad_norm": 4.246065139770508,
      "learning_rate": 0.0001954874027851588,
      "loss": 1.3414,
      "step": 15800
    },
    {
      "epoch": 0.47323321291016035,
      "grad_norm": 3.4039735794067383,
      "learning_rate": 0.00019547311698923995,
      "loss": 1.3303,
      "step": 15850
    },
    {
      "epoch": 0.47472606216224283,
      "grad_norm": 3.279521942138672,
      "learning_rate": 0.00019545883119332112,
      "loss": 1.3447,
      "step": 15900
    },
    {
      "epoch": 0.4762189114143254,
      "grad_norm": 2.9335134029388428,
      "learning_rate": 0.00019544454539740228,
      "loss": 1.3998,
      "step": 15950
    },
    {
      "epoch": 0.4777117606664079,
      "grad_norm": 3.317011833190918,
      "learning_rate": 0.00019543025960148345,
      "loss": 1.3506,
      "step": 16000
    },
    {
      "epoch": 0.47920460991849045,
      "grad_norm": 3.5980935096740723,
      "learning_rate": 0.0001954159738055646,
      "loss": 1.3622,
      "step": 16050
    },
    {
      "epoch": 0.48069745917057294,
      "grad_norm": 4.726743698120117,
      "learning_rate": 0.00019540168800964578,
      "loss": 1.3414,
      "step": 16100
    },
    {
      "epoch": 0.4821903084226555,
      "grad_norm": 5.129758358001709,
      "learning_rate": 0.00019538740221372694,
      "loss": 1.3873,
      "step": 16150
    },
    {
      "epoch": 0.483683157674738,
      "grad_norm": 5.122271537780762,
      "learning_rate": 0.0001953731164178081,
      "loss": 1.428,
      "step": 16200
    },
    {
      "epoch": 0.4851760069268205,
      "grad_norm": 3.359868049621582,
      "learning_rate": 0.0001953588306218893,
      "loss": 1.3111,
      "step": 16250
    },
    {
      "epoch": 0.48666885617890304,
      "grad_norm": 5.066514492034912,
      "learning_rate": 0.00019534454482597044,
      "loss": 1.2961,
      "step": 16300
    },
    {
      "epoch": 0.4881617054309856,
      "grad_norm": 4.902595520019531,
      "learning_rate": 0.0001953302590300516,
      "loss": 1.3682,
      "step": 16350
    },
    {
      "epoch": 0.4896545546830681,
      "grad_norm": 5.0537028312683105,
      "learning_rate": 0.00019531597323413277,
      "loss": 1.315,
      "step": 16400
    },
    {
      "epoch": 0.4911474039351506,
      "grad_norm": 3.7002792358398438,
      "learning_rate": 0.00019530168743821393,
      "loss": 1.3441,
      "step": 16450
    },
    {
      "epoch": 0.49264025318723315,
      "grad_norm": 4.845950603485107,
      "learning_rate": 0.00019528740164229512,
      "loss": 1.3887,
      "step": 16500
    },
    {
      "epoch": 0.4941331024393157,
      "grad_norm": 4.933434963226318,
      "learning_rate": 0.00019527311584637626,
      "loss": 1.3865,
      "step": 16550
    },
    {
      "epoch": 0.4956259516913982,
      "grad_norm": 3.8103625774383545,
      "learning_rate": 0.00019525883005045745,
      "loss": 1.3757,
      "step": 16600
    },
    {
      "epoch": 0.4971188009434807,
      "grad_norm": 4.501999855041504,
      "learning_rate": 0.0001952445442545386,
      "loss": 1.4482,
      "step": 16650
    },
    {
      "epoch": 0.49861165019556325,
      "grad_norm": 5.600002765655518,
      "learning_rate": 0.00019523025845861978,
      "loss": 1.4209,
      "step": 16700
    },
    {
      "epoch": 0.5001044994476458,
      "grad_norm": 5.138682842254639,
      "learning_rate": 0.00019521597266270092,
      "loss": 1.4287,
      "step": 16750
    },
    {
      "epoch": 0.5015973486997283,
      "grad_norm": 5.575449466705322,
      "learning_rate": 0.0001952016868667821,
      "loss": 1.3992,
      "step": 16800
    },
    {
      "epoch": 0.5030901979518109,
      "grad_norm": 3.6443893909454346,
      "learning_rate": 0.00019518740107086327,
      "loss": 1.3661,
      "step": 16850
    },
    {
      "epoch": 0.5045830472038934,
      "grad_norm": 3.904905319213867,
      "learning_rate": 0.00019517311527494444,
      "loss": 1.4448,
      "step": 16900
    },
    {
      "epoch": 0.5060758964559758,
      "grad_norm": 4.380904197692871,
      "learning_rate": 0.0001951588294790256,
      "loss": 1.4076,
      "step": 16950
    },
    {
      "epoch": 0.5075687457080584,
      "grad_norm": 3.5924415588378906,
      "learning_rate": 0.00019514454368310677,
      "loss": 1.3833,
      "step": 17000
    },
    {
      "epoch": 0.5090615949601409,
      "grad_norm": 3.7007193565368652,
      "learning_rate": 0.00019513025788718793,
      "loss": 1.391,
      "step": 17050
    },
    {
      "epoch": 0.5105544442122234,
      "grad_norm": 3.9879095554351807,
      "learning_rate": 0.0001951159720912691,
      "loss": 1.3419,
      "step": 17100
    },
    {
      "epoch": 0.512047293464306,
      "grad_norm": 5.663998126983643,
      "learning_rate": 0.00019510168629535026,
      "loss": 1.3481,
      "step": 17150
    },
    {
      "epoch": 0.5135401427163885,
      "grad_norm": 3.9803707599639893,
      "learning_rate": 0.00019508740049943143,
      "loss": 1.4034,
      "step": 17200
    },
    {
      "epoch": 0.5150329919684711,
      "grad_norm": 3.718477725982666,
      "learning_rate": 0.0001950731147035126,
      "loss": 1.406,
      "step": 17250
    },
    {
      "epoch": 0.5165258412205536,
      "grad_norm": 4.864751815795898,
      "learning_rate": 0.00019505882890759378,
      "loss": 1.375,
      "step": 17300
    },
    {
      "epoch": 0.518018690472636,
      "grad_norm": 3.697645664215088,
      "learning_rate": 0.00019504454311167492,
      "loss": 1.4283,
      "step": 17350
    },
    {
      "epoch": 0.5195115397247186,
      "grad_norm": 4.063074111938477,
      "learning_rate": 0.0001950302573157561,
      "loss": 1.378,
      "step": 17400
    },
    {
      "epoch": 0.5210043889768011,
      "grad_norm": 4.223004341125488,
      "learning_rate": 0.00019501597151983725,
      "loss": 1.3789,
      "step": 17450
    },
    {
      "epoch": 0.5224972382288836,
      "grad_norm": 3.329366683959961,
      "learning_rate": 0.00019500168572391844,
      "loss": 1.346,
      "step": 17500
    },
    {
      "epoch": 0.5239900874809662,
      "grad_norm": 4.774710178375244,
      "learning_rate": 0.00019498739992799958,
      "loss": 1.3895,
      "step": 17550
    },
    {
      "epoch": 0.5254829367330487,
      "grad_norm": 6.2145490646362305,
      "learning_rate": 0.00019497311413208077,
      "loss": 1.3715,
      "step": 17600
    },
    {
      "epoch": 0.5269757859851312,
      "grad_norm": 3.9069626331329346,
      "learning_rate": 0.00019495882833616194,
      "loss": 1.4572,
      "step": 17650
    },
    {
      "epoch": 0.5284686352372138,
      "grad_norm": 3.347576141357422,
      "learning_rate": 0.0001949445425402431,
      "loss": 1.359,
      "step": 17700
    },
    {
      "epoch": 0.5299614844892963,
      "grad_norm": 5.305202484130859,
      "learning_rate": 0.00019493025674432427,
      "loss": 1.4038,
      "step": 17750
    },
    {
      "epoch": 0.5314543337413788,
      "grad_norm": 3.865619659423828,
      "learning_rate": 0.00019491597094840543,
      "loss": 1.3913,
      "step": 17800
    },
    {
      "epoch": 0.5329471829934613,
      "grad_norm": 4.791336536407471,
      "learning_rate": 0.0001949016851524866,
      "loss": 1.3852,
      "step": 17850
    },
    {
      "epoch": 0.5344400322455438,
      "grad_norm": 3.7827060222625732,
      "learning_rate": 0.00019488739935656776,
      "loss": 1.3108,
      "step": 17900
    },
    {
      "epoch": 0.5359328814976264,
      "grad_norm": 4.945117473602295,
      "learning_rate": 0.00019487311356064892,
      "loss": 1.346,
      "step": 17950
    },
    {
      "epoch": 0.5374257307497089,
      "grad_norm": 4.561169147491455,
      "learning_rate": 0.0001948588277647301,
      "loss": 1.3904,
      "step": 18000
    },
    {
      "epoch": 0.5389185800017914,
      "grad_norm": 4.608798027038574,
      "learning_rate": 0.00019484454196881125,
      "loss": 1.4133,
      "step": 18050
    },
    {
      "epoch": 0.540411429253874,
      "grad_norm": 4.303143501281738,
      "learning_rate": 0.00019483025617289245,
      "loss": 1.3994,
      "step": 18100
    },
    {
      "epoch": 0.5419042785059565,
      "grad_norm": 5.815835952758789,
      "learning_rate": 0.00019481597037697358,
      "loss": 1.392,
      "step": 18150
    },
    {
      "epoch": 0.543397127758039,
      "grad_norm": 5.349491596221924,
      "learning_rate": 0.00019480168458105477,
      "loss": 1.418,
      "step": 18200
    },
    {
      "epoch": 0.5448899770101215,
      "grad_norm": 3.7407824993133545,
      "learning_rate": 0.0001947873987851359,
      "loss": 1.3747,
      "step": 18250
    },
    {
      "epoch": 0.546382826262204,
      "grad_norm": 5.2810163497924805,
      "learning_rate": 0.0001947731129892171,
      "loss": 1.4023,
      "step": 18300
    },
    {
      "epoch": 0.5478756755142866,
      "grad_norm": 4.417948246002197,
      "learning_rate": 0.00019475882719329824,
      "loss": 1.4005,
      "step": 18350
    },
    {
      "epoch": 0.5493685247663691,
      "grad_norm": 5.287749290466309,
      "learning_rate": 0.00019474454139737943,
      "loss": 1.4211,
      "step": 18400
    },
    {
      "epoch": 0.5508613740184516,
      "grad_norm": 4.0996809005737305,
      "learning_rate": 0.0001947302556014606,
      "loss": 1.3559,
      "step": 18450
    },
    {
      "epoch": 0.5523542232705342,
      "grad_norm": 5.229327201843262,
      "learning_rate": 0.00019471596980554176,
      "loss": 1.4549,
      "step": 18500
    },
    {
      "epoch": 0.5538470725226167,
      "grad_norm": 4.409546852111816,
      "learning_rate": 0.00019470168400962293,
      "loss": 1.3683,
      "step": 18550
    },
    {
      "epoch": 0.5553399217746992,
      "grad_norm": 5.4077229499816895,
      "learning_rate": 0.0001946873982137041,
      "loss": 1.398,
      "step": 18600
    },
    {
      "epoch": 0.5568327710267817,
      "grad_norm": 5.208966255187988,
      "learning_rate": 0.00019467311241778526,
      "loss": 1.3276,
      "step": 18650
    },
    {
      "epoch": 0.5583256202788642,
      "grad_norm": 4.8162617683410645,
      "learning_rate": 0.00019465882662186642,
      "loss": 1.3314,
      "step": 18700
    },
    {
      "epoch": 0.5598184695309467,
      "grad_norm": 5.04697322845459,
      "learning_rate": 0.00019464454082594759,
      "loss": 1.3883,
      "step": 18750
    },
    {
      "epoch": 0.5613113187830293,
      "grad_norm": 4.038108825683594,
      "learning_rate": 0.00019463025503002875,
      "loss": 1.4377,
      "step": 18800
    },
    {
      "epoch": 0.5628041680351118,
      "grad_norm": 4.910576820373535,
      "learning_rate": 0.00019461596923410992,
      "loss": 1.3713,
      "step": 18850
    },
    {
      "epoch": 0.5642970172871944,
      "grad_norm": 5.3433756828308105,
      "learning_rate": 0.00019460168343819108,
      "loss": 1.4359,
      "step": 18900
    },
    {
      "epoch": 0.5657898665392769,
      "grad_norm": 3.9515552520751953,
      "learning_rate": 0.00019458739764227224,
      "loss": 1.3608,
      "step": 18950
    },
    {
      "epoch": 0.5672827157913594,
      "grad_norm": 4.515705108642578,
      "learning_rate": 0.0001945731118463534,
      "loss": 1.3274,
      "step": 19000
    },
    {
      "epoch": 0.568775565043442,
      "grad_norm": 4.434077262878418,
      "learning_rate": 0.00019455882605043457,
      "loss": 1.3681,
      "step": 19050
    },
    {
      "epoch": 0.5702684142955244,
      "grad_norm": 4.534008979797363,
      "learning_rate": 0.00019454454025451574,
      "loss": 1.3863,
      "step": 19100
    },
    {
      "epoch": 0.5717612635476069,
      "grad_norm": 4.200322151184082,
      "learning_rate": 0.00019453025445859693,
      "loss": 1.3743,
      "step": 19150
    },
    {
      "epoch": 0.5732541127996895,
      "grad_norm": 5.686845779418945,
      "learning_rate": 0.00019451596866267807,
      "loss": 1.3225,
      "step": 19200
    },
    {
      "epoch": 0.574746962051772,
      "grad_norm": 7.821211814880371,
      "learning_rate": 0.00019450168286675926,
      "loss": 1.3775,
      "step": 19250
    },
    {
      "epoch": 0.5762398113038545,
      "grad_norm": 5.200834274291992,
      "learning_rate": 0.0001944873970708404,
      "loss": 1.3423,
      "step": 19300
    },
    {
      "epoch": 0.5777326605559371,
      "grad_norm": 5.26302433013916,
      "learning_rate": 0.0001944731112749216,
      "loss": 1.3813,
      "step": 19350
    },
    {
      "epoch": 0.5792255098080196,
      "grad_norm": 3.3207719326019287,
      "learning_rate": 0.00019445882547900273,
      "loss": 1.3922,
      "step": 19400
    },
    {
      "epoch": 0.5807183590601022,
      "grad_norm": 4.619020938873291,
      "learning_rate": 0.00019444453968308392,
      "loss": 1.3533,
      "step": 19450
    },
    {
      "epoch": 0.5822112083121846,
      "grad_norm": 5.780002593994141,
      "learning_rate": 0.00019443025388716508,
      "loss": 1.4035,
      "step": 19500
    },
    {
      "epoch": 0.5837040575642671,
      "grad_norm": 4.961215496063232,
      "learning_rate": 0.00019441596809124625,
      "loss": 1.3687,
      "step": 19550
    },
    {
      "epoch": 0.5851969068163497,
      "grad_norm": 4.50115442276001,
      "learning_rate": 0.0001944016822953274,
      "loss": 1.342,
      "step": 19600
    },
    {
      "epoch": 0.5866897560684322,
      "grad_norm": 3.9477944374084473,
      "learning_rate": 0.00019438739649940858,
      "loss": 1.394,
      "step": 19650
    },
    {
      "epoch": 0.5881826053205147,
      "grad_norm": 3.7466814517974854,
      "learning_rate": 0.00019437311070348974,
      "loss": 1.3414,
      "step": 19700
    },
    {
      "epoch": 0.5896754545725973,
      "grad_norm": 4.382058143615723,
      "learning_rate": 0.0001943588249075709,
      "loss": 1.3669,
      "step": 19750
    },
    {
      "epoch": 0.5911683038246798,
      "grad_norm": 3.7016665935516357,
      "learning_rate": 0.00019434453911165207,
      "loss": 1.4548,
      "step": 19800
    },
    {
      "epoch": 0.5926611530767623,
      "grad_norm": 4.4738030433654785,
      "learning_rate": 0.00019433025331573324,
      "loss": 1.4273,
      "step": 19850
    },
    {
      "epoch": 0.5941540023288449,
      "grad_norm": 5.2445454597473145,
      "learning_rate": 0.0001943159675198144,
      "loss": 1.3746,
      "step": 19900
    },
    {
      "epoch": 0.5956468515809273,
      "grad_norm": 3.766219139099121,
      "learning_rate": 0.0001943016817238956,
      "loss": 1.4391,
      "step": 19950
    },
    {
      "epoch": 0.5971397008330099,
      "grad_norm": 6.310808181762695,
      "learning_rate": 0.00019428739592797673,
      "loss": 1.3316,
      "step": 20000
    },
    {
      "epoch": 0.5986325500850924,
      "grad_norm": 4.055521488189697,
      "learning_rate": 0.00019427311013205792,
      "loss": 1.3784,
      "step": 20050
    },
    {
      "epoch": 0.6001253993371749,
      "grad_norm": 4.933177471160889,
      "learning_rate": 0.00019425882433613906,
      "loss": 1.3352,
      "step": 20100
    },
    {
      "epoch": 0.6016182485892575,
      "grad_norm": 3.8867061138153076,
      "learning_rate": 0.00019424453854022025,
      "loss": 1.4169,
      "step": 20150
    },
    {
      "epoch": 0.60311109784134,
      "grad_norm": 3.364475727081299,
      "learning_rate": 0.0001942302527443014,
      "loss": 1.3767,
      "step": 20200
    },
    {
      "epoch": 0.6046039470934225,
      "grad_norm": 3.48152232170105,
      "learning_rate": 0.00019421596694838258,
      "loss": 1.3659,
      "step": 20250
    },
    {
      "epoch": 0.6060967963455051,
      "grad_norm": 3.3658649921417236,
      "learning_rate": 0.00019420168115246374,
      "loss": 1.3745,
      "step": 20300
    },
    {
      "epoch": 0.6075896455975875,
      "grad_norm": 4.441917896270752,
      "learning_rate": 0.0001941873953565449,
      "loss": 1.3827,
      "step": 20350
    },
    {
      "epoch": 0.60908249484967,
      "grad_norm": 5.014800548553467,
      "learning_rate": 0.00019417310956062607,
      "loss": 1.3792,
      "step": 20400
    },
    {
      "epoch": 0.6105753441017526,
      "grad_norm": 5.472316265106201,
      "learning_rate": 0.00019415882376470724,
      "loss": 1.394,
      "step": 20450
    },
    {
      "epoch": 0.6120681933538351,
      "grad_norm": 5.35073184967041,
      "learning_rate": 0.0001941445379687884,
      "loss": 1.401,
      "step": 20500
    },
    {
      "epoch": 0.6135610426059177,
      "grad_norm": 4.284445762634277,
      "learning_rate": 0.00019413025217286957,
      "loss": 1.4354,
      "step": 20550
    },
    {
      "epoch": 0.6150538918580002,
      "grad_norm": 3.561774492263794,
      "learning_rate": 0.00019411596637695073,
      "loss": 1.3295,
      "step": 20600
    },
    {
      "epoch": 0.6165467411100827,
      "grad_norm": 4.479186534881592,
      "learning_rate": 0.0001941016805810319,
      "loss": 1.3629,
      "step": 20650
    },
    {
      "epoch": 0.6180395903621653,
      "grad_norm": 4.186618804931641,
      "learning_rate": 0.00019408739478511306,
      "loss": 1.3399,
      "step": 20700
    },
    {
      "epoch": 0.6195324396142478,
      "grad_norm": 3.589655637741089,
      "learning_rate": 0.00019407310898919425,
      "loss": 1.3516,
      "step": 20750
    },
    {
      "epoch": 0.6210252888663302,
      "grad_norm": 4.330646514892578,
      "learning_rate": 0.0001940588231932754,
      "loss": 1.3148,
      "step": 20800
    },
    {
      "epoch": 0.6225181381184128,
      "grad_norm": 6.324933052062988,
      "learning_rate": 0.00019404453739735658,
      "loss": 1.331,
      "step": 20850
    },
    {
      "epoch": 0.6240109873704953,
      "grad_norm": 4.652800559997559,
      "learning_rate": 0.00019403025160143772,
      "loss": 1.3604,
      "step": 20900
    },
    {
      "epoch": 0.6255038366225778,
      "grad_norm": 4.831106662750244,
      "learning_rate": 0.0001940159658055189,
      "loss": 1.3322,
      "step": 20950
    },
    {
      "epoch": 0.6269966858746604,
      "grad_norm": 6.0938920974731445,
      "learning_rate": 0.00019400168000960005,
      "loss": 1.4106,
      "step": 21000
    },
    {
      "epoch": 0.6284895351267429,
      "grad_norm": 4.424108028411865,
      "learning_rate": 0.00019398739421368124,
      "loss": 1.4001,
      "step": 21050
    },
    {
      "epoch": 0.6299823843788255,
      "grad_norm": 4.329803466796875,
      "learning_rate": 0.0001939731084177624,
      "loss": 1.4048,
      "step": 21100
    },
    {
      "epoch": 0.631475233630908,
      "grad_norm": 3.915818929672241,
      "learning_rate": 0.00019395882262184357,
      "loss": 1.3619,
      "step": 21150
    },
    {
      "epoch": 0.6329680828829904,
      "grad_norm": 3.9562571048736572,
      "learning_rate": 0.00019394453682592474,
      "loss": 1.3636,
      "step": 21200
    },
    {
      "epoch": 0.634460932135073,
      "grad_norm": 5.229249954223633,
      "learning_rate": 0.0001939302510300059,
      "loss": 1.3502,
      "step": 21250
    },
    {
      "epoch": 0.6359537813871555,
      "grad_norm": 4.149145126342773,
      "learning_rate": 0.00019391596523408706,
      "loss": 1.3603,
      "step": 21300
    },
    {
      "epoch": 0.637446630639238,
      "grad_norm": 4.068868637084961,
      "learning_rate": 0.00019390167943816823,
      "loss": 1.3945,
      "step": 21350
    },
    {
      "epoch": 0.6389394798913206,
      "grad_norm": 3.6808931827545166,
      "learning_rate": 0.0001938873936422494,
      "loss": 1.4015,
      "step": 21400
    },
    {
      "epoch": 0.6404323291434031,
      "grad_norm": 4.391795635223389,
      "learning_rate": 0.00019387310784633056,
      "loss": 1.3428,
      "step": 21450
    },
    {
      "epoch": 0.6419251783954856,
      "grad_norm": 5.109655380249023,
      "learning_rate": 0.00019385882205041172,
      "loss": 1.3874,
      "step": 21500
    },
    {
      "epoch": 0.6434180276475682,
      "grad_norm": 4.560513496398926,
      "learning_rate": 0.0001938445362544929,
      "loss": 1.3833,
      "step": 21550
    },
    {
      "epoch": 0.6449108768996507,
      "grad_norm": 3.6108620166778564,
      "learning_rate": 0.00019383025045857405,
      "loss": 1.3981,
      "step": 21600
    },
    {
      "epoch": 0.6464037261517332,
      "grad_norm": 4.100367546081543,
      "learning_rate": 0.00019381596466265522,
      "loss": 1.3139,
      "step": 21650
    },
    {
      "epoch": 0.6478965754038157,
      "grad_norm": 3.1513540744781494,
      "learning_rate": 0.00019380167886673638,
      "loss": 1.3359,
      "step": 21700
    },
    {
      "epoch": 0.6493894246558982,
      "grad_norm": 4.793807506561279,
      "learning_rate": 0.00019378739307081755,
      "loss": 1.4025,
      "step": 21750
    },
    {
      "epoch": 0.6508822739079808,
      "grad_norm": 4.403114318847656,
      "learning_rate": 0.0001937731072748987,
      "loss": 1.349,
      "step": 21800
    },
    {
      "epoch": 0.6523751231600633,
      "grad_norm": 4.445423603057861,
      "learning_rate": 0.00019375882147897988,
      "loss": 1.4641,
      "step": 21850
    },
    {
      "epoch": 0.6538679724121458,
      "grad_norm": 5.721547603607178,
      "learning_rate": 0.00019374453568306107,
      "loss": 1.3451,
      "step": 21900
    },
    {
      "epoch": 0.6553608216642284,
      "grad_norm": 4.411465644836426,
      "learning_rate": 0.0001937302498871422,
      "loss": 1.3893,
      "step": 21950
    },
    {
      "epoch": 0.6568536709163109,
      "grad_norm": 3.181412696838379,
      "learning_rate": 0.0001937159640912234,
      "loss": 1.385,
      "step": 22000
    },
    {
      "epoch": 0.6583465201683933,
      "grad_norm": 7.374556541442871,
      "learning_rate": 0.00019370167829530453,
      "loss": 1.4051,
      "step": 22050
    },
    {
      "epoch": 0.6598393694204759,
      "grad_norm": 4.452084064483643,
      "learning_rate": 0.00019368739249938573,
      "loss": 1.3389,
      "step": 22100
    },
    {
      "epoch": 0.6613322186725584,
      "grad_norm": 5.60634708404541,
      "learning_rate": 0.0001936731067034669,
      "loss": 1.3924,
      "step": 22150
    },
    {
      "epoch": 0.662825067924641,
      "grad_norm": 4.35741662979126,
      "learning_rate": 0.00019365882090754806,
      "loss": 1.3957,
      "step": 22200
    },
    {
      "epoch": 0.6643179171767235,
      "grad_norm": 4.173916816711426,
      "learning_rate": 0.00019364453511162922,
      "loss": 1.3669,
      "step": 22250
    },
    {
      "epoch": 0.665810766428806,
      "grad_norm": 3.8707377910614014,
      "learning_rate": 0.00019363024931571038,
      "loss": 1.3478,
      "step": 22300
    },
    {
      "epoch": 0.6673036156808886,
      "grad_norm": 4.765937328338623,
      "learning_rate": 0.00019361596351979155,
      "loss": 1.4007,
      "step": 22350
    },
    {
      "epoch": 0.6687964649329711,
      "grad_norm": 4.315809726715088,
      "learning_rate": 0.00019360167772387271,
      "loss": 1.3622,
      "step": 22400
    },
    {
      "epoch": 0.6702893141850536,
      "grad_norm": 5.977219104766846,
      "learning_rate": 0.00019358739192795388,
      "loss": 1.3836,
      "step": 22450
    },
    {
      "epoch": 0.6717821634371361,
      "grad_norm": 4.420149803161621,
      "learning_rate": 0.00019357310613203504,
      "loss": 1.3488,
      "step": 22500
    },
    {
      "epoch": 0.6732750126892186,
      "grad_norm": 3.9537293910980225,
      "learning_rate": 0.0001935588203361162,
      "loss": 1.3892,
      "step": 22550
    },
    {
      "epoch": 0.6747678619413011,
      "grad_norm": 5.468355655670166,
      "learning_rate": 0.0001935445345401974,
      "loss": 1.4475,
      "step": 22600
    },
    {
      "epoch": 0.6762607111933837,
      "grad_norm": 4.3148674964904785,
      "learning_rate": 0.00019353024874427854,
      "loss": 1.3866,
      "step": 22650
    },
    {
      "epoch": 0.6777535604454662,
      "grad_norm": 4.004809379577637,
      "learning_rate": 0.00019351596294835973,
      "loss": 1.3811,
      "step": 22700
    },
    {
      "epoch": 0.6792464096975488,
      "grad_norm": 4.404988765716553,
      "learning_rate": 0.00019350167715244087,
      "loss": 1.3648,
      "step": 22750
    },
    {
      "epoch": 0.6807392589496313,
      "grad_norm": 5.115052223205566,
      "learning_rate": 0.00019348739135652206,
      "loss": 1.3523,
      "step": 22800
    },
    {
      "epoch": 0.6822321082017138,
      "grad_norm": 5.083719730377197,
      "learning_rate": 0.0001934731055606032,
      "loss": 1.3978,
      "step": 22850
    },
    {
      "epoch": 0.6837249574537964,
      "grad_norm": 4.038282871246338,
      "learning_rate": 0.0001934588197646844,
      "loss": 1.2788,
      "step": 22900
    },
    {
      "epoch": 0.6852178067058788,
      "grad_norm": 5.438407897949219,
      "learning_rate": 0.00019344453396876555,
      "loss": 1.4181,
      "step": 22950
    },
    {
      "epoch": 0.6867106559579613,
      "grad_norm": 5.381191730499268,
      "learning_rate": 0.00019343024817284672,
      "loss": 1.4145,
      "step": 23000
    },
    {
      "epoch": 0.6882035052100439,
      "grad_norm": 3.183706283569336,
      "learning_rate": 0.00019341596237692788,
      "loss": 1.3271,
      "step": 23050
    },
    {
      "epoch": 0.6896963544621264,
      "grad_norm": 4.063404083251953,
      "learning_rate": 0.00019340167658100905,
      "loss": 1.3717,
      "step": 23100
    },
    {
      "epoch": 0.6911892037142089,
      "grad_norm": 5.677481651306152,
      "learning_rate": 0.0001933873907850902,
      "loss": 1.344,
      "step": 23150
    },
    {
      "epoch": 0.6926820529662915,
      "grad_norm": 5.376470565795898,
      "learning_rate": 0.00019337310498917138,
      "loss": 1.295,
      "step": 23200
    },
    {
      "epoch": 0.694174902218374,
      "grad_norm": 3.8844990730285645,
      "learning_rate": 0.00019335881919325254,
      "loss": 1.381,
      "step": 23250
    },
    {
      "epoch": 0.6956677514704566,
      "grad_norm": 5.285017013549805,
      "learning_rate": 0.0001933445333973337,
      "loss": 1.3505,
      "step": 23300
    },
    {
      "epoch": 0.697160600722539,
      "grad_norm": 4.133642196655273,
      "learning_rate": 0.00019333024760141487,
      "loss": 1.3604,
      "step": 23350
    },
    {
      "epoch": 0.6986534499746215,
      "grad_norm": 5.037447929382324,
      "learning_rate": 0.00019331596180549606,
      "loss": 1.3209,
      "step": 23400
    },
    {
      "epoch": 0.7001462992267041,
      "grad_norm": 4.916257858276367,
      "learning_rate": 0.0001933016760095772,
      "loss": 1.4084,
      "step": 23450
    },
    {
      "epoch": 0.7016391484787866,
      "grad_norm": 3.469505786895752,
      "learning_rate": 0.0001932873902136584,
      "loss": 1.3705,
      "step": 23500
    },
    {
      "epoch": 0.7031319977308691,
      "grad_norm": 3.624896287918091,
      "learning_rate": 0.00019327310441773953,
      "loss": 1.3507,
      "step": 23550
    },
    {
      "epoch": 0.7046248469829517,
      "grad_norm": 4.352174758911133,
      "learning_rate": 0.00019325881862182072,
      "loss": 1.3184,
      "step": 23600
    },
    {
      "epoch": 0.7061176962350342,
      "grad_norm": 6.4549150466918945,
      "learning_rate": 0.00019324453282590186,
      "loss": 1.4161,
      "step": 23650
    },
    {
      "epoch": 0.7076105454871167,
      "grad_norm": 6.221735000610352,
      "learning_rate": 0.00019323024702998305,
      "loss": 1.3425,
      "step": 23700
    },
    {
      "epoch": 0.7091033947391993,
      "grad_norm": 4.285811901092529,
      "learning_rate": 0.00019321596123406421,
      "loss": 1.3418,
      "step": 23750
    },
    {
      "epoch": 0.7105962439912817,
      "grad_norm": 4.533527374267578,
      "learning_rate": 0.00019320167543814538,
      "loss": 1.3723,
      "step": 23800
    },
    {
      "epoch": 0.7120890932433643,
      "grad_norm": 3.838109016418457,
      "learning_rate": 0.00019318738964222654,
      "loss": 1.3624,
      "step": 23850
    },
    {
      "epoch": 0.7135819424954468,
      "grad_norm": 3.8401126861572266,
      "learning_rate": 0.0001931731038463077,
      "loss": 1.4239,
      "step": 23900
    },
    {
      "epoch": 0.7150747917475293,
      "grad_norm": 5.9049787521362305,
      "learning_rate": 0.00019315881805038887,
      "loss": 1.3642,
      "step": 23950
    },
    {
      "epoch": 0.7165676409996119,
      "grad_norm": 5.397033214569092,
      "learning_rate": 0.00019314453225447004,
      "loss": 1.3298,
      "step": 24000
    },
    {
      "epoch": 0.7180604902516944,
      "grad_norm": 5.649387359619141,
      "learning_rate": 0.0001931302464585512,
      "loss": 1.4102,
      "step": 24050
    },
    {
      "epoch": 0.7195533395037769,
      "grad_norm": 5.697938919067383,
      "learning_rate": 0.00019311596066263237,
      "loss": 1.3765,
      "step": 24100
    },
    {
      "epoch": 0.7210461887558595,
      "grad_norm": 4.244998455047607,
      "learning_rate": 0.00019310167486671353,
      "loss": 1.3471,
      "step": 24150
    },
    {
      "epoch": 0.722539038007942,
      "grad_norm": 4.579226493835449,
      "learning_rate": 0.0001930873890707947,
      "loss": 1.3897,
      "step": 24200
    },
    {
      "epoch": 0.7240318872600244,
      "grad_norm": 8.80657958984375,
      "learning_rate": 0.00019307310327487586,
      "loss": 1.3628,
      "step": 24250
    },
    {
      "epoch": 0.725524736512107,
      "grad_norm": 5.42709493637085,
      "learning_rate": 0.00019305881747895703,
      "loss": 1.3587,
      "step": 24300
    },
    {
      "epoch": 0.7270175857641895,
      "grad_norm": 4.2680840492248535,
      "learning_rate": 0.0001930445316830382,
      "loss": 1.3414,
      "step": 24350
    },
    {
      "epoch": 0.7285104350162721,
      "grad_norm": 5.641107559204102,
      "learning_rate": 0.00019303024588711935,
      "loss": 1.4095,
      "step": 24400
    },
    {
      "epoch": 0.7300032842683546,
      "grad_norm": 5.29530143737793,
      "learning_rate": 0.00019301596009120052,
      "loss": 1.389,
      "step": 24450
    },
    {
      "epoch": 0.7314961335204371,
      "grad_norm": 4.110136032104492,
      "learning_rate": 0.00019300167429528168,
      "loss": 1.4641,
      "step": 24500
    },
    {
      "epoch": 0.7329889827725197,
      "grad_norm": 4.390302658081055,
      "learning_rate": 0.00019298738849936288,
      "loss": 1.4028,
      "step": 24550
    },
    {
      "epoch": 0.7344818320246022,
      "grad_norm": 3.717409372329712,
      "learning_rate": 0.000192973102703444,
      "loss": 1.436,
      "step": 24600
    },
    {
      "epoch": 0.7359746812766846,
      "grad_norm": 5.399808406829834,
      "learning_rate": 0.0001929588169075252,
      "loss": 1.2887,
      "step": 24650
    },
    {
      "epoch": 0.7374675305287672,
      "grad_norm": 5.027936935424805,
      "learning_rate": 0.00019294453111160634,
      "loss": 1.3591,
      "step": 24700
    },
    {
      "epoch": 0.7389603797808497,
      "grad_norm": 5.12235689163208,
      "learning_rate": 0.00019293024531568753,
      "loss": 1.4305,
      "step": 24750
    },
    {
      "epoch": 0.7404532290329322,
      "grad_norm": 4.445746898651123,
      "learning_rate": 0.0001929159595197687,
      "loss": 1.3609,
      "step": 24800
    },
    {
      "epoch": 0.7419460782850148,
      "grad_norm": 4.590288162231445,
      "learning_rate": 0.00019290167372384986,
      "loss": 1.3811,
      "step": 24850
    },
    {
      "epoch": 0.7434389275370973,
      "grad_norm": 3.841573476791382,
      "learning_rate": 0.00019288738792793103,
      "loss": 1.3632,
      "step": 24900
    },
    {
      "epoch": 0.7449317767891799,
      "grad_norm": 5.2155327796936035,
      "learning_rate": 0.0001928731021320122,
      "loss": 1.4422,
      "step": 24950
    },
    {
      "epoch": 0.7464246260412624,
      "grad_norm": 4.942720413208008,
      "learning_rate": 0.00019285881633609336,
      "loss": 1.3916,
      "step": 25000
    },
    {
      "epoch": 0.7479174752933448,
      "grad_norm": 5.292357921600342,
      "learning_rate": 0.00019284453054017452,
      "loss": 1.3048,
      "step": 25050
    },
    {
      "epoch": 0.7494103245454274,
      "grad_norm": 6.690375804901123,
      "learning_rate": 0.0001928302447442557,
      "loss": 1.3967,
      "step": 25100
    },
    {
      "epoch": 0.7509031737975099,
      "grad_norm": 4.590394496917725,
      "learning_rate": 0.00019281595894833685,
      "loss": 1.3925,
      "step": 25150
    },
    {
      "epoch": 0.7523960230495924,
      "grad_norm": 4.220889091491699,
      "learning_rate": 0.00019280167315241802,
      "loss": 1.3263,
      "step": 25200
    },
    {
      "epoch": 0.753888872301675,
      "grad_norm": 4.847371578216553,
      "learning_rate": 0.00019278738735649918,
      "loss": 1.4141,
      "step": 25250
    },
    {
      "epoch": 0.7553817215537575,
      "grad_norm": 5.392793655395508,
      "learning_rate": 0.00019277310156058035,
      "loss": 1.3147,
      "step": 25300
    },
    {
      "epoch": 0.75687457080584,
      "grad_norm": 4.261468887329102,
      "learning_rate": 0.00019275881576466154,
      "loss": 1.3102,
      "step": 25350
    },
    {
      "epoch": 0.7583674200579226,
      "grad_norm": 4.498802661895752,
      "learning_rate": 0.00019274452996874267,
      "loss": 1.3912,
      "step": 25400
    },
    {
      "epoch": 0.759860269310005,
      "grad_norm": 5.2235283851623535,
      "learning_rate": 0.00019273024417282387,
      "loss": 1.3375,
      "step": 25450
    },
    {
      "epoch": 0.7613531185620876,
      "grad_norm": 6.409016132354736,
      "learning_rate": 0.000192715958376905,
      "loss": 1.3339,
      "step": 25500
    },
    {
      "epoch": 0.7628459678141701,
      "grad_norm": 4.4392805099487305,
      "learning_rate": 0.0001927016725809862,
      "loss": 1.3599,
      "step": 25550
    },
    {
      "epoch": 0.7643388170662526,
      "grad_norm": 5.580776691436768,
      "learning_rate": 0.00019268738678506736,
      "loss": 1.3812,
      "step": 25600
    },
    {
      "epoch": 0.7658316663183352,
      "grad_norm": 5.473046779632568,
      "learning_rate": 0.00019267310098914852,
      "loss": 1.3582,
      "step": 25650
    },
    {
      "epoch": 0.7673245155704177,
      "grad_norm": 6.353076934814453,
      "learning_rate": 0.0001926588151932297,
      "loss": 1.3388,
      "step": 25700
    },
    {
      "epoch": 0.7688173648225002,
      "grad_norm": 4.247453689575195,
      "learning_rate": 0.00019264452939731085,
      "loss": 1.3745,
      "step": 25750
    },
    {
      "epoch": 0.7703102140745828,
      "grad_norm": 5.048892498016357,
      "learning_rate": 0.00019263024360139202,
      "loss": 1.405,
      "step": 25800
    },
    {
      "epoch": 0.7718030633266653,
      "grad_norm": 4.883440017700195,
      "learning_rate": 0.00019261595780547318,
      "loss": 1.4082,
      "step": 25850
    },
    {
      "epoch": 0.7732959125787477,
      "grad_norm": 4.221151828765869,
      "learning_rate": 0.00019260167200955435,
      "loss": 1.3766,
      "step": 25900
    },
    {
      "epoch": 0.7747887618308303,
      "grad_norm": 8.68738079071045,
      "learning_rate": 0.0001925873862136355,
      "loss": 1.4046,
      "step": 25950
    },
    {
      "epoch": 0.7762816110829128,
      "grad_norm": 4.479017734527588,
      "learning_rate": 0.00019257310041771668,
      "loss": 1.3452,
      "step": 26000
    },
    {
      "epoch": 0.7777744603349954,
      "grad_norm": 4.299473285675049,
      "learning_rate": 0.00019255881462179787,
      "loss": 1.3702,
      "step": 26050
    },
    {
      "epoch": 0.7792673095870779,
      "grad_norm": 5.272589683532715,
      "learning_rate": 0.000192544528825879,
      "loss": 1.3725,
      "step": 26100
    },
    {
      "epoch": 0.7807601588391604,
      "grad_norm": 4.644061088562012,
      "learning_rate": 0.0001925302430299602,
      "loss": 1.4143,
      "step": 26150
    },
    {
      "epoch": 0.782253008091243,
      "grad_norm": 4.7033185958862305,
      "learning_rate": 0.00019251595723404134,
      "loss": 1.3666,
      "step": 26200
    },
    {
      "epoch": 0.7837458573433255,
      "grad_norm": 3.5262560844421387,
      "learning_rate": 0.00019250167143812253,
      "loss": 1.4135,
      "step": 26250
    },
    {
      "epoch": 0.785238706595408,
      "grad_norm": 3.8599159717559814,
      "learning_rate": 0.00019248738564220367,
      "loss": 1.3258,
      "step": 26300
    },
    {
      "epoch": 0.7867315558474905,
      "grad_norm": 5.743364334106445,
      "learning_rate": 0.00019247309984628486,
      "loss": 1.3946,
      "step": 26350
    },
    {
      "epoch": 0.788224405099573,
      "grad_norm": 5.478078365325928,
      "learning_rate": 0.00019245881405036602,
      "loss": 1.3881,
      "step": 26400
    },
    {
      "epoch": 0.7897172543516555,
      "grad_norm": 5.912649631500244,
      "learning_rate": 0.0001924445282544472,
      "loss": 1.4139,
      "step": 26450
    },
    {
      "epoch": 0.7912101036037381,
      "grad_norm": 3.753570079803467,
      "learning_rate": 0.00019243024245852835,
      "loss": 1.3267,
      "step": 26500
    },
    {
      "epoch": 0.7927029528558206,
      "grad_norm": 4.155190944671631,
      "learning_rate": 0.00019241595666260952,
      "loss": 1.3539,
      "step": 26550
    },
    {
      "epoch": 0.7941958021079032,
      "grad_norm": 4.314638614654541,
      "learning_rate": 0.00019240167086669068,
      "loss": 1.3635,
      "step": 26600
    },
    {
      "epoch": 0.7956886513599857,
      "grad_norm": 4.177329063415527,
      "learning_rate": 0.00019238738507077185,
      "loss": 1.3642,
      "step": 26650
    },
    {
      "epoch": 0.7971815006120682,
      "grad_norm": 4.3052144050598145,
      "learning_rate": 0.000192373099274853,
      "loss": 1.3011,
      "step": 26700
    },
    {
      "epoch": 0.7986743498641508,
      "grad_norm": 5.800657272338867,
      "learning_rate": 0.00019235881347893417,
      "loss": 1.3247,
      "step": 26750
    },
    {
      "epoch": 0.8001671991162332,
      "grad_norm": 4.211179733276367,
      "learning_rate": 0.00019234452768301534,
      "loss": 1.3399,
      "step": 26800
    },
    {
      "epoch": 0.8016600483683157,
      "grad_norm": 4.8080315589904785,
      "learning_rate": 0.0001923302418870965,
      "loss": 1.3439,
      "step": 26850
    },
    {
      "epoch": 0.8031528976203983,
      "grad_norm": 5.03045129776001,
      "learning_rate": 0.00019231595609117767,
      "loss": 1.4009,
      "step": 26900
    },
    {
      "epoch": 0.8046457468724808,
      "grad_norm": 8.932162284851074,
      "learning_rate": 0.00019230167029525883,
      "loss": 1.3298,
      "step": 26950
    },
    {
      "epoch": 0.8061385961245633,
      "grad_norm": 5.651429176330566,
      "learning_rate": 0.00019228738449934,
      "loss": 1.4044,
      "step": 27000
    },
    {
      "epoch": 0.8076314453766459,
      "grad_norm": 4.428074836730957,
      "learning_rate": 0.00019227309870342116,
      "loss": 1.3988,
      "step": 27050
    },
    {
      "epoch": 0.8091242946287284,
      "grad_norm": 6.512781620025635,
      "learning_rate": 0.00019225881290750233,
      "loss": 1.3529,
      "step": 27100
    },
    {
      "epoch": 0.810617143880811,
      "grad_norm": 5.393406867980957,
      "learning_rate": 0.0001922445271115835,
      "loss": 1.3266,
      "step": 27150
    },
    {
      "epoch": 0.8121099931328934,
      "grad_norm": 3.74702525138855,
      "learning_rate": 0.00019223024131566468,
      "loss": 1.416,
      "step": 27200
    },
    {
      "epoch": 0.8136028423849759,
      "grad_norm": 3.856290817260742,
      "learning_rate": 0.00019221595551974582,
      "loss": 1.4082,
      "step": 27250
    },
    {
      "epoch": 0.8150956916370585,
      "grad_norm": 4.597848415374756,
      "learning_rate": 0.000192201669723827,
      "loss": 1.3751,
      "step": 27300
    },
    {
      "epoch": 0.816588540889141,
      "grad_norm": 3.875593662261963,
      "learning_rate": 0.00019218738392790815,
      "loss": 1.4024,
      "step": 27350
    },
    {
      "epoch": 0.8180813901412235,
      "grad_norm": 5.470495700836182,
      "learning_rate": 0.00019217309813198934,
      "loss": 1.394,
      "step": 27400
    },
    {
      "epoch": 0.8195742393933061,
      "grad_norm": 4.562834739685059,
      "learning_rate": 0.00019215881233607048,
      "loss": 1.3161,
      "step": 27450
    },
    {
      "epoch": 0.8210670886453886,
      "grad_norm": 4.183352470397949,
      "learning_rate": 0.00019214452654015167,
      "loss": 1.3427,
      "step": 27500
    },
    {
      "epoch": 0.8225599378974711,
      "grad_norm": 5.345236301422119,
      "learning_rate": 0.00019213024074423284,
      "loss": 1.3551,
      "step": 27550
    },
    {
      "epoch": 0.8240527871495537,
      "grad_norm": 5.217405319213867,
      "learning_rate": 0.000192115954948314,
      "loss": 1.3174,
      "step": 27600
    },
    {
      "epoch": 0.8255456364016361,
      "grad_norm": 4.486965179443359,
      "learning_rate": 0.00019210166915239517,
      "loss": 1.3931,
      "step": 27650
    },
    {
      "epoch": 0.8270384856537187,
      "grad_norm": 4.183075904846191,
      "learning_rate": 0.00019208738335647633,
      "loss": 1.3672,
      "step": 27700
    },
    {
      "epoch": 0.8285313349058012,
      "grad_norm": 4.916398525238037,
      "learning_rate": 0.0001920730975605575,
      "loss": 1.3993,
      "step": 27750
    },
    {
      "epoch": 0.8300241841578837,
      "grad_norm": 3.99765682220459,
      "learning_rate": 0.00019205881176463866,
      "loss": 1.362,
      "step": 27800
    },
    {
      "epoch": 0.8315170334099663,
      "grad_norm": 3.948981523513794,
      "learning_rate": 0.00019204452596871982,
      "loss": 1.3912,
      "step": 27850
    },
    {
      "epoch": 0.8330098826620488,
      "grad_norm": 6.524683475494385,
      "learning_rate": 0.000192030240172801,
      "loss": 1.3845,
      "step": 27900
    },
    {
      "epoch": 0.8345027319141313,
      "grad_norm": 4.598319053649902,
      "learning_rate": 0.00019201595437688215,
      "loss": 1.38,
      "step": 27950
    },
    {
      "epoch": 0.8359955811662139,
      "grad_norm": 4.287365436553955,
      "learning_rate": 0.00019200166858096335,
      "loss": 1.3687,
      "step": 28000
    },
    {
      "epoch": 0.8374884304182963,
      "grad_norm": 5.497170448303223,
      "learning_rate": 0.00019198738278504448,
      "loss": 1.4324,
      "step": 28050
    },
    {
      "epoch": 0.8389812796703788,
      "grad_norm": 3.9373042583465576,
      "learning_rate": 0.00019197309698912567,
      "loss": 1.3892,
      "step": 28100
    },
    {
      "epoch": 0.8404741289224614,
      "grad_norm": 5.157922267913818,
      "learning_rate": 0.0001919588111932068,
      "loss": 1.3764,
      "step": 28150
    },
    {
      "epoch": 0.8419669781745439,
      "grad_norm": 4.0024895668029785,
      "learning_rate": 0.000191944525397288,
      "loss": 1.3571,
      "step": 28200
    },
    {
      "epoch": 0.8434598274266265,
      "grad_norm": 4.894957542419434,
      "learning_rate": 0.00019193023960136917,
      "loss": 1.3997,
      "step": 28250
    },
    {
      "epoch": 0.844952676678709,
      "grad_norm": 4.829247951507568,
      "learning_rate": 0.00019191595380545033,
      "loss": 1.3751,
      "step": 28300
    },
    {
      "epoch": 0.8464455259307915,
      "grad_norm": 4.123593330383301,
      "learning_rate": 0.0001919016680095315,
      "loss": 1.4183,
      "step": 28350
    },
    {
      "epoch": 0.8479383751828741,
      "grad_norm": 4.012912273406982,
      "learning_rate": 0.00019188738221361266,
      "loss": 1.3757,
      "step": 28400
    },
    {
      "epoch": 0.8494312244349566,
      "grad_norm": 5.17979097366333,
      "learning_rate": 0.00019187309641769383,
      "loss": 1.3975,
      "step": 28450
    },
    {
      "epoch": 0.850924073687039,
      "grad_norm": 4.665642261505127,
      "learning_rate": 0.000191858810621775,
      "loss": 1.4539,
      "step": 28500
    },
    {
      "epoch": 0.8524169229391216,
      "grad_norm": 4.333798408508301,
      "learning_rate": 0.00019184452482585616,
      "loss": 1.3711,
      "step": 28550
    },
    {
      "epoch": 0.8539097721912041,
      "grad_norm": 4.485865116119385,
      "learning_rate": 0.00019183023902993732,
      "loss": 1.4299,
      "step": 28600
    },
    {
      "epoch": 0.8554026214432866,
      "grad_norm": 6.3318328857421875,
      "learning_rate": 0.00019181595323401849,
      "loss": 1.3621,
      "step": 28650
    },
    {
      "epoch": 0.8568954706953692,
      "grad_norm": 3.915356159210205,
      "learning_rate": 0.00019180166743809965,
      "loss": 1.4561,
      "step": 28700
    },
    {
      "epoch": 0.8583883199474517,
      "grad_norm": 4.613978862762451,
      "learning_rate": 0.00019178738164218081,
      "loss": 1.3267,
      "step": 28750
    },
    {
      "epoch": 0.8598811691995343,
      "grad_norm": 4.935427188873291,
      "learning_rate": 0.000191773095846262,
      "loss": 1.3962,
      "step": 28800
    },
    {
      "epoch": 0.8613740184516168,
      "grad_norm": 4.204285144805908,
      "learning_rate": 0.00019175881005034314,
      "loss": 1.3644,
      "step": 28850
    },
    {
      "epoch": 0.8628668677036992,
      "grad_norm": 3.901266098022461,
      "learning_rate": 0.00019174452425442434,
      "loss": 1.344,
      "step": 28900
    },
    {
      "epoch": 0.8643597169557818,
      "grad_norm": 5.1740922927856445,
      "learning_rate": 0.00019173023845850547,
      "loss": 1.3852,
      "step": 28950
    },
    {
      "epoch": 0.8658525662078643,
      "grad_norm": 4.984529972076416,
      "learning_rate": 0.00019171595266258667,
      "loss": 1.3853,
      "step": 29000
    },
    {
      "epoch": 0.8673454154599468,
      "grad_norm": 4.719232082366943,
      "learning_rate": 0.00019170166686666783,
      "loss": 1.3962,
      "step": 29050
    },
    {
      "epoch": 0.8688382647120294,
      "grad_norm": 6.084499359130859,
      "learning_rate": 0.000191687381070749,
      "loss": 1.3377,
      "step": 29100
    },
    {
      "epoch": 0.8703311139641119,
      "grad_norm": 5.261173248291016,
      "learning_rate": 0.00019167309527483016,
      "loss": 1.3673,
      "step": 29150
    },
    {
      "epoch": 0.8718239632161944,
      "grad_norm": 5.18539571762085,
      "learning_rate": 0.00019165880947891132,
      "loss": 1.358,
      "step": 29200
    },
    {
      "epoch": 0.873316812468277,
      "grad_norm": 4.789880275726318,
      "learning_rate": 0.0001916445236829925,
      "loss": 1.3426,
      "step": 29250
    },
    {
      "epoch": 0.8748096617203595,
      "grad_norm": 3.2945287227630615,
      "learning_rate": 0.00019163023788707365,
      "loss": 1.3308,
      "step": 29300
    },
    {
      "epoch": 0.876302510972442,
      "grad_norm": 5.060371398925781,
      "learning_rate": 0.00019161595209115482,
      "loss": 1.3497,
      "step": 29350
    },
    {
      "epoch": 0.8777953602245245,
      "grad_norm": 4.771635055541992,
      "learning_rate": 0.00019160166629523598,
      "loss": 1.3517,
      "step": 29400
    },
    {
      "epoch": 0.879288209476607,
      "grad_norm": 4.408042907714844,
      "learning_rate": 0.00019158738049931715,
      "loss": 1.3636,
      "step": 29450
    },
    {
      "epoch": 0.8807810587286896,
      "grad_norm": 5.183899879455566,
      "learning_rate": 0.0001915730947033983,
      "loss": 1.3569,
      "step": 29500
    },
    {
      "epoch": 0.8822739079807721,
      "grad_norm": 4.176016807556152,
      "learning_rate": 0.00019155880890747948,
      "loss": 1.3659,
      "step": 29550
    },
    {
      "epoch": 0.8837667572328546,
      "grad_norm": 4.783830165863037,
      "learning_rate": 0.00019154452311156064,
      "loss": 1.368,
      "step": 29600
    },
    {
      "epoch": 0.8852596064849372,
      "grad_norm": 3.7985739707946777,
      "learning_rate": 0.0001915302373156418,
      "loss": 1.4091,
      "step": 29650
    },
    {
      "epoch": 0.8867524557370197,
      "grad_norm": 4.401581287384033,
      "learning_rate": 0.00019151595151972297,
      "loss": 1.3613,
      "step": 29700
    },
    {
      "epoch": 0.8882453049891021,
      "grad_norm": 6.214754581451416,
      "learning_rate": 0.00019150166572380414,
      "loss": 1.4077,
      "step": 29750
    },
    {
      "epoch": 0.8897381542411847,
      "grad_norm": 4.197826862335205,
      "learning_rate": 0.0001914873799278853,
      "loss": 1.347,
      "step": 29800
    },
    {
      "epoch": 0.8912310034932672,
      "grad_norm": 3.1535027027130127,
      "learning_rate": 0.0001914730941319665,
      "loss": 1.3452,
      "step": 29850
    },
    {
      "epoch": 0.8927238527453498,
      "grad_norm": 4.618538856506348,
      "learning_rate": 0.00019145880833604763,
      "loss": 1.3417,
      "step": 29900
    },
    {
      "epoch": 0.8942167019974323,
      "grad_norm": 3.913506031036377,
      "learning_rate": 0.00019144452254012882,
      "loss": 1.3503,
      "step": 29950
    },
    {
      "epoch": 0.8957095512495148,
      "grad_norm": 4.013761043548584,
      "learning_rate": 0.00019143023674420996,
      "loss": 1.4109,
      "step": 30000
    },
    {
      "epoch": 0.8972024005015974,
      "grad_norm": 4.6257758140563965,
      "learning_rate": 0.00019141595094829115,
      "loss": 1.3679,
      "step": 30050
    },
    {
      "epoch": 0.8986952497536799,
      "grad_norm": 4.422520637512207,
      "learning_rate": 0.0001914016651523723,
      "loss": 1.3893,
      "step": 30100
    },
    {
      "epoch": 0.9001880990057624,
      "grad_norm": 4.911538124084473,
      "learning_rate": 0.00019138737935645348,
      "loss": 1.3346,
      "step": 30150
    },
    {
      "epoch": 0.901680948257845,
      "grad_norm": 6.410045623779297,
      "learning_rate": 0.00019137309356053464,
      "loss": 1.3651,
      "step": 30200
    },
    {
      "epoch": 0.9031737975099274,
      "grad_norm": 3.8565444946289062,
      "learning_rate": 0.0001913588077646158,
      "loss": 1.3436,
      "step": 30250
    },
    {
      "epoch": 0.9046666467620099,
      "grad_norm": 4.632288455963135,
      "learning_rate": 0.00019134452196869697,
      "loss": 1.3858,
      "step": 30300
    },
    {
      "epoch": 0.9061594960140925,
      "grad_norm": 5.382070064544678,
      "learning_rate": 0.00019133023617277814,
      "loss": 1.3872,
      "step": 30350
    },
    {
      "epoch": 0.907652345266175,
      "grad_norm": 4.425014019012451,
      "learning_rate": 0.0001913159503768593,
      "loss": 1.4488,
      "step": 30400
    },
    {
      "epoch": 0.9091451945182576,
      "grad_norm": 3.7266223430633545,
      "learning_rate": 0.00019130166458094047,
      "loss": 1.376,
      "step": 30450
    },
    {
      "epoch": 0.9106380437703401,
      "grad_norm": 7.25547981262207,
      "learning_rate": 0.00019128737878502163,
      "loss": 1.3488,
      "step": 30500
    },
    {
      "epoch": 0.9121308930224226,
      "grad_norm": 5.824578285217285,
      "learning_rate": 0.0001912730929891028,
      "loss": 1.4001,
      "step": 30550
    },
    {
      "epoch": 0.9136237422745052,
      "grad_norm": 5.808544635772705,
      "learning_rate": 0.00019125880719318396,
      "loss": 1.3132,
      "step": 30600
    },
    {
      "epoch": 0.9151165915265876,
      "grad_norm": 4.343785285949707,
      "learning_rate": 0.00019124452139726515,
      "loss": 1.3592,
      "step": 30650
    },
    {
      "epoch": 0.9166094407786701,
      "grad_norm": 5.0626325607299805,
      "learning_rate": 0.0001912302356013463,
      "loss": 1.4418,
      "step": 30700
    },
    {
      "epoch": 0.9181022900307527,
      "grad_norm": 4.336055755615234,
      "learning_rate": 0.00019121594980542748,
      "loss": 1.3624,
      "step": 30750
    },
    {
      "epoch": 0.9195951392828352,
      "grad_norm": 6.215260982513428,
      "learning_rate": 0.00019120166400950862,
      "loss": 1.4053,
      "step": 30800
    },
    {
      "epoch": 0.9210879885349177,
      "grad_norm": 4.496364116668701,
      "learning_rate": 0.0001911873782135898,
      "loss": 1.4117,
      "step": 30850
    },
    {
      "epoch": 0.9225808377870003,
      "grad_norm": 4.023138046264648,
      "learning_rate": 0.00019117309241767095,
      "loss": 1.3783,
      "step": 30900
    },
    {
      "epoch": 0.9240736870390828,
      "grad_norm": 3.8177783489227295,
      "learning_rate": 0.00019115880662175214,
      "loss": 1.3127,
      "step": 30950
    },
    {
      "epoch": 0.9255665362911654,
      "grad_norm": 3.893087387084961,
      "learning_rate": 0.0001911445208258333,
      "loss": 1.3359,
      "step": 31000
    },
    {
      "epoch": 0.9270593855432478,
      "grad_norm": 3.9967384338378906,
      "learning_rate": 0.00019113023502991447,
      "loss": 1.3897,
      "step": 31050
    },
    {
      "epoch": 0.9285522347953303,
      "grad_norm": 3.2351789474487305,
      "learning_rate": 0.00019111594923399563,
      "loss": 1.3406,
      "step": 31100
    },
    {
      "epoch": 0.9300450840474129,
      "grad_norm": 4.571500778198242,
      "learning_rate": 0.0001911016634380768,
      "loss": 1.3783,
      "step": 31150
    },
    {
      "epoch": 0.9315379332994954,
      "grad_norm": 6.241518020629883,
      "learning_rate": 0.00019108737764215796,
      "loss": 1.3659,
      "step": 31200
    },
    {
      "epoch": 0.9330307825515779,
      "grad_norm": 3.2547826766967773,
      "learning_rate": 0.00019107309184623913,
      "loss": 1.3137,
      "step": 31250
    },
    {
      "epoch": 0.9345236318036605,
      "grad_norm": 5.386812686920166,
      "learning_rate": 0.0001910588060503203,
      "loss": 1.3974,
      "step": 31300
    },
    {
      "epoch": 0.936016481055743,
      "grad_norm": 4.239005088806152,
      "learning_rate": 0.00019104452025440146,
      "loss": 1.3789,
      "step": 31350
    },
    {
      "epoch": 0.9375093303078255,
      "grad_norm": 3.7252326011657715,
      "learning_rate": 0.00019103023445848262,
      "loss": 1.3622,
      "step": 31400
    },
    {
      "epoch": 0.9390021795599081,
      "grad_norm": 5.890969276428223,
      "learning_rate": 0.00019101594866256381,
      "loss": 1.3905,
      "step": 31450
    },
    {
      "epoch": 0.9404950288119905,
      "grad_norm": 5.796853542327881,
      "learning_rate": 0.00019100166286664495,
      "loss": 1.3797,
      "step": 31500
    },
    {
      "epoch": 0.9419878780640731,
      "grad_norm": 4.130026817321777,
      "learning_rate": 0.00019098737707072614,
      "loss": 1.3365,
      "step": 31550
    },
    {
      "epoch": 0.9434807273161556,
      "grad_norm": 3.65081524848938,
      "learning_rate": 0.00019097309127480728,
      "loss": 1.4347,
      "step": 31600
    },
    {
      "epoch": 0.9449735765682381,
      "grad_norm": 4.91404390335083,
      "learning_rate": 0.00019095880547888847,
      "loss": 1.4056,
      "step": 31650
    },
    {
      "epoch": 0.9464664258203207,
      "grad_norm": 4.7272114753723145,
      "learning_rate": 0.00019094451968296964,
      "loss": 1.2828,
      "step": 31700
    },
    {
      "epoch": 0.9479592750724032,
      "grad_norm": 5.111660957336426,
      "learning_rate": 0.0001909302338870508,
      "loss": 1.3486,
      "step": 31750
    },
    {
      "epoch": 0.9494521243244857,
      "grad_norm": 3.1206741333007812,
      "learning_rate": 0.00019091594809113197,
      "loss": 1.3416,
      "step": 31800
    },
    {
      "epoch": 0.9509449735765683,
      "grad_norm": 4.359163761138916,
      "learning_rate": 0.00019090166229521313,
      "loss": 1.3471,
      "step": 31850
    },
    {
      "epoch": 0.9524378228286507,
      "grad_norm": 4.667807102203369,
      "learning_rate": 0.0001908873764992943,
      "loss": 1.4042,
      "step": 31900
    },
    {
      "epoch": 0.9539306720807332,
      "grad_norm": 4.410369873046875,
      "learning_rate": 0.00019087309070337546,
      "loss": 1.3909,
      "step": 31950
    },
    {
      "epoch": 0.9554235213328158,
      "grad_norm": 4.987631320953369,
      "learning_rate": 0.00019085880490745663,
      "loss": 1.3548,
      "step": 32000
    },
    {
      "epoch": 0.9569163705848983,
      "grad_norm": 4.66121768951416,
      "learning_rate": 0.0001908445191115378,
      "loss": 1.3753,
      "step": 32050
    },
    {
      "epoch": 0.9584092198369809,
      "grad_norm": 6.010061740875244,
      "learning_rate": 0.00019083023331561896,
      "loss": 1.4128,
      "step": 32100
    },
    {
      "epoch": 0.9599020690890634,
      "grad_norm": 3.118246078491211,
      "learning_rate": 0.00019081594751970012,
      "loss": 1.3625,
      "step": 32150
    },
    {
      "epoch": 0.9613949183411459,
      "grad_norm": 3.821732759475708,
      "learning_rate": 0.00019080166172378128,
      "loss": 1.3691,
      "step": 32200
    },
    {
      "epoch": 0.9628877675932285,
      "grad_norm": 5.321105480194092,
      "learning_rate": 0.00019078737592786245,
      "loss": 1.3499,
      "step": 32250
    },
    {
      "epoch": 0.964380616845311,
      "grad_norm": 5.963992595672607,
      "learning_rate": 0.00019077309013194361,
      "loss": 1.3145,
      "step": 32300
    },
    {
      "epoch": 0.9658734660973934,
      "grad_norm": 4.071386337280273,
      "learning_rate": 0.00019075880433602478,
      "loss": 1.3215,
      "step": 32350
    },
    {
      "epoch": 0.967366315349476,
      "grad_norm": 3.7641048431396484,
      "learning_rate": 0.00019074451854010594,
      "loss": 1.3846,
      "step": 32400
    },
    {
      "epoch": 0.9688591646015585,
      "grad_norm": 5.985085487365723,
      "learning_rate": 0.0001907302327441871,
      "loss": 1.3383,
      "step": 32450
    },
    {
      "epoch": 0.970352013853641,
      "grad_norm": 6.602528095245361,
      "learning_rate": 0.0001907159469482683,
      "loss": 1.4195,
      "step": 32500
    },
    {
      "epoch": 0.9718448631057236,
      "grad_norm": 5.370469570159912,
      "learning_rate": 0.00019070166115234944,
      "loss": 1.3485,
      "step": 32550
    },
    {
      "epoch": 0.9733377123578061,
      "grad_norm": 4.588457107543945,
      "learning_rate": 0.00019068737535643063,
      "loss": 1.3896,
      "step": 32600
    },
    {
      "epoch": 0.9748305616098887,
      "grad_norm": 6.389341354370117,
      "learning_rate": 0.00019067308956051177,
      "loss": 1.3118,
      "step": 32650
    },
    {
      "epoch": 0.9763234108619712,
      "grad_norm": 4.051975250244141,
      "learning_rate": 0.00019065880376459296,
      "loss": 1.4065,
      "step": 32700
    },
    {
      "epoch": 0.9778162601140536,
      "grad_norm": 5.474813938140869,
      "learning_rate": 0.0001906445179686741,
      "loss": 1.3967,
      "step": 32750
    },
    {
      "epoch": 0.9793091093661362,
      "grad_norm": 4.033454895019531,
      "learning_rate": 0.0001906302321727553,
      "loss": 1.3212,
      "step": 32800
    },
    {
      "epoch": 0.9808019586182187,
      "grad_norm": 4.46946382522583,
      "learning_rate": 0.00019061594637683645,
      "loss": 1.3917,
      "step": 32850
    },
    {
      "epoch": 0.9822948078703012,
      "grad_norm": 4.98955774307251,
      "learning_rate": 0.00019060166058091762,
      "loss": 1.3814,
      "step": 32900
    },
    {
      "epoch": 0.9837876571223838,
      "grad_norm": 4.879435062408447,
      "learning_rate": 0.00019058737478499878,
      "loss": 1.3486,
      "step": 32950
    },
    {
      "epoch": 0.9852805063744663,
      "grad_norm": 4.0941925048828125,
      "learning_rate": 0.00019057308898907995,
      "loss": 1.3289,
      "step": 33000
    },
    {
      "epoch": 0.9867733556265488,
      "grad_norm": 5.2992706298828125,
      "learning_rate": 0.0001905588031931611,
      "loss": 1.3458,
      "step": 33050
    },
    {
      "epoch": 0.9882662048786314,
      "grad_norm": 4.284411430358887,
      "learning_rate": 0.00019054451739724228,
      "loss": 1.3932,
      "step": 33100
    },
    {
      "epoch": 0.9897590541307139,
      "grad_norm": 4.969634532928467,
      "learning_rate": 0.00019053023160132344,
      "loss": 1.4044,
      "step": 33150
    },
    {
      "epoch": 0.9912519033827965,
      "grad_norm": 5.532953262329102,
      "learning_rate": 0.0001905159458054046,
      "loss": 1.3658,
      "step": 33200
    },
    {
      "epoch": 0.9927447526348789,
      "grad_norm": 4.402670383453369,
      "learning_rate": 0.00019050166000948577,
      "loss": 1.343,
      "step": 33250
    },
    {
      "epoch": 0.9942376018869614,
      "grad_norm": 4.458900451660156,
      "learning_rate": 0.00019048737421356696,
      "loss": 1.4002,
      "step": 33300
    },
    {
      "epoch": 0.995730451139044,
      "grad_norm": 3.934340238571167,
      "learning_rate": 0.0001904730884176481,
      "loss": 1.413,
      "step": 33350
    },
    {
      "epoch": 0.9972233003911265,
      "grad_norm": 4.274466514587402,
      "learning_rate": 0.0001904588026217293,
      "loss": 1.3552,
      "step": 33400
    },
    {
      "epoch": 0.998716149643209,
      "grad_norm": 4.774006366729736,
      "learning_rate": 0.00019044451682581043,
      "loss": 1.3825,
      "step": 33450
    },
    {
      "epoch": 1.0002089988952916,
      "grad_norm": 3.870762825012207,
      "learning_rate": 0.00019043023102989162,
      "loss": 1.3715,
      "step": 33500
    },
    {
      "epoch": 1.001701848147374,
      "grad_norm": 3.595686912536621,
      "learning_rate": 0.00019041594523397276,
      "loss": 1.3352,
      "step": 33550
    },
    {
      "epoch": 1.0031946973994565,
      "grad_norm": 5.492071151733398,
      "learning_rate": 0.00019040165943805395,
      "loss": 1.2927,
      "step": 33600
    },
    {
      "epoch": 1.004687546651539,
      "grad_norm": 5.381584167480469,
      "learning_rate": 0.00019038737364213511,
      "loss": 1.2545,
      "step": 33650
    },
    {
      "epoch": 1.0061803959036217,
      "grad_norm": 5.983656883239746,
      "learning_rate": 0.00019037308784621628,
      "loss": 1.2393,
      "step": 33700
    },
    {
      "epoch": 1.0076732451557042,
      "grad_norm": 4.956115245819092,
      "learning_rate": 0.00019035880205029744,
      "loss": 1.2931,
      "step": 33750
    },
    {
      "epoch": 1.0091660944077867,
      "grad_norm": 4.604572296142578,
      "learning_rate": 0.0001903445162543786,
      "loss": 1.2786,
      "step": 33800
    },
    {
      "epoch": 1.0106589436598692,
      "grad_norm": 5.043067932128906,
      "learning_rate": 0.00019033023045845977,
      "loss": 1.2933,
      "step": 33850
    },
    {
      "epoch": 1.0121517929119517,
      "grad_norm": 4.53336238861084,
      "learning_rate": 0.00019031594466254094,
      "loss": 1.3114,
      "step": 33900
    },
    {
      "epoch": 1.0136446421640344,
      "grad_norm": 5.106605052947998,
      "learning_rate": 0.0001903016588666221,
      "loss": 1.2994,
      "step": 33950
    },
    {
      "epoch": 1.0151374914161169,
      "grad_norm": 5.796351909637451,
      "learning_rate": 0.00019028737307070327,
      "loss": 1.2555,
      "step": 34000
    },
    {
      "epoch": 1.0166303406681994,
      "grad_norm": 4.970977306365967,
      "learning_rate": 0.00019027308727478443,
      "loss": 1.3343,
      "step": 34050
    },
    {
      "epoch": 1.0181231899202818,
      "grad_norm": 4.233397960662842,
      "learning_rate": 0.00019025880147886562,
      "loss": 1.2485,
      "step": 34100
    },
    {
      "epoch": 1.0196160391723643,
      "grad_norm": 4.012045383453369,
      "learning_rate": 0.00019024451568294676,
      "loss": 1.2816,
      "step": 34150
    },
    {
      "epoch": 1.0211088884244468,
      "grad_norm": 4.715073585510254,
      "learning_rate": 0.00019023022988702795,
      "loss": 1.2541,
      "step": 34200
    },
    {
      "epoch": 1.0226017376765295,
      "grad_norm": 3.7285947799682617,
      "learning_rate": 0.0001902159440911091,
      "loss": 1.3094,
      "step": 34250
    },
    {
      "epoch": 1.024094586928612,
      "grad_norm": 4.012038707733154,
      "learning_rate": 0.00019020165829519028,
      "loss": 1.2791,
      "step": 34300
    },
    {
      "epoch": 1.0255874361806945,
      "grad_norm": 4.04899263381958,
      "learning_rate": 0.00019018737249927142,
      "loss": 1.3085,
      "step": 34350
    },
    {
      "epoch": 1.027080285432777,
      "grad_norm": 4.380834102630615,
      "learning_rate": 0.0001901730867033526,
      "loss": 1.3212,
      "step": 34400
    },
    {
      "epoch": 1.0285731346848594,
      "grad_norm": 5.275148868560791,
      "learning_rate": 0.00019015880090743378,
      "loss": 1.2749,
      "step": 34450
    },
    {
      "epoch": 1.0300659839369422,
      "grad_norm": 4.162272930145264,
      "learning_rate": 0.00019014451511151494,
      "loss": 1.296,
      "step": 34500
    },
    {
      "epoch": 1.0315588331890246,
      "grad_norm": 4.128006935119629,
      "learning_rate": 0.0001901302293155961,
      "loss": 1.2547,
      "step": 34550
    },
    {
      "epoch": 1.0330516824411071,
      "grad_norm": 3.930121421813965,
      "learning_rate": 0.00019011594351967727,
      "loss": 1.2831,
      "step": 34600
    },
    {
      "epoch": 1.0345445316931896,
      "grad_norm": 4.515873908996582,
      "learning_rate": 0.00019010165772375843,
      "loss": 1.2747,
      "step": 34650
    },
    {
      "epoch": 1.036037380945272,
      "grad_norm": 4.484467506408691,
      "learning_rate": 0.0001900873719278396,
      "loss": 1.289,
      "step": 34700
    },
    {
      "epoch": 1.0375302301973546,
      "grad_norm": 3.8937742710113525,
      "learning_rate": 0.00019007308613192076,
      "loss": 1.2939,
      "step": 34750
    },
    {
      "epoch": 1.0390230794494373,
      "grad_norm": 4.367059230804443,
      "learning_rate": 0.00019005880033600193,
      "loss": 1.2522,
      "step": 34800
    },
    {
      "epoch": 1.0405159287015198,
      "grad_norm": 5.143396854400635,
      "learning_rate": 0.0001900445145400831,
      "loss": 1.2996,
      "step": 34850
    },
    {
      "epoch": 1.0420087779536022,
      "grad_norm": 5.060800552368164,
      "learning_rate": 0.00019003022874416426,
      "loss": 1.3122,
      "step": 34900
    },
    {
      "epoch": 1.0435016272056847,
      "grad_norm": 4.775914192199707,
      "learning_rate": 0.00019001594294824542,
      "loss": 1.2595,
      "step": 34950
    },
    {
      "epoch": 1.0449944764577672,
      "grad_norm": 5.256932258605957,
      "learning_rate": 0.0001900016571523266,
      "loss": 1.3148,
      "step": 35000
    },
    {
      "epoch": 1.04648732570985,
      "grad_norm": 4.253088474273682,
      "learning_rate": 0.00018998737135640775,
      "loss": 1.2575,
      "step": 35050
    },
    {
      "epoch": 1.0479801749619324,
      "grad_norm": 4.784812927246094,
      "learning_rate": 0.00018997308556048892,
      "loss": 1.2911,
      "step": 35100
    },
    {
      "epoch": 1.049473024214015,
      "grad_norm": 3.9015183448791504,
      "learning_rate": 0.0001899587997645701,
      "loss": 1.2647,
      "step": 35150
    },
    {
      "epoch": 1.0509658734660974,
      "grad_norm": 4.729675769805908,
      "learning_rate": 0.00018994451396865125,
      "loss": 1.2876,
      "step": 35200
    },
    {
      "epoch": 1.0524587227181799,
      "grad_norm": 4.921074867248535,
      "learning_rate": 0.00018993022817273244,
      "loss": 1.2956,
      "step": 35250
    },
    {
      "epoch": 1.0539515719702623,
      "grad_norm": 4.466576099395752,
      "learning_rate": 0.00018991594237681357,
      "loss": 1.2938,
      "step": 35300
    },
    {
      "epoch": 1.055444421222345,
      "grad_norm": 4.142183780670166,
      "learning_rate": 0.00018990165658089477,
      "loss": 1.282,
      "step": 35350
    },
    {
      "epoch": 1.0569372704744275,
      "grad_norm": 4.372234344482422,
      "learning_rate": 0.0001898873707849759,
      "loss": 1.2793,
      "step": 35400
    },
    {
      "epoch": 1.05843011972651,
      "grad_norm": 5.349823474884033,
      "learning_rate": 0.0001898730849890571,
      "loss": 1.3678,
      "step": 35450
    },
    {
      "epoch": 1.0599229689785925,
      "grad_norm": 4.57612419128418,
      "learning_rate": 0.00018985879919313826,
      "loss": 1.3256,
      "step": 35500
    },
    {
      "epoch": 1.061415818230675,
      "grad_norm": 5.3342156410217285,
      "learning_rate": 0.00018984451339721942,
      "loss": 1.2897,
      "step": 35550
    },
    {
      "epoch": 1.0629086674827577,
      "grad_norm": 5.361871242523193,
      "learning_rate": 0.0001898302276013006,
      "loss": 1.2933,
      "step": 35600
    },
    {
      "epoch": 1.0644015167348402,
      "grad_norm": 3.6657068729400635,
      "learning_rate": 0.00018981594180538175,
      "loss": 1.3359,
      "step": 35650
    },
    {
      "epoch": 1.0658943659869227,
      "grad_norm": 4.428374290466309,
      "learning_rate": 0.00018980165600946292,
      "loss": 1.3278,
      "step": 35700
    },
    {
      "epoch": 1.0673872152390051,
      "grad_norm": 5.079537391662598,
      "learning_rate": 0.00018978737021354408,
      "loss": 1.2699,
      "step": 35750
    },
    {
      "epoch": 1.0688800644910876,
      "grad_norm": 5.303153038024902,
      "learning_rate": 0.00018977308441762525,
      "loss": 1.2985,
      "step": 35800
    },
    {
      "epoch": 1.0703729137431701,
      "grad_norm": 4.319901943206787,
      "learning_rate": 0.0001897587986217064,
      "loss": 1.2661,
      "step": 35850
    },
    {
      "epoch": 1.0718657629952528,
      "grad_norm": 6.0582451820373535,
      "learning_rate": 0.00018974451282578758,
      "loss": 1.2837,
      "step": 35900
    },
    {
      "epoch": 1.0733586122473353,
      "grad_norm": 5.738245964050293,
      "learning_rate": 0.00018973022702986877,
      "loss": 1.3075,
      "step": 35950
    },
    {
      "epoch": 1.0748514614994178,
      "grad_norm": 4.602933883666992,
      "learning_rate": 0.0001897159412339499,
      "loss": 1.2585,
      "step": 36000
    },
    {
      "epoch": 1.0763443107515003,
      "grad_norm": 4.273153305053711,
      "learning_rate": 0.0001897016554380311,
      "loss": 1.3382,
      "step": 36050
    },
    {
      "epoch": 1.0778371600035828,
      "grad_norm": 4.225873947143555,
      "learning_rate": 0.00018968736964211224,
      "loss": 1.2964,
      "step": 36100
    },
    {
      "epoch": 1.0793300092556652,
      "grad_norm": 4.828727722167969,
      "learning_rate": 0.00018967308384619343,
      "loss": 1.3734,
      "step": 36150
    },
    {
      "epoch": 1.080822858507748,
      "grad_norm": 3.7251532077789307,
      "learning_rate": 0.00018965879805027457,
      "loss": 1.2706,
      "step": 36200
    },
    {
      "epoch": 1.0823157077598304,
      "grad_norm": 4.92055606842041,
      "learning_rate": 0.00018964451225435576,
      "loss": 1.3187,
      "step": 36250
    },
    {
      "epoch": 1.083808557011913,
      "grad_norm": 4.827970027923584,
      "learning_rate": 0.00018963022645843692,
      "loss": 1.2574,
      "step": 36300
    },
    {
      "epoch": 1.0853014062639954,
      "grad_norm": 4.742151737213135,
      "learning_rate": 0.00018961594066251809,
      "loss": 1.3076,
      "step": 36350
    },
    {
      "epoch": 1.086794255516078,
      "grad_norm": 3.70794939994812,
      "learning_rate": 0.00018960165486659925,
      "loss": 1.3075,
      "step": 36400
    },
    {
      "epoch": 1.0882871047681606,
      "grad_norm": 4.062809944152832,
      "learning_rate": 0.00018958736907068042,
      "loss": 1.2628,
      "step": 36450
    },
    {
      "epoch": 1.089779954020243,
      "grad_norm": 5.072466850280762,
      "learning_rate": 0.00018957308327476158,
      "loss": 1.3278,
      "step": 36500
    },
    {
      "epoch": 1.0912728032723256,
      "grad_norm": 4.530171871185303,
      "learning_rate": 0.00018955879747884274,
      "loss": 1.3233,
      "step": 36550
    },
    {
      "epoch": 1.092765652524408,
      "grad_norm": 3.9391825199127197,
      "learning_rate": 0.0001895445116829239,
      "loss": 1.2973,
      "step": 36600
    },
    {
      "epoch": 1.0942585017764905,
      "grad_norm": 6.774008274078369,
      "learning_rate": 0.00018953022588700507,
      "loss": 1.2801,
      "step": 36650
    },
    {
      "epoch": 1.0957513510285732,
      "grad_norm": 4.565464496612549,
      "learning_rate": 0.00018951594009108624,
      "loss": 1.3718,
      "step": 36700
    },
    {
      "epoch": 1.0972442002806557,
      "grad_norm": 5.68241548538208,
      "learning_rate": 0.00018950165429516743,
      "loss": 1.2928,
      "step": 36750
    },
    {
      "epoch": 1.0987370495327382,
      "grad_norm": 4.235448360443115,
      "learning_rate": 0.00018948736849924857,
      "loss": 1.3097,
      "step": 36800
    },
    {
      "epoch": 1.1002298987848207,
      "grad_norm": 4.672369003295898,
      "learning_rate": 0.00018947308270332976,
      "loss": 1.2875,
      "step": 36850
    },
    {
      "epoch": 1.1017227480369032,
      "grad_norm": 6.449750900268555,
      "learning_rate": 0.0001894587969074109,
      "loss": 1.2823,
      "step": 36900
    },
    {
      "epoch": 1.1032155972889857,
      "grad_norm": 5.093149662017822,
      "learning_rate": 0.0001894445111114921,
      "loss": 1.315,
      "step": 36950
    },
    {
      "epoch": 1.1047084465410684,
      "grad_norm": 5.433828353881836,
      "learning_rate": 0.00018943022531557323,
      "loss": 1.2781,
      "step": 37000
    },
    {
      "epoch": 1.1062012957931509,
      "grad_norm": 5.176681995391846,
      "learning_rate": 0.00018941593951965442,
      "loss": 1.2652,
      "step": 37050
    },
    {
      "epoch": 1.1076941450452333,
      "grad_norm": 4.09026575088501,
      "learning_rate": 0.00018940165372373558,
      "loss": 1.3045,
      "step": 37100
    },
    {
      "epoch": 1.1091869942973158,
      "grad_norm": 5.223085880279541,
      "learning_rate": 0.00018938736792781675,
      "loss": 1.3009,
      "step": 37150
    },
    {
      "epoch": 1.1106798435493983,
      "grad_norm": 5.105752944946289,
      "learning_rate": 0.0001893730821318979,
      "loss": 1.2915,
      "step": 37200
    },
    {
      "epoch": 1.1121726928014808,
      "grad_norm": 3.945962905883789,
      "learning_rate": 0.00018935879633597908,
      "loss": 1.341,
      "step": 37250
    },
    {
      "epoch": 1.1136655420535635,
      "grad_norm": 4.856802463531494,
      "learning_rate": 0.00018934451054006024,
      "loss": 1.3318,
      "step": 37300
    },
    {
      "epoch": 1.115158391305646,
      "grad_norm": 4.858597755432129,
      "learning_rate": 0.0001893302247441414,
      "loss": 1.2756,
      "step": 37350
    },
    {
      "epoch": 1.1166512405577285,
      "grad_norm": 4.594838619232178,
      "learning_rate": 0.00018931593894822257,
      "loss": 1.2712,
      "step": 37400
    },
    {
      "epoch": 1.118144089809811,
      "grad_norm": 4.745405673980713,
      "learning_rate": 0.00018930165315230374,
      "loss": 1.2878,
      "step": 37450
    },
    {
      "epoch": 1.1196369390618934,
      "grad_norm": 4.112718105316162,
      "learning_rate": 0.0001892873673563849,
      "loss": 1.3337,
      "step": 37500
    },
    {
      "epoch": 1.1211297883139761,
      "grad_norm": 4.297769069671631,
      "learning_rate": 0.00018927308156046607,
      "loss": 1.3324,
      "step": 37550
    },
    {
      "epoch": 1.1226226375660586,
      "grad_norm": 5.46763801574707,
      "learning_rate": 0.00018925879576454723,
      "loss": 1.276,
      "step": 37600
    },
    {
      "epoch": 1.124115486818141,
      "grad_norm": 5.648976802825928,
      "learning_rate": 0.0001892445099686284,
      "loss": 1.3238,
      "step": 37650
    },
    {
      "epoch": 1.1256083360702236,
      "grad_norm": 4.185100078582764,
      "learning_rate": 0.00018923022417270956,
      "loss": 1.2848,
      "step": 37700
    },
    {
      "epoch": 1.127101185322306,
      "grad_norm": 4.676313877105713,
      "learning_rate": 0.00018921593837679072,
      "loss": 1.3038,
      "step": 37750
    },
    {
      "epoch": 1.1285940345743888,
      "grad_norm": 4.426568508148193,
      "learning_rate": 0.00018920165258087192,
      "loss": 1.3343,
      "step": 37800
    },
    {
      "epoch": 1.1300868838264713,
      "grad_norm": 4.887205600738525,
      "learning_rate": 0.00018918736678495305,
      "loss": 1.3456,
      "step": 37850
    },
    {
      "epoch": 1.1315797330785538,
      "grad_norm": 5.455615043640137,
      "learning_rate": 0.00018917308098903424,
      "loss": 1.3258,
      "step": 37900
    },
    {
      "epoch": 1.1330725823306362,
      "grad_norm": 5.676678657531738,
      "learning_rate": 0.00018915879519311538,
      "loss": 1.3702,
      "step": 37950
    },
    {
      "epoch": 1.1345654315827187,
      "grad_norm": 3.9528415203094482,
      "learning_rate": 0.00018914450939719657,
      "loss": 1.3097,
      "step": 38000
    },
    {
      "epoch": 1.1360582808348014,
      "grad_norm": 4.538026332855225,
      "learning_rate": 0.0001891302236012777,
      "loss": 1.316,
      "step": 38050
    },
    {
      "epoch": 1.137551130086884,
      "grad_norm": 4.46422815322876,
      "learning_rate": 0.0001891159378053589,
      "loss": 1.2479,
      "step": 38100
    },
    {
      "epoch": 1.1390439793389664,
      "grad_norm": 4.823958396911621,
      "learning_rate": 0.00018910165200944007,
      "loss": 1.2616,
      "step": 38150
    },
    {
      "epoch": 1.1405368285910489,
      "grad_norm": 5.7128167152404785,
      "learning_rate": 0.00018908736621352123,
      "loss": 1.2853,
      "step": 38200
    },
    {
      "epoch": 1.1420296778431314,
      "grad_norm": 3.5420877933502197,
      "learning_rate": 0.0001890730804176024,
      "loss": 1.3008,
      "step": 38250
    },
    {
      "epoch": 1.1435225270952138,
      "grad_norm": 4.617981910705566,
      "learning_rate": 0.00018905879462168356,
      "loss": 1.2888,
      "step": 38300
    },
    {
      "epoch": 1.1450153763472963,
      "grad_norm": 4.542544364929199,
      "learning_rate": 0.00018904450882576473,
      "loss": 1.3009,
      "step": 38350
    },
    {
      "epoch": 1.146508225599379,
      "grad_norm": 4.301428318023682,
      "learning_rate": 0.0001890302230298459,
      "loss": 1.2834,
      "step": 38400
    },
    {
      "epoch": 1.1480010748514615,
      "grad_norm": 4.136764049530029,
      "learning_rate": 0.00018901593723392706,
      "loss": 1.3895,
      "step": 38450
    },
    {
      "epoch": 1.149493924103544,
      "grad_norm": 5.59256649017334,
      "learning_rate": 0.00018900165143800822,
      "loss": 1.295,
      "step": 38500
    },
    {
      "epoch": 1.1509867733556265,
      "grad_norm": 3.4925365447998047,
      "learning_rate": 0.00018898736564208939,
      "loss": 1.3385,
      "step": 38550
    },
    {
      "epoch": 1.152479622607709,
      "grad_norm": 4.884555816650391,
      "learning_rate": 0.00018897307984617058,
      "loss": 1.2776,
      "step": 38600
    },
    {
      "epoch": 1.1539724718597917,
      "grad_norm": 3.866908550262451,
      "learning_rate": 0.00018895879405025171,
      "loss": 1.3173,
      "step": 38650
    },
    {
      "epoch": 1.1554653211118742,
      "grad_norm": 4.8104939460754395,
      "learning_rate": 0.0001889445082543329,
      "loss": 1.2988,
      "step": 38700
    },
    {
      "epoch": 1.1569581703639567,
      "grad_norm": 3.814675807952881,
      "learning_rate": 0.00018893022245841404,
      "loss": 1.348,
      "step": 38750
    },
    {
      "epoch": 1.1584510196160391,
      "grad_norm": 5.720306873321533,
      "learning_rate": 0.00018891593666249524,
      "loss": 1.2844,
      "step": 38800
    },
    {
      "epoch": 1.1599438688681216,
      "grad_norm": 4.06850528717041,
      "learning_rate": 0.00018890165086657637,
      "loss": 1.3514,
      "step": 38850
    },
    {
      "epoch": 1.1614367181202043,
      "grad_norm": 6.193358898162842,
      "learning_rate": 0.00018888736507065757,
      "loss": 1.3587,
      "step": 38900
    },
    {
      "epoch": 1.1629295673722868,
      "grad_norm": 4.8998212814331055,
      "learning_rate": 0.00018887307927473873,
      "loss": 1.317,
      "step": 38950
    },
    {
      "epoch": 1.1644224166243693,
      "grad_norm": 4.4293107986450195,
      "learning_rate": 0.0001888587934788199,
      "loss": 1.3473,
      "step": 39000
    },
    {
      "epoch": 1.1659152658764518,
      "grad_norm": 6.795536994934082,
      "learning_rate": 0.00018884450768290106,
      "loss": 1.2793,
      "step": 39050
    },
    {
      "epoch": 1.1674081151285343,
      "grad_norm": 4.158294200897217,
      "learning_rate": 0.00018883022188698222,
      "loss": 1.3175,
      "step": 39100
    },
    {
      "epoch": 1.168900964380617,
      "grad_norm": 5.839204788208008,
      "learning_rate": 0.0001888159360910634,
      "loss": 1.2931,
      "step": 39150
    },
    {
      "epoch": 1.1703938136326995,
      "grad_norm": 6.633917331695557,
      "learning_rate": 0.00018880165029514455,
      "loss": 1.3053,
      "step": 39200
    },
    {
      "epoch": 1.171886662884782,
      "grad_norm": 4.409125328063965,
      "learning_rate": 0.00018878736449922572,
      "loss": 1.3074,
      "step": 39250
    },
    {
      "epoch": 1.1733795121368644,
      "grad_norm": 4.820318698883057,
      "learning_rate": 0.00018877307870330688,
      "loss": 1.329,
      "step": 39300
    },
    {
      "epoch": 1.174872361388947,
      "grad_norm": 5.104337215423584,
      "learning_rate": 0.00018875879290738805,
      "loss": 1.2628,
      "step": 39350
    },
    {
      "epoch": 1.1763652106410294,
      "grad_norm": 5.449405670166016,
      "learning_rate": 0.00018874450711146924,
      "loss": 1.2672,
      "step": 39400
    },
    {
      "epoch": 1.1778580598931119,
      "grad_norm": 5.3521504402160645,
      "learning_rate": 0.00018873022131555038,
      "loss": 1.3236,
      "step": 39450
    },
    {
      "epoch": 1.1793509091451946,
      "grad_norm": 5.748175144195557,
      "learning_rate": 0.00018871593551963157,
      "loss": 1.284,
      "step": 39500
    },
    {
      "epoch": 1.180843758397277,
      "grad_norm": 4.418118953704834,
      "learning_rate": 0.0001887016497237127,
      "loss": 1.299,
      "step": 39550
    },
    {
      "epoch": 1.1823366076493596,
      "grad_norm": 4.267107009887695,
      "learning_rate": 0.0001886873639277939,
      "loss": 1.3287,
      "step": 39600
    },
    {
      "epoch": 1.183829456901442,
      "grad_norm": 4.607259273529053,
      "learning_rate": 0.00018867307813187503,
      "loss": 1.3087,
      "step": 39650
    },
    {
      "epoch": 1.1853223061535245,
      "grad_norm": 4.254667282104492,
      "learning_rate": 0.00018865879233595623,
      "loss": 1.3015,
      "step": 39700
    },
    {
      "epoch": 1.1868151554056072,
      "grad_norm": 5.307114124298096,
      "learning_rate": 0.0001886445065400374,
      "loss": 1.2802,
      "step": 39750
    },
    {
      "epoch": 1.1883080046576897,
      "grad_norm": 3.9305639266967773,
      "learning_rate": 0.00018863022074411856,
      "loss": 1.2975,
      "step": 39800
    },
    {
      "epoch": 1.1898008539097722,
      "grad_norm": 4.650544166564941,
      "learning_rate": 0.00018861593494819972,
      "loss": 1.3239,
      "step": 39850
    },
    {
      "epoch": 1.1912937031618547,
      "grad_norm": 4.182717800140381,
      "learning_rate": 0.00018860164915228086,
      "loss": 1.3948,
      "step": 39900
    },
    {
      "epoch": 1.1927865524139372,
      "grad_norm": 5.322524547576904,
      "learning_rate": 0.00018858736335636205,
      "loss": 1.3411,
      "step": 39950
    },
    {
      "epoch": 1.1942794016660199,
      "grad_norm": 5.221969127655029,
      "learning_rate": 0.0001885730775604432,
      "loss": 1.3047,
      "step": 40000
    },
    {
      "epoch": 1.1957722509181024,
      "grad_norm": 4.373313903808594,
      "learning_rate": 0.00018855879176452438,
      "loss": 1.366,
      "step": 40050
    },
    {
      "epoch": 1.1972651001701848,
      "grad_norm": 4.842396259307861,
      "learning_rate": 0.00018854450596860554,
      "loss": 1.3294,
      "step": 40100
    },
    {
      "epoch": 1.1987579494222673,
      "grad_norm": 5.024688720703125,
      "learning_rate": 0.0001885302201726867,
      "loss": 1.3531,
      "step": 40150
    },
    {
      "epoch": 1.2002507986743498,
      "grad_norm": 5.733693599700928,
      "learning_rate": 0.00018851593437676787,
      "loss": 1.3177,
      "step": 40200
    },
    {
      "epoch": 1.2017436479264325,
      "grad_norm": 4.634098529815674,
      "learning_rate": 0.00018850164858084904,
      "loss": 1.3534,
      "step": 40250
    },
    {
      "epoch": 1.203236497178515,
      "grad_norm": 5.662175178527832,
      "learning_rate": 0.0001884873627849302,
      "loss": 1.2813,
      "step": 40300
    },
    {
      "epoch": 1.2047293464305975,
      "grad_norm": 4.25814151763916,
      "learning_rate": 0.00018847307698901137,
      "loss": 1.3704,
      "step": 40350
    },
    {
      "epoch": 1.20622219568268,
      "grad_norm": 4.733590126037598,
      "learning_rate": 0.00018845879119309253,
      "loss": 1.2999,
      "step": 40400
    },
    {
      "epoch": 1.2077150449347624,
      "grad_norm": 5.873198509216309,
      "learning_rate": 0.0001884445053971737,
      "loss": 1.3697,
      "step": 40450
    },
    {
      "epoch": 1.209207894186845,
      "grad_norm": 4.847556114196777,
      "learning_rate": 0.00018843021960125486,
      "loss": 1.2706,
      "step": 40500
    },
    {
      "epoch": 1.2107007434389274,
      "grad_norm": 4.76710319519043,
      "learning_rate": 0.00018841593380533605,
      "loss": 1.3873,
      "step": 40550
    },
    {
      "epoch": 1.2121935926910101,
      "grad_norm": 4.693939685821533,
      "learning_rate": 0.0001884016480094172,
      "loss": 1.2752,
      "step": 40600
    },
    {
      "epoch": 1.2136864419430926,
      "grad_norm": 5.377971649169922,
      "learning_rate": 0.00018838736221349838,
      "loss": 1.3083,
      "step": 40650
    },
    {
      "epoch": 1.215179291195175,
      "grad_norm": 4.257247447967529,
      "learning_rate": 0.00018837307641757952,
      "loss": 1.3318,
      "step": 40700
    },
    {
      "epoch": 1.2166721404472576,
      "grad_norm": 4.820394039154053,
      "learning_rate": 0.0001883587906216607,
      "loss": 1.3063,
      "step": 40750
    },
    {
      "epoch": 1.21816498969934,
      "grad_norm": 3.7783172130584717,
      "learning_rate": 0.00018834450482574188,
      "loss": 1.2765,
      "step": 40800
    },
    {
      "epoch": 1.2196578389514228,
      "grad_norm": 5.447601795196533,
      "learning_rate": 0.00018833021902982304,
      "loss": 1.3067,
      "step": 40850
    },
    {
      "epoch": 1.2211506882035053,
      "grad_norm": 4.7725510597229,
      "learning_rate": 0.0001883159332339042,
      "loss": 1.3366,
      "step": 40900
    },
    {
      "epoch": 1.2226435374555877,
      "grad_norm": 5.6868672370910645,
      "learning_rate": 0.00018830164743798537,
      "loss": 1.3441,
      "step": 40950
    },
    {
      "epoch": 1.2241363867076702,
      "grad_norm": 8.460538864135742,
      "learning_rate": 0.00018828736164206653,
      "loss": 1.3305,
      "step": 41000
    },
    {
      "epoch": 1.2256292359597527,
      "grad_norm": 5.594083309173584,
      "learning_rate": 0.0001882730758461477,
      "loss": 1.2825,
      "step": 41050
    },
    {
      "epoch": 1.2271220852118354,
      "grad_norm": 6.125982284545898,
      "learning_rate": 0.00018825879005022886,
      "loss": 1.3253,
      "step": 41100
    },
    {
      "epoch": 1.228614934463918,
      "grad_norm": 4.173381328582764,
      "learning_rate": 0.00018824450425431003,
      "loss": 1.3363,
      "step": 41150
    },
    {
      "epoch": 1.2301077837160004,
      "grad_norm": 4.884373664855957,
      "learning_rate": 0.0001882302184583912,
      "loss": 1.3124,
      "step": 41200
    },
    {
      "epoch": 1.2316006329680829,
      "grad_norm": 6.418700695037842,
      "learning_rate": 0.00018821593266247239,
      "loss": 1.3123,
      "step": 41250
    },
    {
      "epoch": 1.2330934822201653,
      "grad_norm": 4.717264175415039,
      "learning_rate": 0.00018820164686655352,
      "loss": 1.287,
      "step": 41300
    },
    {
      "epoch": 1.234586331472248,
      "grad_norm": 5.55648136138916,
      "learning_rate": 0.00018818736107063471,
      "loss": 1.3426,
      "step": 41350
    },
    {
      "epoch": 1.2360791807243305,
      "grad_norm": 5.190313339233398,
      "learning_rate": 0.00018817307527471585,
      "loss": 1.2945,
      "step": 41400
    },
    {
      "epoch": 1.237572029976413,
      "grad_norm": 6.027787208557129,
      "learning_rate": 0.00018815878947879704,
      "loss": 1.3313,
      "step": 41450
    },
    {
      "epoch": 1.2390648792284955,
      "grad_norm": 4.4704413414001465,
      "learning_rate": 0.00018814450368287818,
      "loss": 1.3031,
      "step": 41500
    },
    {
      "epoch": 1.240557728480578,
      "grad_norm": 4.408693790435791,
      "learning_rate": 0.00018813021788695937,
      "loss": 1.287,
      "step": 41550
    },
    {
      "epoch": 1.2420505777326605,
      "grad_norm": 5.230643272399902,
      "learning_rate": 0.00018811593209104054,
      "loss": 1.2846,
      "step": 41600
    },
    {
      "epoch": 1.243543426984743,
      "grad_norm": 4.442893981933594,
      "learning_rate": 0.0001881016462951217,
      "loss": 1.3199,
      "step": 41650
    },
    {
      "epoch": 1.2450362762368257,
      "grad_norm": 3.9362857341766357,
      "learning_rate": 0.00018808736049920287,
      "loss": 1.3301,
      "step": 41700
    },
    {
      "epoch": 1.2465291254889082,
      "grad_norm": 4.765340805053711,
      "learning_rate": 0.00018807307470328403,
      "loss": 1.3092,
      "step": 41750
    },
    {
      "epoch": 1.2480219747409906,
      "grad_norm": 4.7868146896362305,
      "learning_rate": 0.0001880587889073652,
      "loss": 1.3077,
      "step": 41800
    },
    {
      "epoch": 1.2495148239930731,
      "grad_norm": 5.724763870239258,
      "learning_rate": 0.00018804450311144636,
      "loss": 1.3035,
      "step": 41850
    },
    {
      "epoch": 1.2510076732451556,
      "grad_norm": 5.832367897033691,
      "learning_rate": 0.00018803021731552753,
      "loss": 1.3703,
      "step": 41900
    },
    {
      "epoch": 1.2525005224972383,
      "grad_norm": 5.079206466674805,
      "learning_rate": 0.0001880159315196087,
      "loss": 1.3328,
      "step": 41950
    },
    {
      "epoch": 1.2539933717493208,
      "grad_norm": 5.704753398895264,
      "learning_rate": 0.00018800164572368986,
      "loss": 1.354,
      "step": 42000
    },
    {
      "epoch": 1.2554862210014033,
      "grad_norm": 5.396346569061279,
      "learning_rate": 0.00018798735992777105,
      "loss": 1.2794,
      "step": 42050
    },
    {
      "epoch": 1.2569790702534858,
      "grad_norm": 5.461711406707764,
      "learning_rate": 0.00018797307413185218,
      "loss": 1.3335,
      "step": 42100
    },
    {
      "epoch": 1.2584719195055682,
      "grad_norm": 3.3528285026550293,
      "learning_rate": 0.00018795878833593338,
      "loss": 1.296,
      "step": 42150
    },
    {
      "epoch": 1.259964768757651,
      "grad_norm": 4.413696765899658,
      "learning_rate": 0.00018794450254001451,
      "loss": 1.2828,
      "step": 42200
    },
    {
      "epoch": 1.2614576180097334,
      "grad_norm": 5.430131435394287,
      "learning_rate": 0.0001879302167440957,
      "loss": 1.3107,
      "step": 42250
    },
    {
      "epoch": 1.262950467261816,
      "grad_norm": 4.623260974884033,
      "learning_rate": 0.00018791593094817684,
      "loss": 1.3252,
      "step": 42300
    },
    {
      "epoch": 1.2644433165138984,
      "grad_norm": 5.377962112426758,
      "learning_rate": 0.00018790164515225803,
      "loss": 1.3073,
      "step": 42350
    },
    {
      "epoch": 1.265936165765981,
      "grad_norm": 5.34269905090332,
      "learning_rate": 0.0001878873593563392,
      "loss": 1.3705,
      "step": 42400
    },
    {
      "epoch": 1.2674290150180636,
      "grad_norm": 5.574904918670654,
      "learning_rate": 0.00018787307356042036,
      "loss": 1.3848,
      "step": 42450
    },
    {
      "epoch": 1.268921864270146,
      "grad_norm": 4.49995231628418,
      "learning_rate": 0.00018785878776450153,
      "loss": 1.3169,
      "step": 42500
    },
    {
      "epoch": 1.2704147135222286,
      "grad_norm": 3.9160735607147217,
      "learning_rate": 0.00018784450196858267,
      "loss": 1.2994,
      "step": 42550
    },
    {
      "epoch": 1.271907562774311,
      "grad_norm": 6.018070697784424,
      "learning_rate": 0.00018783021617266386,
      "loss": 1.3213,
      "step": 42600
    },
    {
      "epoch": 1.2734004120263935,
      "grad_norm": 5.017271041870117,
      "learning_rate": 0.000187815930376745,
      "loss": 1.2986,
      "step": 42650
    },
    {
      "epoch": 1.274893261278476,
      "grad_norm": 4.147556304931641,
      "learning_rate": 0.0001878016445808262,
      "loss": 1.2738,
      "step": 42700
    },
    {
      "epoch": 1.2763861105305585,
      "grad_norm": 4.864907741546631,
      "learning_rate": 0.00018778735878490735,
      "loss": 1.3414,
      "step": 42750
    },
    {
      "epoch": 1.2778789597826412,
      "grad_norm": 4.564859390258789,
      "learning_rate": 0.00018777307298898852,
      "loss": 1.3028,
      "step": 42800
    },
    {
      "epoch": 1.2793718090347237,
      "grad_norm": 4.96591854095459,
      "learning_rate": 0.00018775878719306968,
      "loss": 1.3197,
      "step": 42850
    },
    {
      "epoch": 1.2808646582868062,
      "grad_norm": 4.384543418884277,
      "learning_rate": 0.00018774450139715085,
      "loss": 1.2802,
      "step": 42900
    },
    {
      "epoch": 1.2823575075388887,
      "grad_norm": 3.956608295440674,
      "learning_rate": 0.000187730215601232,
      "loss": 1.2764,
      "step": 42950
    },
    {
      "epoch": 1.2838503567909711,
      "grad_norm": 4.815617561340332,
      "learning_rate": 0.00018771592980531318,
      "loss": 1.2384,
      "step": 43000
    },
    {
      "epoch": 1.2853432060430539,
      "grad_norm": 4.381842136383057,
      "learning_rate": 0.00018770164400939434,
      "loss": 1.3016,
      "step": 43050
    },
    {
      "epoch": 1.2868360552951363,
      "grad_norm": 6.777477741241455,
      "learning_rate": 0.0001876873582134755,
      "loss": 1.2956,
      "step": 43100
    },
    {
      "epoch": 1.2883289045472188,
      "grad_norm": 4.342850208282471,
      "learning_rate": 0.00018767307241755667,
      "loss": 1.338,
      "step": 43150
    },
    {
      "epoch": 1.2898217537993013,
      "grad_norm": 5.081860065460205,
      "learning_rate": 0.00018765878662163786,
      "loss": 1.292,
      "step": 43200
    },
    {
      "epoch": 1.2913146030513838,
      "grad_norm": 4.7942986488342285,
      "learning_rate": 0.000187644500825719,
      "loss": 1.2969,
      "step": 43250
    },
    {
      "epoch": 1.2928074523034665,
      "grad_norm": 3.766878604888916,
      "learning_rate": 0.0001876302150298002,
      "loss": 1.3561,
      "step": 43300
    },
    {
      "epoch": 1.294300301555549,
      "grad_norm": 5.662395000457764,
      "learning_rate": 0.00018761592923388133,
      "loss": 1.2993,
      "step": 43350
    },
    {
      "epoch": 1.2957931508076315,
      "grad_norm": 4.3839545249938965,
      "learning_rate": 0.00018760164343796252,
      "loss": 1.3283,
      "step": 43400
    },
    {
      "epoch": 1.297286000059714,
      "grad_norm": 5.862764835357666,
      "learning_rate": 0.00018758735764204368,
      "loss": 1.3876,
      "step": 43450
    },
    {
      "epoch": 1.2987788493117964,
      "grad_norm": 4.883128643035889,
      "learning_rate": 0.00018757307184612485,
      "loss": 1.3456,
      "step": 43500
    },
    {
      "epoch": 1.3002716985638791,
      "grad_norm": 5.6903533935546875,
      "learning_rate": 0.000187558786050206,
      "loss": 1.351,
      "step": 43550
    },
    {
      "epoch": 1.3017645478159616,
      "grad_norm": 3.8554911613464355,
      "learning_rate": 0.00018754450025428718,
      "loss": 1.3385,
      "step": 43600
    },
    {
      "epoch": 1.303257397068044,
      "grad_norm": 4.693499565124512,
      "learning_rate": 0.00018753021445836834,
      "loss": 1.2923,
      "step": 43650
    },
    {
      "epoch": 1.3047502463201266,
      "grad_norm": 4.351410388946533,
      "learning_rate": 0.0001875159286624495,
      "loss": 1.37,
      "step": 43700
    },
    {
      "epoch": 1.306243095572209,
      "grad_norm": 3.915044069290161,
      "learning_rate": 0.00018750164286653067,
      "loss": 1.3378,
      "step": 43750
    },
    {
      "epoch": 1.3077359448242916,
      "grad_norm": 5.484678745269775,
      "learning_rate": 0.00018748735707061184,
      "loss": 1.2753,
      "step": 43800
    },
    {
      "epoch": 1.309228794076374,
      "grad_norm": 4.64120626449585,
      "learning_rate": 0.000187473071274693,
      "loss": 1.346,
      "step": 43850
    },
    {
      "epoch": 1.3107216433284568,
      "grad_norm": 6.287013053894043,
      "learning_rate": 0.00018745878547877417,
      "loss": 1.3233,
      "step": 43900
    },
    {
      "epoch": 1.3122144925805392,
      "grad_norm": 4.840941905975342,
      "learning_rate": 0.00018744449968285533,
      "loss": 1.3651,
      "step": 43950
    },
    {
      "epoch": 1.3137073418326217,
      "grad_norm": 4.027961730957031,
      "learning_rate": 0.00018743021388693652,
      "loss": 1.3241,
      "step": 44000
    },
    {
      "epoch": 1.3152001910847042,
      "grad_norm": 5.4013190269470215,
      "learning_rate": 0.00018741592809101766,
      "loss": 1.3048,
      "step": 44050
    },
    {
      "epoch": 1.3166930403367867,
      "grad_norm": 5.291625022888184,
      "learning_rate": 0.00018740164229509885,
      "loss": 1.3171,
      "step": 44100
    },
    {
      "epoch": 1.3181858895888694,
      "grad_norm": 3.7480239868164062,
      "learning_rate": 0.00018738735649918,
      "loss": 1.3527,
      "step": 44150
    },
    {
      "epoch": 1.3196787388409519,
      "grad_norm": 3.080198287963867,
      "learning_rate": 0.00018737307070326118,
      "loss": 1.3782,
      "step": 44200
    },
    {
      "epoch": 1.3211715880930344,
      "grad_norm": 5.115420818328857,
      "learning_rate": 0.00018735878490734235,
      "loss": 1.3359,
      "step": 44250
    },
    {
      "epoch": 1.3226644373451169,
      "grad_norm": 4.522632598876953,
      "learning_rate": 0.0001873444991114235,
      "loss": 1.3557,
      "step": 44300
    },
    {
      "epoch": 1.3241572865971993,
      "grad_norm": 5.980693817138672,
      "learning_rate": 0.00018733021331550468,
      "loss": 1.3356,
      "step": 44350
    },
    {
      "epoch": 1.325650135849282,
      "grad_norm": 4.4641947746276855,
      "learning_rate": 0.00018731592751958584,
      "loss": 1.3179,
      "step": 44400
    },
    {
      "epoch": 1.3271429851013645,
      "grad_norm": 5.057164669036865,
      "learning_rate": 0.000187301641723667,
      "loss": 1.2966,
      "step": 44450
    },
    {
      "epoch": 1.328635834353447,
      "grad_norm": 4.383616924285889,
      "learning_rate": 0.00018728735592774817,
      "loss": 1.317,
      "step": 44500
    },
    {
      "epoch": 1.3301286836055295,
      "grad_norm": 5.158847332000732,
      "learning_rate": 0.00018727307013182933,
      "loss": 1.3278,
      "step": 44550
    },
    {
      "epoch": 1.331621532857612,
      "grad_norm": 4.217287063598633,
      "learning_rate": 0.0001872587843359105,
      "loss": 1.2695,
      "step": 44600
    },
    {
      "epoch": 1.3331143821096947,
      "grad_norm": 3.9984827041625977,
      "learning_rate": 0.00018724449853999166,
      "loss": 1.287,
      "step": 44650
    },
    {
      "epoch": 1.3346072313617772,
      "grad_norm": 3.536656379699707,
      "learning_rate": 0.00018723021274407285,
      "loss": 1.3528,
      "step": 44700
    },
    {
      "epoch": 1.3361000806138597,
      "grad_norm": 8.800326347351074,
      "learning_rate": 0.000187215926948154,
      "loss": 1.3632,
      "step": 44750
    },
    {
      "epoch": 1.3375929298659421,
      "grad_norm": 5.808169364929199,
      "learning_rate": 0.00018720164115223518,
      "loss": 1.3802,
      "step": 44800
    },
    {
      "epoch": 1.3390857791180246,
      "grad_norm": 3.631208896636963,
      "learning_rate": 0.00018718735535631632,
      "loss": 1.3953,
      "step": 44850
    },
    {
      "epoch": 1.340578628370107,
      "grad_norm": 3.4300222396850586,
      "learning_rate": 0.0001871730695603975,
      "loss": 1.3154,
      "step": 44900
    },
    {
      "epoch": 1.3420714776221896,
      "grad_norm": 4.281446933746338,
      "learning_rate": 0.00018715878376447865,
      "loss": 1.3453,
      "step": 44950
    },
    {
      "epoch": 1.3435643268742723,
      "grad_norm": 4.406055450439453,
      "learning_rate": 0.00018714449796855984,
      "loss": 1.3131,
      "step": 45000
    },
    {
      "epoch": 1.3450571761263548,
      "grad_norm": 4.559666156768799,
      "learning_rate": 0.000187130212172641,
      "loss": 1.3005,
      "step": 45050
    },
    {
      "epoch": 1.3465500253784373,
      "grad_norm": 4.3830342292785645,
      "learning_rate": 0.00018711592637672217,
      "loss": 1.2796,
      "step": 45100
    },
    {
      "epoch": 1.3480428746305197,
      "grad_norm": 3.840522289276123,
      "learning_rate": 0.00018710164058080334,
      "loss": 1.3683,
      "step": 45150
    },
    {
      "epoch": 1.3495357238826022,
      "grad_norm": 6.221634387969971,
      "learning_rate": 0.00018708735478488447,
      "loss": 1.3393,
      "step": 45200
    },
    {
      "epoch": 1.351028573134685,
      "grad_norm": 5.086376190185547,
      "learning_rate": 0.00018707306898896567,
      "loss": 1.3007,
      "step": 45250
    },
    {
      "epoch": 1.3525214223867674,
      "grad_norm": 5.3263373374938965,
      "learning_rate": 0.0001870587831930468,
      "loss": 1.3334,
      "step": 45300
    },
    {
      "epoch": 1.35401427163885,
      "grad_norm": 3.790532350540161,
      "learning_rate": 0.000187044497397128,
      "loss": 1.3282,
      "step": 45350
    },
    {
      "epoch": 1.3555071208909324,
      "grad_norm": 5.929101943969727,
      "learning_rate": 0.00018703021160120916,
      "loss": 1.3048,
      "step": 45400
    },
    {
      "epoch": 1.3569999701430149,
      "grad_norm": 4.900948524475098,
      "learning_rate": 0.00018701592580529032,
      "loss": 1.3282,
      "step": 45450
    },
    {
      "epoch": 1.3584928193950976,
      "grad_norm": 4.735415935516357,
      "learning_rate": 0.0001870016400093715,
      "loss": 1.3076,
      "step": 45500
    },
    {
      "epoch": 1.35998566864718,
      "grad_norm": 3.394000768661499,
      "learning_rate": 0.00018698735421345265,
      "loss": 1.3283,
      "step": 45550
    },
    {
      "epoch": 1.3614785178992626,
      "grad_norm": 4.902571678161621,
      "learning_rate": 0.00018697306841753382,
      "loss": 1.3456,
      "step": 45600
    },
    {
      "epoch": 1.362971367151345,
      "grad_norm": 3.773273229598999,
      "learning_rate": 0.00018695878262161498,
      "loss": 1.3095,
      "step": 45650
    },
    {
      "epoch": 1.3644642164034275,
      "grad_norm": 3.6063497066497803,
      "learning_rate": 0.00018694449682569615,
      "loss": 1.3088,
      "step": 45700
    },
    {
      "epoch": 1.3659570656555102,
      "grad_norm": 5.380884170532227,
      "learning_rate": 0.0001869302110297773,
      "loss": 1.2992,
      "step": 45750
    },
    {
      "epoch": 1.3674499149075927,
      "grad_norm": 3.108400821685791,
      "learning_rate": 0.00018691592523385848,
      "loss": 1.3638,
      "step": 45800
    },
    {
      "epoch": 1.3689427641596752,
      "grad_norm": 5.948930740356445,
      "learning_rate": 0.00018690163943793967,
      "loss": 1.4066,
      "step": 45850
    },
    {
      "epoch": 1.3704356134117577,
      "grad_norm": 3.5984768867492676,
      "learning_rate": 0.0001868873536420208,
      "loss": 1.3167,
      "step": 45900
    },
    {
      "epoch": 1.3719284626638402,
      "grad_norm": 4.595841884613037,
      "learning_rate": 0.000186873067846102,
      "loss": 1.2317,
      "step": 45950
    },
    {
      "epoch": 1.3734213119159226,
      "grad_norm": 4.499626159667969,
      "learning_rate": 0.00018685878205018314,
      "loss": 1.3094,
      "step": 46000
    },
    {
      "epoch": 1.3749141611680051,
      "grad_norm": 4.876091480255127,
      "learning_rate": 0.00018684449625426433,
      "loss": 1.3723,
      "step": 46050
    },
    {
      "epoch": 1.3764070104200878,
      "grad_norm": 5.354704856872559,
      "learning_rate": 0.00018683021045834547,
      "loss": 1.3672,
      "step": 46100
    },
    {
      "epoch": 1.3778998596721703,
      "grad_norm": 4.629252910614014,
      "learning_rate": 0.00018681592466242666,
      "loss": 1.3295,
      "step": 46150
    },
    {
      "epoch": 1.3793927089242528,
      "grad_norm": 3.6931405067443848,
      "learning_rate": 0.00018680163886650782,
      "loss": 1.2989,
      "step": 46200
    },
    {
      "epoch": 1.3808855581763353,
      "grad_norm": 6.378023624420166,
      "learning_rate": 0.00018678735307058899,
      "loss": 1.3022,
      "step": 46250
    },
    {
      "epoch": 1.3823784074284178,
      "grad_norm": 4.780524253845215,
      "learning_rate": 0.00018677306727467015,
      "loss": 1.298,
      "step": 46300
    },
    {
      "epoch": 1.3838712566805005,
      "grad_norm": 4.4194560050964355,
      "learning_rate": 0.00018675878147875132,
      "loss": 1.3396,
      "step": 46350
    },
    {
      "epoch": 1.385364105932583,
      "grad_norm": 5.250397205352783,
      "learning_rate": 0.00018674449568283248,
      "loss": 1.347,
      "step": 46400
    },
    {
      "epoch": 1.3868569551846655,
      "grad_norm": 5.2740607261657715,
      "learning_rate": 0.00018673020988691364,
      "loss": 1.3962,
      "step": 46450
    },
    {
      "epoch": 1.388349804436748,
      "grad_norm": 4.3260884284973145,
      "learning_rate": 0.0001867159240909948,
      "loss": 1.3236,
      "step": 46500
    },
    {
      "epoch": 1.3898426536888304,
      "grad_norm": 4.773256778717041,
      "learning_rate": 0.00018670163829507597,
      "loss": 1.324,
      "step": 46550
    },
    {
      "epoch": 1.3913355029409131,
      "grad_norm": 6.666245460510254,
      "learning_rate": 0.00018668735249915714,
      "loss": 1.3242,
      "step": 46600
    },
    {
      "epoch": 1.3928283521929956,
      "grad_norm": 7.051351547241211,
      "learning_rate": 0.00018667306670323833,
      "loss": 1.3228,
      "step": 46650
    },
    {
      "epoch": 1.394321201445078,
      "grad_norm": 7.103764057159424,
      "learning_rate": 0.00018665878090731947,
      "loss": 1.3815,
      "step": 46700
    },
    {
      "epoch": 1.3958140506971606,
      "grad_norm": 4.093582630157471,
      "learning_rate": 0.00018664449511140066,
      "loss": 1.3352,
      "step": 46750
    },
    {
      "epoch": 1.397306899949243,
      "grad_norm": 4.276589870452881,
      "learning_rate": 0.0001866302093154818,
      "loss": 1.2578,
      "step": 46800
    },
    {
      "epoch": 1.3987997492013258,
      "grad_norm": 4.325264930725098,
      "learning_rate": 0.000186615923519563,
      "loss": 1.297,
      "step": 46850
    },
    {
      "epoch": 1.4002925984534083,
      "grad_norm": 4.978733062744141,
      "learning_rate": 0.00018660163772364415,
      "loss": 1.3142,
      "step": 46900
    },
    {
      "epoch": 1.4017854477054907,
      "grad_norm": 4.2968854904174805,
      "learning_rate": 0.00018658735192772532,
      "loss": 1.3085,
      "step": 46950
    },
    {
      "epoch": 1.4032782969575732,
      "grad_norm": 3.587477922439575,
      "learning_rate": 0.00018657306613180648,
      "loss": 1.3348,
      "step": 47000
    },
    {
      "epoch": 1.4047711462096557,
      "grad_norm": 5.918453693389893,
      "learning_rate": 0.00018655878033588765,
      "loss": 1.3163,
      "step": 47050
    },
    {
      "epoch": 1.4062639954617382,
      "grad_norm": 4.217932224273682,
      "learning_rate": 0.0001865444945399688,
      "loss": 1.358,
      "step": 47100
    },
    {
      "epoch": 1.4077568447138207,
      "grad_norm": 4.987464904785156,
      "learning_rate": 0.00018653020874404998,
      "loss": 1.2865,
      "step": 47150
    },
    {
      "epoch": 1.4092496939659034,
      "grad_norm": 4.03493070602417,
      "learning_rate": 0.00018651592294813114,
      "loss": 1.2961,
      "step": 47200
    },
    {
      "epoch": 1.4107425432179859,
      "grad_norm": 4.768476486206055,
      "learning_rate": 0.0001865016371522123,
      "loss": 1.3301,
      "step": 47250
    },
    {
      "epoch": 1.4122353924700684,
      "grad_norm": 4.223608016967773,
      "learning_rate": 0.00018648735135629347,
      "loss": 1.358,
      "step": 47300
    },
    {
      "epoch": 1.4137282417221508,
      "grad_norm": 5.97542142868042,
      "learning_rate": 0.00018647306556037464,
      "loss": 1.3001,
      "step": 47350
    },
    {
      "epoch": 1.4152210909742333,
      "grad_norm": 3.4876928329467773,
      "learning_rate": 0.0001864587797644558,
      "loss": 1.358,
      "step": 47400
    },
    {
      "epoch": 1.416713940226316,
      "grad_norm": 3.8894033432006836,
      "learning_rate": 0.000186444493968537,
      "loss": 1.3079,
      "step": 47450
    },
    {
      "epoch": 1.4182067894783985,
      "grad_norm": 5.399030685424805,
      "learning_rate": 0.00018643020817261813,
      "loss": 1.3059,
      "step": 47500
    },
    {
      "epoch": 1.419699638730481,
      "grad_norm": 3.9021189212799072,
      "learning_rate": 0.00018641592237669932,
      "loss": 1.2862,
      "step": 47550
    },
    {
      "epoch": 1.4211924879825635,
      "grad_norm": 4.103232383728027,
      "learning_rate": 0.00018640163658078046,
      "loss": 1.3533,
      "step": 47600
    },
    {
      "epoch": 1.422685337234646,
      "grad_norm": 4.994762897491455,
      "learning_rate": 0.00018638735078486165,
      "loss": 1.3515,
      "step": 47650
    },
    {
      "epoch": 1.4241781864867287,
      "grad_norm": 4.363476276397705,
      "learning_rate": 0.00018637306498894282,
      "loss": 1.3524,
      "step": 47700
    },
    {
      "epoch": 1.4256710357388112,
      "grad_norm": 4.2915873527526855,
      "learning_rate": 0.00018635877919302398,
      "loss": 1.3245,
      "step": 47750
    },
    {
      "epoch": 1.4271638849908936,
      "grad_norm": 4.654321670532227,
      "learning_rate": 0.00018634449339710514,
      "loss": 1.369,
      "step": 47800
    },
    {
      "epoch": 1.4286567342429761,
      "grad_norm": 4.786460876464844,
      "learning_rate": 0.00018633020760118628,
      "loss": 1.3485,
      "step": 47850
    },
    {
      "epoch": 1.4301495834950586,
      "grad_norm": 4.434101104736328,
      "learning_rate": 0.00018631592180526747,
      "loss": 1.3131,
      "step": 47900
    },
    {
      "epoch": 1.4316424327471413,
      "grad_norm": 4.622028827667236,
      "learning_rate": 0.0001863016360093486,
      "loss": 1.3383,
      "step": 47950
    },
    {
      "epoch": 1.4331352819992238,
      "grad_norm": 5.579216480255127,
      "learning_rate": 0.0001862873502134298,
      "loss": 1.3214,
      "step": 48000
    },
    {
      "epoch": 1.4346281312513063,
      "grad_norm": 5.444978713989258,
      "learning_rate": 0.00018627306441751097,
      "loss": 1.3243,
      "step": 48050
    },
    {
      "epoch": 1.4361209805033888,
      "grad_norm": 4.104014873504639,
      "learning_rate": 0.00018625877862159213,
      "loss": 1.241,
      "step": 48100
    },
    {
      "epoch": 1.4376138297554713,
      "grad_norm": 4.517495632171631,
      "learning_rate": 0.0001862444928256733,
      "loss": 1.4197,
      "step": 48150
    },
    {
      "epoch": 1.4391066790075537,
      "grad_norm": 5.38925838470459,
      "learning_rate": 0.00018623020702975446,
      "loss": 1.3125,
      "step": 48200
    },
    {
      "epoch": 1.4405995282596362,
      "grad_norm": 6.211074352264404,
      "learning_rate": 0.00018621592123383563,
      "loss": 1.2659,
      "step": 48250
    },
    {
      "epoch": 1.442092377511719,
      "grad_norm": 5.461385726928711,
      "learning_rate": 0.0001862016354379168,
      "loss": 1.3043,
      "step": 48300
    },
    {
      "epoch": 1.4435852267638014,
      "grad_norm": 8.296045303344727,
      "learning_rate": 0.00018618734964199796,
      "loss": 1.3422,
      "step": 48350
    },
    {
      "epoch": 1.445078076015884,
      "grad_norm": 4.640946388244629,
      "learning_rate": 0.00018617306384607912,
      "loss": 1.3317,
      "step": 48400
    },
    {
      "epoch": 1.4465709252679664,
      "grad_norm": 4.698127269744873,
      "learning_rate": 0.00018615877805016029,
      "loss": 1.3241,
      "step": 48450
    },
    {
      "epoch": 1.4480637745200489,
      "grad_norm": 4.810885429382324,
      "learning_rate": 0.00018614449225424148,
      "loss": 1.3204,
      "step": 48500
    },
    {
      "epoch": 1.4495566237721316,
      "grad_norm": 3.57252836227417,
      "learning_rate": 0.00018613020645832261,
      "loss": 1.3412,
      "step": 48550
    },
    {
      "epoch": 1.451049473024214,
      "grad_norm": 4.79307222366333,
      "learning_rate": 0.0001861159206624038,
      "loss": 1.325,
      "step": 48600
    },
    {
      "epoch": 1.4525423222762965,
      "grad_norm": 5.495840072631836,
      "learning_rate": 0.00018610163486648494,
      "loss": 1.3191,
      "step": 48650
    },
    {
      "epoch": 1.454035171528379,
      "grad_norm": 4.352155685424805,
      "learning_rate": 0.00018608734907056614,
      "loss": 1.3129,
      "step": 48700
    },
    {
      "epoch": 1.4555280207804615,
      "grad_norm": 4.606874942779541,
      "learning_rate": 0.00018607306327464727,
      "loss": 1.3212,
      "step": 48750
    },
    {
      "epoch": 1.4570208700325442,
      "grad_norm": 5.2679219245910645,
      "learning_rate": 0.00018605877747872846,
      "loss": 1.3531,
      "step": 48800
    },
    {
      "epoch": 1.4585137192846267,
      "grad_norm": 4.821314811706543,
      "learning_rate": 0.00018604449168280963,
      "loss": 1.3195,
      "step": 48850
    },
    {
      "epoch": 1.4600065685367092,
      "grad_norm": 4.137834072113037,
      "learning_rate": 0.0001860302058868908,
      "loss": 1.2751,
      "step": 48900
    },
    {
      "epoch": 1.4614994177887917,
      "grad_norm": 6.577256202697754,
      "learning_rate": 0.00018601592009097196,
      "loss": 1.3971,
      "step": 48950
    },
    {
      "epoch": 1.4629922670408742,
      "grad_norm": 4.749688148498535,
      "learning_rate": 0.00018600163429505312,
      "loss": 1.3557,
      "step": 49000
    },
    {
      "epoch": 1.4644851162929569,
      "grad_norm": 5.410492897033691,
      "learning_rate": 0.0001859873484991343,
      "loss": 1.2467,
      "step": 49050
    },
    {
      "epoch": 1.4659779655450393,
      "grad_norm": 4.303316593170166,
      "learning_rate": 0.00018597306270321545,
      "loss": 1.3166,
      "step": 49100
    },
    {
      "epoch": 1.4674708147971218,
      "grad_norm": 5.683919906616211,
      "learning_rate": 0.00018595877690729662,
      "loss": 1.2932,
      "step": 49150
    },
    {
      "epoch": 1.4689636640492043,
      "grad_norm": 5.251864433288574,
      "learning_rate": 0.00018594449111137778,
      "loss": 1.3179,
      "step": 49200
    },
    {
      "epoch": 1.4704565133012868,
      "grad_norm": 4.565097808837891,
      "learning_rate": 0.00018593020531545895,
      "loss": 1.327,
      "step": 49250
    },
    {
      "epoch": 1.4719493625533693,
      "grad_norm": 5.532310485839844,
      "learning_rate": 0.00018591591951954014,
      "loss": 1.3868,
      "step": 49300
    },
    {
      "epoch": 1.4734422118054518,
      "grad_norm": 4.5699782371521,
      "learning_rate": 0.00018590163372362128,
      "loss": 1.2583,
      "step": 49350
    },
    {
      "epoch": 1.4749350610575345,
      "grad_norm": 4.792230606079102,
      "learning_rate": 0.00018588734792770247,
      "loss": 1.3805,
      "step": 49400
    },
    {
      "epoch": 1.476427910309617,
      "grad_norm": 4.52764368057251,
      "learning_rate": 0.0001858730621317836,
      "loss": 1.3465,
      "step": 49450
    },
    {
      "epoch": 1.4779207595616994,
      "grad_norm": 5.21061372756958,
      "learning_rate": 0.0001858587763358648,
      "loss": 1.3778,
      "step": 49500
    },
    {
      "epoch": 1.479413608813782,
      "grad_norm": 4.855043888092041,
      "learning_rate": 0.00018584449053994593,
      "loss": 1.3313,
      "step": 49550
    },
    {
      "epoch": 1.4809064580658644,
      "grad_norm": 4.446802616119385,
      "learning_rate": 0.00018583020474402713,
      "loss": 1.3224,
      "step": 49600
    },
    {
      "epoch": 1.4823993073179471,
      "grad_norm": 5.054227828979492,
      "learning_rate": 0.0001858159189481083,
      "loss": 1.3782,
      "step": 49650
    },
    {
      "epoch": 1.4838921565700296,
      "grad_norm": 6.877427101135254,
      "learning_rate": 0.00018580163315218946,
      "loss": 1.3399,
      "step": 49700
    },
    {
      "epoch": 1.485385005822112,
      "grad_norm": 6.227908611297607,
      "learning_rate": 0.00018578734735627062,
      "loss": 1.2996,
      "step": 49750
    },
    {
      "epoch": 1.4868778550741946,
      "grad_norm": 6.2445759773254395,
      "learning_rate": 0.00018577306156035179,
      "loss": 1.3005,
      "step": 49800
    },
    {
      "epoch": 1.488370704326277,
      "grad_norm": 6.488719463348389,
      "learning_rate": 0.00018575877576443295,
      "loss": 1.3755,
      "step": 49850
    },
    {
      "epoch": 1.4898635535783598,
      "grad_norm": 5.3022589683532715,
      "learning_rate": 0.00018574448996851411,
      "loss": 1.3402,
      "step": 49900
    },
    {
      "epoch": 1.4913564028304422,
      "grad_norm": 4.986827373504639,
      "learning_rate": 0.00018573020417259528,
      "loss": 1.3695,
      "step": 49950
    },
    {
      "epoch": 1.4928492520825247,
      "grad_norm": 5.669780731201172,
      "learning_rate": 0.00018571591837667644,
      "loss": 1.314,
      "step": 50000
    },
    {
      "epoch": 1.4943421013346072,
      "grad_norm": 5.21511697769165,
      "learning_rate": 0.0001857016325807576,
      "loss": 1.363,
      "step": 50050
    },
    {
      "epoch": 1.4958349505866897,
      "grad_norm": 4.303940296173096,
      "learning_rate": 0.0001856873467848388,
      "loss": 1.2926,
      "step": 50100
    },
    {
      "epoch": 1.4973277998387724,
      "grad_norm": 4.7364020347595215,
      "learning_rate": 0.00018567306098891994,
      "loss": 1.3325,
      "step": 50150
    },
    {
      "epoch": 1.4988206490908549,
      "grad_norm": 4.325379371643066,
      "learning_rate": 0.00018565877519300113,
      "loss": 1.31,
      "step": 50200
    },
    {
      "epoch": 1.5003134983429374,
      "grad_norm": 5.442165374755859,
      "learning_rate": 0.00018564448939708227,
      "loss": 1.3445,
      "step": 50250
    },
    {
      "epoch": 1.5018063475950199,
      "grad_norm": 5.861702919006348,
      "learning_rate": 0.00018563020360116346,
      "loss": 1.3392,
      "step": 50300
    },
    {
      "epoch": 1.5032991968471023,
      "grad_norm": 4.277650356292725,
      "learning_rate": 0.00018561591780524462,
      "loss": 1.3127,
      "step": 50350
    },
    {
      "epoch": 1.504792046099185,
      "grad_norm": 5.670056343078613,
      "learning_rate": 0.00018560163200932576,
      "loss": 1.346,
      "step": 50400
    },
    {
      "epoch": 1.5062848953512673,
      "grad_norm": 4.479572772979736,
      "learning_rate": 0.00018558734621340695,
      "loss": 1.3254,
      "step": 50450
    },
    {
      "epoch": 1.50777774460335,
      "grad_norm": 4.820215225219727,
      "learning_rate": 0.0001855730604174881,
      "loss": 1.3043,
      "step": 50500
    },
    {
      "epoch": 1.5092705938554325,
      "grad_norm": 5.21779203414917,
      "learning_rate": 0.00018555877462156928,
      "loss": 1.2953,
      "step": 50550
    },
    {
      "epoch": 1.510763443107515,
      "grad_norm": 4.987244129180908,
      "learning_rate": 0.00018554448882565042,
      "loss": 1.3008,
      "step": 50600
    },
    {
      "epoch": 1.5122562923595977,
      "grad_norm": 4.658651828765869,
      "learning_rate": 0.0001855302030297316,
      "loss": 1.3146,
      "step": 50650
    },
    {
      "epoch": 1.51374914161168,
      "grad_norm": 4.724532604217529,
      "learning_rate": 0.00018551591723381278,
      "loss": 1.3467,
      "step": 50700
    },
    {
      "epoch": 1.5152419908637627,
      "grad_norm": 6.199383735656738,
      "learning_rate": 0.00018550163143789394,
      "loss": 1.3244,
      "step": 50750
    },
    {
      "epoch": 1.5167348401158451,
      "grad_norm": 4.255542278289795,
      "learning_rate": 0.0001854873456419751,
      "loss": 1.3373,
      "step": 50800
    },
    {
      "epoch": 1.5182276893679276,
      "grad_norm": 5.011137962341309,
      "learning_rate": 0.00018547305984605627,
      "loss": 1.2945,
      "step": 50850
    },
    {
      "epoch": 1.51972053862001,
      "grad_norm": 4.276299476623535,
      "learning_rate": 0.00018545877405013743,
      "loss": 1.3503,
      "step": 50900
    },
    {
      "epoch": 1.5212133878720926,
      "grad_norm": 5.867751598358154,
      "learning_rate": 0.0001854444882542186,
      "loss": 1.345,
      "step": 50950
    },
    {
      "epoch": 1.5227062371241753,
      "grad_norm": 3.6653764247894287,
      "learning_rate": 0.00018543020245829976,
      "loss": 1.3353,
      "step": 51000
    },
    {
      "epoch": 1.5241990863762576,
      "grad_norm": 4.758911609649658,
      "learning_rate": 0.00018541591666238093,
      "loss": 1.3488,
      "step": 51050
    },
    {
      "epoch": 1.5256919356283403,
      "grad_norm": 4.028106689453125,
      "learning_rate": 0.0001854016308664621,
      "loss": 1.3215,
      "step": 51100
    },
    {
      "epoch": 1.5271847848804228,
      "grad_norm": 4.353278636932373,
      "learning_rate": 0.00018538734507054329,
      "loss": 1.2984,
      "step": 51150
    },
    {
      "epoch": 1.5286776341325052,
      "grad_norm": 7.112466812133789,
      "learning_rate": 0.00018537305927462442,
      "loss": 1.3057,
      "step": 51200
    },
    {
      "epoch": 1.530170483384588,
      "grad_norm": 4.703035354614258,
      "learning_rate": 0.00018535877347870561,
      "loss": 1.3289,
      "step": 51250
    },
    {
      "epoch": 1.5316633326366702,
      "grad_norm": 4.110894203186035,
      "learning_rate": 0.00018534448768278675,
      "loss": 1.3587,
      "step": 51300
    },
    {
      "epoch": 1.533156181888753,
      "grad_norm": 5.426900863647461,
      "learning_rate": 0.00018533020188686794,
      "loss": 1.316,
      "step": 51350
    },
    {
      "epoch": 1.5346490311408354,
      "grad_norm": 4.927971839904785,
      "learning_rate": 0.00018531591609094908,
      "loss": 1.3775,
      "step": 51400
    },
    {
      "epoch": 1.5361418803929179,
      "grad_norm": 4.867948055267334,
      "learning_rate": 0.00018530163029503027,
      "loss": 1.3038,
      "step": 51450
    },
    {
      "epoch": 1.5376347296450006,
      "grad_norm": 4.8490471839904785,
      "learning_rate": 0.00018528734449911144,
      "loss": 1.3851,
      "step": 51500
    },
    {
      "epoch": 1.5391275788970828,
      "grad_norm": 4.634105682373047,
      "learning_rate": 0.0001852730587031926,
      "loss": 1.3215,
      "step": 51550
    },
    {
      "epoch": 1.5406204281491656,
      "grad_norm": 3.7799713611602783,
      "learning_rate": 0.00018525877290727377,
      "loss": 1.3371,
      "step": 51600
    },
    {
      "epoch": 1.542113277401248,
      "grad_norm": 3.9134316444396973,
      "learning_rate": 0.00018524448711135493,
      "loss": 1.3073,
      "step": 51650
    },
    {
      "epoch": 1.5436061266533305,
      "grad_norm": 3.5114309787750244,
      "learning_rate": 0.0001852302013154361,
      "loss": 1.3859,
      "step": 51700
    },
    {
      "epoch": 1.5450989759054132,
      "grad_norm": 3.7832348346710205,
      "learning_rate": 0.00018521591551951726,
      "loss": 1.369,
      "step": 51750
    },
    {
      "epoch": 1.5465918251574955,
      "grad_norm": 3.666222333908081,
      "learning_rate": 0.00018520162972359843,
      "loss": 1.3552,
      "step": 51800
    },
    {
      "epoch": 1.5480846744095782,
      "grad_norm": 4.288083076477051,
      "learning_rate": 0.0001851873439276796,
      "loss": 1.3139,
      "step": 51850
    },
    {
      "epoch": 1.5495775236616607,
      "grad_norm": 6.1794939041137695,
      "learning_rate": 0.00018517305813176075,
      "loss": 1.2984,
      "step": 51900
    },
    {
      "epoch": 1.5510703729137432,
      "grad_norm": 4.89127779006958,
      "learning_rate": 0.00018515877233584195,
      "loss": 1.3251,
      "step": 51950
    },
    {
      "epoch": 1.5525632221658257,
      "grad_norm": 3.1737000942230225,
      "learning_rate": 0.00018514448653992308,
      "loss": 1.342,
      "step": 52000
    },
    {
      "epoch": 1.5540560714179081,
      "grad_norm": 5.618998050689697,
      "learning_rate": 0.00018513020074400428,
      "loss": 1.3234,
      "step": 52050
    },
    {
      "epoch": 1.5555489206699908,
      "grad_norm": 4.3047685623168945,
      "learning_rate": 0.0001851159149480854,
      "loss": 1.3305,
      "step": 52100
    },
    {
      "epoch": 1.557041769922073,
      "grad_norm": 5.245848178863525,
      "learning_rate": 0.0001851016291521666,
      "loss": 1.3052,
      "step": 52150
    },
    {
      "epoch": 1.5585346191741558,
      "grad_norm": 5.337231636047363,
      "learning_rate": 0.00018508734335624774,
      "loss": 1.3353,
      "step": 52200
    },
    {
      "epoch": 1.5600274684262383,
      "grad_norm": 4.924198627471924,
      "learning_rate": 0.00018507305756032893,
      "loss": 1.3022,
      "step": 52250
    },
    {
      "epoch": 1.5615203176783208,
      "grad_norm": 4.464673042297363,
      "learning_rate": 0.0001850587717644101,
      "loss": 1.3168,
      "step": 52300
    },
    {
      "epoch": 1.5630131669304035,
      "grad_norm": 3.7909131050109863,
      "learning_rate": 0.00018504448596849126,
      "loss": 1.3307,
      "step": 52350
    },
    {
      "epoch": 1.5645060161824857,
      "grad_norm": 4.331950664520264,
      "learning_rate": 0.00018503020017257243,
      "loss": 1.3574,
      "step": 52400
    },
    {
      "epoch": 1.5659988654345685,
      "grad_norm": 4.650835990905762,
      "learning_rate": 0.0001850159143766536,
      "loss": 1.3356,
      "step": 52450
    },
    {
      "epoch": 1.567491714686651,
      "grad_norm": 5.968230247497559,
      "learning_rate": 0.00018500162858073476,
      "loss": 1.3515,
      "step": 52500
    },
    {
      "epoch": 1.5689845639387334,
      "grad_norm": 3.9560632705688477,
      "learning_rate": 0.00018498734278481592,
      "loss": 1.3262,
      "step": 52550
    },
    {
      "epoch": 1.5704774131908161,
      "grad_norm": 3.6332058906555176,
      "learning_rate": 0.0001849730569888971,
      "loss": 1.3708,
      "step": 52600
    },
    {
      "epoch": 1.5719702624428984,
      "grad_norm": 5.171941757202148,
      "learning_rate": 0.00018495877119297825,
      "loss": 1.3201,
      "step": 52650
    },
    {
      "epoch": 1.573463111694981,
      "grad_norm": 4.531052589416504,
      "learning_rate": 0.00018494448539705942,
      "loss": 1.364,
      "step": 52700
    },
    {
      "epoch": 1.5749559609470636,
      "grad_norm": 3.860273838043213,
      "learning_rate": 0.0001849301996011406,
      "loss": 1.3729,
      "step": 52750
    },
    {
      "epoch": 1.576448810199146,
      "grad_norm": 4.562480926513672,
      "learning_rate": 0.00018491591380522175,
      "loss": 1.2745,
      "step": 52800
    },
    {
      "epoch": 1.5779416594512288,
      "grad_norm": 4.759164333343506,
      "learning_rate": 0.00018490162800930294,
      "loss": 1.3399,
      "step": 52850
    },
    {
      "epoch": 1.579434508703311,
      "grad_norm": 6.885429859161377,
      "learning_rate": 0.00018488734221338408,
      "loss": 1.2915,
      "step": 52900
    },
    {
      "epoch": 1.5809273579553937,
      "grad_norm": 4.555217742919922,
      "learning_rate": 0.00018487305641746527,
      "loss": 1.3897,
      "step": 52950
    },
    {
      "epoch": 1.5824202072074762,
      "grad_norm": 5.686909198760986,
      "learning_rate": 0.0001848587706215464,
      "loss": 1.3234,
      "step": 53000
    },
    {
      "epoch": 1.5839130564595587,
      "grad_norm": 3.994685411453247,
      "learning_rate": 0.00018484448482562757,
      "loss": 1.3265,
      "step": 53050
    },
    {
      "epoch": 1.5854059057116412,
      "grad_norm": 4.623023986816406,
      "learning_rate": 0.00018483019902970876,
      "loss": 1.3479,
      "step": 53100
    },
    {
      "epoch": 1.5868987549637237,
      "grad_norm": 6.000095844268799,
      "learning_rate": 0.0001848159132337899,
      "loss": 1.3333,
      "step": 53150
    },
    {
      "epoch": 1.5883916042158064,
      "grad_norm": 7.388947010040283,
      "learning_rate": 0.0001848016274378711,
      "loss": 1.306,
      "step": 53200
    },
    {
      "epoch": 1.5898844534678886,
      "grad_norm": 4.04311990737915,
      "learning_rate": 0.00018478734164195223,
      "loss": 1.3263,
      "step": 53250
    },
    {
      "epoch": 1.5913773027199714,
      "grad_norm": 5.370228290557861,
      "learning_rate": 0.00018477305584603342,
      "loss": 1.3279,
      "step": 53300
    },
    {
      "epoch": 1.5928701519720538,
      "grad_norm": 5.453476905822754,
      "learning_rate": 0.00018475877005011458,
      "loss": 1.3477,
      "step": 53350
    },
    {
      "epoch": 1.5943630012241363,
      "grad_norm": 3.9221408367156982,
      "learning_rate": 0.00018474448425419575,
      "loss": 1.3391,
      "step": 53400
    },
    {
      "epoch": 1.595855850476219,
      "grad_norm": 4.4245781898498535,
      "learning_rate": 0.0001847301984582769,
      "loss": 1.3263,
      "step": 53450
    },
    {
      "epoch": 1.5973486997283013,
      "grad_norm": 4.352360725402832,
      "learning_rate": 0.00018471591266235808,
      "loss": 1.3059,
      "step": 53500
    },
    {
      "epoch": 1.598841548980384,
      "grad_norm": 5.2914509773254395,
      "learning_rate": 0.00018470162686643924,
      "loss": 1.3353,
      "step": 53550
    },
    {
      "epoch": 1.6003343982324665,
      "grad_norm": 5.199918746948242,
      "learning_rate": 0.0001846873410705204,
      "loss": 1.3667,
      "step": 53600
    },
    {
      "epoch": 1.601827247484549,
      "grad_norm": 4.404350280761719,
      "learning_rate": 0.00018467305527460157,
      "loss": 1.3245,
      "step": 53650
    },
    {
      "epoch": 1.6033200967366317,
      "grad_norm": 4.452194690704346,
      "learning_rate": 0.00018465876947868274,
      "loss": 1.3194,
      "step": 53700
    },
    {
      "epoch": 1.604812945988714,
      "grad_norm": 4.387017726898193,
      "learning_rate": 0.0001846444836827639,
      "loss": 1.2448,
      "step": 53750
    },
    {
      "epoch": 1.6063057952407966,
      "grad_norm": 5.375933647155762,
      "learning_rate": 0.0001846301978868451,
      "loss": 1.3515,
      "step": 53800
    },
    {
      "epoch": 1.6077986444928791,
      "grad_norm": 4.564199447631836,
      "learning_rate": 0.00018461591209092623,
      "loss": 1.3182,
      "step": 53850
    },
    {
      "epoch": 1.6092914937449616,
      "grad_norm": 5.435900688171387,
      "learning_rate": 0.00018460162629500742,
      "loss": 1.3071,
      "step": 53900
    },
    {
      "epoch": 1.6107843429970443,
      "grad_norm": 3.600592613220215,
      "learning_rate": 0.00018458734049908856,
      "loss": 1.4009,
      "step": 53950
    },
    {
      "epoch": 1.6122771922491266,
      "grad_norm": 5.12654447555542,
      "learning_rate": 0.00018457305470316975,
      "loss": 1.3285,
      "step": 54000
    },
    {
      "epoch": 1.6137700415012093,
      "grad_norm": 3.6472108364105225,
      "learning_rate": 0.0001845587689072509,
      "loss": 1.32,
      "step": 54050
    },
    {
      "epoch": 1.6152628907532918,
      "grad_norm": 5.0967559814453125,
      "learning_rate": 0.00018454448311133208,
      "loss": 1.3018,
      "step": 54100
    },
    {
      "epoch": 1.6167557400053743,
      "grad_norm": 5.078010082244873,
      "learning_rate": 0.00018453019731541325,
      "loss": 1.3539,
      "step": 54150
    },
    {
      "epoch": 1.6182485892574567,
      "grad_norm": 4.601762294769287,
      "learning_rate": 0.0001845159115194944,
      "loss": 1.3153,
      "step": 54200
    },
    {
      "epoch": 1.6197414385095392,
      "grad_norm": 3.8307271003723145,
      "learning_rate": 0.00018450162572357557,
      "loss": 1.34,
      "step": 54250
    },
    {
      "epoch": 1.621234287761622,
      "grad_norm": 6.145284175872803,
      "learning_rate": 0.00018448733992765674,
      "loss": 1.3315,
      "step": 54300
    },
    {
      "epoch": 1.6227271370137042,
      "grad_norm": 9.96362590789795,
      "learning_rate": 0.0001844730541317379,
      "loss": 1.3269,
      "step": 54350
    },
    {
      "epoch": 1.624219986265787,
      "grad_norm": 5.126537322998047,
      "learning_rate": 0.00018445876833581907,
      "loss": 1.3604,
      "step": 54400
    },
    {
      "epoch": 1.6257128355178694,
      "grad_norm": 3.604433059692383,
      "learning_rate": 0.00018444448253990023,
      "loss": 1.3651,
      "step": 54450
    },
    {
      "epoch": 1.6272056847699519,
      "grad_norm": 4.835113048553467,
      "learning_rate": 0.0001844301967439814,
      "loss": 1.2787,
      "step": 54500
    },
    {
      "epoch": 1.6286985340220346,
      "grad_norm": 4.534168243408203,
      "learning_rate": 0.00018441591094806256,
      "loss": 1.3354,
      "step": 54550
    },
    {
      "epoch": 1.6301913832741168,
      "grad_norm": 7.929208755493164,
      "learning_rate": 0.00018440162515214375,
      "loss": 1.3361,
      "step": 54600
    },
    {
      "epoch": 1.6316842325261995,
      "grad_norm": 4.582381248474121,
      "learning_rate": 0.0001843873393562249,
      "loss": 1.354,
      "step": 54650
    },
    {
      "epoch": 1.633177081778282,
      "grad_norm": 4.781309604644775,
      "learning_rate": 0.00018437305356030608,
      "loss": 1.3035,
      "step": 54700
    },
    {
      "epoch": 1.6346699310303645,
      "grad_norm": 6.0308308601379395,
      "learning_rate": 0.00018435876776438722,
      "loss": 1.3392,
      "step": 54750
    },
    {
      "epoch": 1.6361627802824472,
      "grad_norm": 5.485566139221191,
      "learning_rate": 0.0001843444819684684,
      "loss": 1.3408,
      "step": 54800
    },
    {
      "epoch": 1.6376556295345295,
      "grad_norm": 6.658644676208496,
      "learning_rate": 0.00018433019617254955,
      "loss": 1.3772,
      "step": 54850
    },
    {
      "epoch": 1.6391484787866122,
      "grad_norm": 4.54780912399292,
      "learning_rate": 0.00018431591037663074,
      "loss": 1.3511,
      "step": 54900
    },
    {
      "epoch": 1.6406413280386947,
      "grad_norm": 5.600758075714111,
      "learning_rate": 0.0001843016245807119,
      "loss": 1.3301,
      "step": 54950
    },
    {
      "epoch": 1.6421341772907772,
      "grad_norm": 4.194091796875,
      "learning_rate": 0.00018428733878479307,
      "loss": 1.3159,
      "step": 55000
    },
    {
      "epoch": 1.6436270265428599,
      "grad_norm": 6.306137561798096,
      "learning_rate": 0.00018427305298887424,
      "loss": 1.3401,
      "step": 55050
    },
    {
      "epoch": 1.6451198757949421,
      "grad_norm": 4.377103328704834,
      "learning_rate": 0.0001842587671929554,
      "loss": 1.34,
      "step": 55100
    },
    {
      "epoch": 1.6466127250470248,
      "grad_norm": 3.450533390045166,
      "learning_rate": 0.00018424448139703657,
      "loss": 1.3367,
      "step": 55150
    },
    {
      "epoch": 1.6481055742991073,
      "grad_norm": 3.5845348834991455,
      "learning_rate": 0.00018423019560111773,
      "loss": 1.3755,
      "step": 55200
    },
    {
      "epoch": 1.6495984235511898,
      "grad_norm": 4.741848468780518,
      "learning_rate": 0.0001842159098051989,
      "loss": 1.3116,
      "step": 55250
    },
    {
      "epoch": 1.6510912728032723,
      "grad_norm": 4.1155619621276855,
      "learning_rate": 0.00018420162400928006,
      "loss": 1.3682,
      "step": 55300
    },
    {
      "epoch": 1.6525841220553548,
      "grad_norm": 4.3722028732299805,
      "learning_rate": 0.00018418733821336122,
      "loss": 1.2916,
      "step": 55350
    },
    {
      "epoch": 1.6540769713074375,
      "grad_norm": 4.6453118324279785,
      "learning_rate": 0.00018417305241744242,
      "loss": 1.3743,
      "step": 55400
    },
    {
      "epoch": 1.6555698205595197,
      "grad_norm": 4.749643802642822,
      "learning_rate": 0.00018415876662152355,
      "loss": 1.3649,
      "step": 55450
    },
    {
      "epoch": 1.6570626698116024,
      "grad_norm": 3.9023261070251465,
      "learning_rate": 0.00018414448082560475,
      "loss": 1.4002,
      "step": 55500
    },
    {
      "epoch": 1.658555519063685,
      "grad_norm": 4.34132719039917,
      "learning_rate": 0.00018413019502968588,
      "loss": 1.3344,
      "step": 55550
    },
    {
      "epoch": 1.6600483683157674,
      "grad_norm": 5.479841709136963,
      "learning_rate": 0.00018411590923376707,
      "loss": 1.3511,
      "step": 55600
    },
    {
      "epoch": 1.6615412175678501,
      "grad_norm": 3.340822696685791,
      "learning_rate": 0.0001841016234378482,
      "loss": 1.3789,
      "step": 55650
    },
    {
      "epoch": 1.6630340668199324,
      "grad_norm": 5.259148597717285,
      "learning_rate": 0.00018408733764192938,
      "loss": 1.3092,
      "step": 55700
    },
    {
      "epoch": 1.664526916072015,
      "grad_norm": 3.590052843093872,
      "learning_rate": 0.00018407305184601057,
      "loss": 1.3596,
      "step": 55750
    },
    {
      "epoch": 1.6660197653240976,
      "grad_norm": 5.688297748565674,
      "learning_rate": 0.0001840587660500917,
      "loss": 1.3629,
      "step": 55800
    },
    {
      "epoch": 1.66751261457618,
      "grad_norm": 5.826660633087158,
      "learning_rate": 0.0001840444802541729,
      "loss": 1.3757,
      "step": 55850
    },
    {
      "epoch": 1.6690054638282628,
      "grad_norm": 3.0926566123962402,
      "learning_rate": 0.00018403019445825404,
      "loss": 1.3507,
      "step": 55900
    },
    {
      "epoch": 1.670498313080345,
      "grad_norm": 7.726663589477539,
      "learning_rate": 0.00018401590866233523,
      "loss": 1.3162,
      "step": 55950
    },
    {
      "epoch": 1.6719911623324277,
      "grad_norm": 4.687432765960693,
      "learning_rate": 0.0001840016228664164,
      "loss": 1.289,
      "step": 56000
    },
    {
      "epoch": 1.6734840115845102,
      "grad_norm": 4.279799461364746,
      "learning_rate": 0.00018398733707049756,
      "loss": 1.3438,
      "step": 56050
    },
    {
      "epoch": 1.6749768608365927,
      "grad_norm": 6.4994797706604,
      "learning_rate": 0.00018397305127457872,
      "loss": 1.3373,
      "step": 56100
    },
    {
      "epoch": 1.6764697100886754,
      "grad_norm": 4.511581897735596,
      "learning_rate": 0.00018395876547865989,
      "loss": 1.3549,
      "step": 56150
    },
    {
      "epoch": 1.6779625593407577,
      "grad_norm": 5.1127190589904785,
      "learning_rate": 0.00018394447968274105,
      "loss": 1.2734,
      "step": 56200
    },
    {
      "epoch": 1.6794554085928404,
      "grad_norm": 4.7369561195373535,
      "learning_rate": 0.00018393019388682222,
      "loss": 1.3193,
      "step": 56250
    },
    {
      "epoch": 1.6809482578449229,
      "grad_norm": 5.101438522338867,
      "learning_rate": 0.00018391590809090338,
      "loss": 1.3311,
      "step": 56300
    },
    {
      "epoch": 1.6824411070970053,
      "grad_norm": 5.68798303604126,
      "learning_rate": 0.00018390162229498454,
      "loss": 1.3723,
      "step": 56350
    },
    {
      "epoch": 1.6839339563490878,
      "grad_norm": 5.620144367218018,
      "learning_rate": 0.0001838873364990657,
      "loss": 1.3051,
      "step": 56400
    },
    {
      "epoch": 1.6854268056011703,
      "grad_norm": 4.335910797119141,
      "learning_rate": 0.00018387305070314687,
      "loss": 1.2625,
      "step": 56450
    },
    {
      "epoch": 1.686919654853253,
      "grad_norm": 5.112996578216553,
      "learning_rate": 0.00018385876490722804,
      "loss": 1.3288,
      "step": 56500
    },
    {
      "epoch": 1.6884125041053353,
      "grad_norm": 5.485559463500977,
      "learning_rate": 0.00018384447911130923,
      "loss": 1.3171,
      "step": 56550
    },
    {
      "epoch": 1.689905353357418,
      "grad_norm": 5.58554744720459,
      "learning_rate": 0.00018383019331539037,
      "loss": 1.3002,
      "step": 56600
    },
    {
      "epoch": 1.6913982026095005,
      "grad_norm": 4.202605247497559,
      "learning_rate": 0.00018381590751947156,
      "loss": 1.3018,
      "step": 56650
    },
    {
      "epoch": 1.692891051861583,
      "grad_norm": 5.206381797790527,
      "learning_rate": 0.0001838016217235527,
      "loss": 1.3696,
      "step": 56700
    },
    {
      "epoch": 1.6943839011136657,
      "grad_norm": 5.87765645980835,
      "learning_rate": 0.0001837873359276339,
      "loss": 1.3616,
      "step": 56750
    },
    {
      "epoch": 1.695876750365748,
      "grad_norm": 3.4311394691467285,
      "learning_rate": 0.00018377305013171505,
      "loss": 1.3249,
      "step": 56800
    },
    {
      "epoch": 1.6973695996178306,
      "grad_norm": 2.7731411457061768,
      "learning_rate": 0.00018375876433579622,
      "loss": 1.3715,
      "step": 56850
    },
    {
      "epoch": 1.698862448869913,
      "grad_norm": 6.399103164672852,
      "learning_rate": 0.00018374447853987738,
      "loss": 1.3654,
      "step": 56900
    },
    {
      "epoch": 1.7003552981219956,
      "grad_norm": 3.8202717304229736,
      "learning_rate": 0.00018373019274395855,
      "loss": 1.3565,
      "step": 56950
    },
    {
      "epoch": 1.7018481473740783,
      "grad_norm": 4.040468215942383,
      "learning_rate": 0.0001837159069480397,
      "loss": 1.3373,
      "step": 57000
    },
    {
      "epoch": 1.7033409966261606,
      "grad_norm": 4.522890090942383,
      "learning_rate": 0.00018370162115212088,
      "loss": 1.3393,
      "step": 57050
    },
    {
      "epoch": 1.7048338458782433,
      "grad_norm": 4.771418571472168,
      "learning_rate": 0.00018368733535620204,
      "loss": 1.2437,
      "step": 57100
    },
    {
      "epoch": 1.7063266951303258,
      "grad_norm": 4.333567142486572,
      "learning_rate": 0.0001836730495602832,
      "loss": 1.2876,
      "step": 57150
    },
    {
      "epoch": 1.7078195443824082,
      "grad_norm": 5.1498703956604,
      "learning_rate": 0.00018365876376436437,
      "loss": 1.4321,
      "step": 57200
    },
    {
      "epoch": 1.709312393634491,
      "grad_norm": 3.8978734016418457,
      "learning_rate": 0.00018364447796844556,
      "loss": 1.2815,
      "step": 57250
    },
    {
      "epoch": 1.7108052428865732,
      "grad_norm": 5.365299224853516,
      "learning_rate": 0.0001836301921725267,
      "loss": 1.2443,
      "step": 57300
    },
    {
      "epoch": 1.712298092138656,
      "grad_norm": 4.474836826324463,
      "learning_rate": 0.0001836159063766079,
      "loss": 1.3634,
      "step": 57350
    },
    {
      "epoch": 1.7137909413907384,
      "grad_norm": 3.8736836910247803,
      "learning_rate": 0.00018360162058068903,
      "loss": 1.3889,
      "step": 57400
    },
    {
      "epoch": 1.7152837906428209,
      "grad_norm": 4.919806957244873,
      "learning_rate": 0.00018358733478477022,
      "loss": 1.3216,
      "step": 57450
    },
    {
      "epoch": 1.7167766398949034,
      "grad_norm": 6.2346954345703125,
      "learning_rate": 0.00018357304898885136,
      "loss": 1.3294,
      "step": 57500
    },
    {
      "epoch": 1.7182694891469859,
      "grad_norm": 4.782268524169922,
      "learning_rate": 0.00018355876319293255,
      "loss": 1.363,
      "step": 57550
    },
    {
      "epoch": 1.7197623383990686,
      "grad_norm": 5.770125389099121,
      "learning_rate": 0.00018354447739701372,
      "loss": 1.3441,
      "step": 57600
    },
    {
      "epoch": 1.721255187651151,
      "grad_norm": 5.742440223693848,
      "learning_rate": 0.00018353019160109488,
      "loss": 1.302,
      "step": 57650
    },
    {
      "epoch": 1.7227480369032335,
      "grad_norm": 5.430489540100098,
      "learning_rate": 0.00018351590580517604,
      "loss": 1.3183,
      "step": 57700
    },
    {
      "epoch": 1.724240886155316,
      "grad_norm": 6.17573356628418,
      "learning_rate": 0.0001835016200092572,
      "loss": 1.3421,
      "step": 57750
    },
    {
      "epoch": 1.7257337354073985,
      "grad_norm": 5.362539291381836,
      "learning_rate": 0.00018348733421333837,
      "loss": 1.3333,
      "step": 57800
    },
    {
      "epoch": 1.7272265846594812,
      "grad_norm": 4.4574294090271,
      "learning_rate": 0.00018347304841741954,
      "loss": 1.2792,
      "step": 57850
    },
    {
      "epoch": 1.7287194339115635,
      "grad_norm": 4.277164936065674,
      "learning_rate": 0.0001834587626215007,
      "loss": 1.342,
      "step": 57900
    },
    {
      "epoch": 1.7302122831636462,
      "grad_norm": 4.775753498077393,
      "learning_rate": 0.00018344447682558187,
      "loss": 1.3618,
      "step": 57950
    },
    {
      "epoch": 1.7317051324157287,
      "grad_norm": 4.7877702713012695,
      "learning_rate": 0.00018343019102966303,
      "loss": 1.3631,
      "step": 58000
    },
    {
      "epoch": 1.7331979816678111,
      "grad_norm": 5.616472244262695,
      "learning_rate": 0.00018341590523374422,
      "loss": 1.3558,
      "step": 58050
    },
    {
      "epoch": 1.7346908309198938,
      "grad_norm": 5.436590671539307,
      "learning_rate": 0.00018340161943782536,
      "loss": 1.3999,
      "step": 58100
    },
    {
      "epoch": 1.736183680171976,
      "grad_norm": 4.245835304260254,
      "learning_rate": 0.00018338733364190655,
      "loss": 1.3592,
      "step": 58150
    },
    {
      "epoch": 1.7376765294240588,
      "grad_norm": 4.440322399139404,
      "learning_rate": 0.0001833730478459877,
      "loss": 1.312,
      "step": 58200
    },
    {
      "epoch": 1.7391693786761413,
      "grad_norm": 4.439347743988037,
      "learning_rate": 0.00018335876205006888,
      "loss": 1.3769,
      "step": 58250
    },
    {
      "epoch": 1.7406622279282238,
      "grad_norm": 4.374370098114014,
      "learning_rate": 0.00018334447625415002,
      "loss": 1.3328,
      "step": 58300
    },
    {
      "epoch": 1.7421550771803065,
      "grad_norm": 3.9568932056427,
      "learning_rate": 0.00018333019045823119,
      "loss": 1.2992,
      "step": 58350
    },
    {
      "epoch": 1.7436479264323888,
      "grad_norm": 4.133091926574707,
      "learning_rate": 0.00018331590466231238,
      "loss": 1.3685,
      "step": 58400
    },
    {
      "epoch": 1.7451407756844715,
      "grad_norm": 4.492006778717041,
      "learning_rate": 0.00018330161886639351,
      "loss": 1.3543,
      "step": 58450
    },
    {
      "epoch": 1.746633624936554,
      "grad_norm": 2.6073009967803955,
      "learning_rate": 0.0001832873330704747,
      "loss": 1.36,
      "step": 58500
    },
    {
      "epoch": 1.7481264741886364,
      "grad_norm": 4.137593746185303,
      "learning_rate": 0.00018327304727455584,
      "loss": 1.3599,
      "step": 58550
    },
    {
      "epoch": 1.749619323440719,
      "grad_norm": 4.932790756225586,
      "learning_rate": 0.00018325876147863704,
      "loss": 1.3265,
      "step": 58600
    },
    {
      "epoch": 1.7511121726928014,
      "grad_norm": 4.623594760894775,
      "learning_rate": 0.00018324447568271817,
      "loss": 1.3186,
      "step": 58650
    },
    {
      "epoch": 1.752605021944884,
      "grad_norm": 4.066229343414307,
      "learning_rate": 0.00018323018988679936,
      "loss": 1.3196,
      "step": 58700
    },
    {
      "epoch": 1.7540978711969666,
      "grad_norm": 4.105048656463623,
      "learning_rate": 0.00018321590409088053,
      "loss": 1.3663,
      "step": 58750
    },
    {
      "epoch": 1.755590720449049,
      "grad_norm": 4.454251766204834,
      "learning_rate": 0.0001832016182949617,
      "loss": 1.2823,
      "step": 58800
    },
    {
      "epoch": 1.7570835697011316,
      "grad_norm": 4.3030242919921875,
      "learning_rate": 0.00018318733249904286,
      "loss": 1.3267,
      "step": 58850
    },
    {
      "epoch": 1.758576418953214,
      "grad_norm": 5.350499629974365,
      "learning_rate": 0.00018317304670312402,
      "loss": 1.2992,
      "step": 58900
    },
    {
      "epoch": 1.7600692682052967,
      "grad_norm": 4.497105598449707,
      "learning_rate": 0.0001831587609072052,
      "loss": 1.2759,
      "step": 58950
    },
    {
      "epoch": 1.761562117457379,
      "grad_norm": 4.683063983917236,
      "learning_rate": 0.00018314447511128635,
      "loss": 1.3342,
      "step": 59000
    },
    {
      "epoch": 1.7630549667094617,
      "grad_norm": 3.4937946796417236,
      "learning_rate": 0.00018313018931536752,
      "loss": 1.2969,
      "step": 59050
    },
    {
      "epoch": 1.7645478159615442,
      "grad_norm": 5.422043800354004,
      "learning_rate": 0.00018311590351944868,
      "loss": 1.3116,
      "step": 59100
    },
    {
      "epoch": 1.7660406652136267,
      "grad_norm": 3.8882923126220703,
      "learning_rate": 0.00018310161772352985,
      "loss": 1.3073,
      "step": 59150
    },
    {
      "epoch": 1.7675335144657094,
      "grad_norm": 5.60578727722168,
      "learning_rate": 0.00018308733192761104,
      "loss": 1.31,
      "step": 59200
    },
    {
      "epoch": 1.7690263637177917,
      "grad_norm": 5.257917881011963,
      "learning_rate": 0.00018307304613169218,
      "loss": 1.3617,
      "step": 59250
    },
    {
      "epoch": 1.7705192129698744,
      "grad_norm": 4.838101863861084,
      "learning_rate": 0.00018305876033577337,
      "loss": 1.367,
      "step": 59300
    },
    {
      "epoch": 1.7720120622219568,
      "grad_norm": 4.63686990737915,
      "learning_rate": 0.0001830444745398545,
      "loss": 1.3653,
      "step": 59350
    },
    {
      "epoch": 1.7735049114740393,
      "grad_norm": 3.8766531944274902,
      "learning_rate": 0.0001830301887439357,
      "loss": 1.298,
      "step": 59400
    },
    {
      "epoch": 1.774997760726122,
      "grad_norm": 5.356032848358154,
      "learning_rate": 0.00018301590294801686,
      "loss": 1.3132,
      "step": 59450
    },
    {
      "epoch": 1.7764906099782043,
      "grad_norm": 4.647774696350098,
      "learning_rate": 0.00018300161715209803,
      "loss": 1.3594,
      "step": 59500
    },
    {
      "epoch": 1.777983459230287,
      "grad_norm": 3.6505117416381836,
      "learning_rate": 0.0001829873313561792,
      "loss": 1.3178,
      "step": 59550
    },
    {
      "epoch": 1.7794763084823695,
      "grad_norm": 4.109731674194336,
      "learning_rate": 0.00018297304556026036,
      "loss": 1.2996,
      "step": 59600
    },
    {
      "epoch": 1.780969157734452,
      "grad_norm": 3.942424774169922,
      "learning_rate": 0.00018295875976434152,
      "loss": 1.3066,
      "step": 59650
    },
    {
      "epoch": 1.7824620069865345,
      "grad_norm": 5.511881351470947,
      "learning_rate": 0.00018294447396842268,
      "loss": 1.3009,
      "step": 59700
    },
    {
      "epoch": 1.783954856238617,
      "grad_norm": 4.0874552726745605,
      "learning_rate": 0.00018293018817250385,
      "loss": 1.3426,
      "step": 59750
    },
    {
      "epoch": 1.7854477054906996,
      "grad_norm": 4.787434101104736,
      "learning_rate": 0.00018291590237658501,
      "loss": 1.3765,
      "step": 59800
    },
    {
      "epoch": 1.7869405547427821,
      "grad_norm": 4.477612018585205,
      "learning_rate": 0.00018290161658066618,
      "loss": 1.3215,
      "step": 59850
    },
    {
      "epoch": 1.7884334039948646,
      "grad_norm": 6.102437973022461,
      "learning_rate": 0.00018288733078474737,
      "loss": 1.2529,
      "step": 59900
    },
    {
      "epoch": 1.789926253246947,
      "grad_norm": 4.351520538330078,
      "learning_rate": 0.0001828730449888285,
      "loss": 1.2889,
      "step": 59950
    },
    {
      "epoch": 1.7914191024990296,
      "grad_norm": 5.043869972229004,
      "learning_rate": 0.0001828587591929097,
      "loss": 1.3149,
      "step": 60000
    },
    {
      "epoch": 1.7929119517511123,
      "grad_norm": 4.172428607940674,
      "learning_rate": 0.00018284447339699084,
      "loss": 1.2703,
      "step": 60050
    },
    {
      "epoch": 1.7944048010031945,
      "grad_norm": 5.1870903968811035,
      "learning_rate": 0.00018283018760107203,
      "loss": 1.3107,
      "step": 60100
    },
    {
      "epoch": 1.7958976502552773,
      "grad_norm": 4.141382694244385,
      "learning_rate": 0.00018281590180515317,
      "loss": 1.3252,
      "step": 60150
    },
    {
      "epoch": 1.7973904995073597,
      "grad_norm": 3.9119277000427246,
      "learning_rate": 0.00018280161600923436,
      "loss": 1.3875,
      "step": 60200
    },
    {
      "epoch": 1.7988833487594422,
      "grad_norm": 4.159456729888916,
      "learning_rate": 0.00018278733021331552,
      "loss": 1.3871,
      "step": 60250
    },
    {
      "epoch": 1.800376198011525,
      "grad_norm": 4.536618709564209,
      "learning_rate": 0.0001827730444173967,
      "loss": 1.2892,
      "step": 60300
    },
    {
      "epoch": 1.8018690472636072,
      "grad_norm": 7.839147090911865,
      "learning_rate": 0.00018275875862147785,
      "loss": 1.2953,
      "step": 60350
    },
    {
      "epoch": 1.80336189651569,
      "grad_norm": 3.6673262119293213,
      "learning_rate": 0.00018274447282555902,
      "loss": 1.3245,
      "step": 60400
    },
    {
      "epoch": 1.8048547457677724,
      "grad_norm": 4.275418281555176,
      "learning_rate": 0.00018273018702964018,
      "loss": 1.3061,
      "step": 60450
    },
    {
      "epoch": 1.8063475950198549,
      "grad_norm": 4.154748916625977,
      "learning_rate": 0.00018271590123372135,
      "loss": 1.3127,
      "step": 60500
    },
    {
      "epoch": 1.8078404442719376,
      "grad_norm": 4.65784215927124,
      "learning_rate": 0.0001827016154378025,
      "loss": 1.3717,
      "step": 60550
    },
    {
      "epoch": 1.8093332935240198,
      "grad_norm": 6.874159812927246,
      "learning_rate": 0.00018268732964188368,
      "loss": 1.3043,
      "step": 60600
    },
    {
      "epoch": 1.8108261427761025,
      "grad_norm": 5.431716442108154,
      "learning_rate": 0.00018267304384596484,
      "loss": 1.3138,
      "step": 60650
    },
    {
      "epoch": 1.812318992028185,
      "grad_norm": 4.631954669952393,
      "learning_rate": 0.00018265875805004603,
      "loss": 1.3241,
      "step": 60700
    },
    {
      "epoch": 1.8138118412802675,
      "grad_norm": 4.336903095245361,
      "learning_rate": 0.00018264447225412717,
      "loss": 1.2709,
      "step": 60750
    },
    {
      "epoch": 1.81530469053235,
      "grad_norm": 4.103146553039551,
      "learning_rate": 0.00018263018645820836,
      "loss": 1.2912,
      "step": 60800
    },
    {
      "epoch": 1.8167975397844325,
      "grad_norm": 4.709009170532227,
      "learning_rate": 0.0001826159006622895,
      "loss": 1.304,
      "step": 60850
    },
    {
      "epoch": 1.8182903890365152,
      "grad_norm": 5.0871148109436035,
      "learning_rate": 0.00018260161486637066,
      "loss": 1.301,
      "step": 60900
    },
    {
      "epoch": 1.8197832382885977,
      "grad_norm": 4.265575408935547,
      "learning_rate": 0.00018258732907045183,
      "loss": 1.3914,
      "step": 60950
    },
    {
      "epoch": 1.8212760875406802,
      "grad_norm": 4.705594539642334,
      "learning_rate": 0.000182573043274533,
      "loss": 1.3056,
      "step": 61000
    },
    {
      "epoch": 1.8227689367927626,
      "grad_norm": 3.86602783203125,
      "learning_rate": 0.00018255875747861418,
      "loss": 1.3394,
      "step": 61050
    },
    {
      "epoch": 1.8242617860448451,
      "grad_norm": 3.9397497177124023,
      "learning_rate": 0.00018254447168269532,
      "loss": 1.3625,
      "step": 61100
    },
    {
      "epoch": 1.8257546352969278,
      "grad_norm": 4.313930988311768,
      "learning_rate": 0.00018253018588677651,
      "loss": 1.3445,
      "step": 61150
    },
    {
      "epoch": 1.82724748454901,
      "grad_norm": 5.848763942718506,
      "learning_rate": 0.00018251590009085765,
      "loss": 1.3896,
      "step": 61200
    },
    {
      "epoch": 1.8287403338010928,
      "grad_norm": 4.323721885681152,
      "learning_rate": 0.00018250161429493884,
      "loss": 1.3616,
      "step": 61250
    },
    {
      "epoch": 1.8302331830531753,
      "grad_norm": 5.140015125274658,
      "learning_rate": 0.00018248732849901998,
      "loss": 1.2641,
      "step": 61300
    },
    {
      "epoch": 1.8317260323052578,
      "grad_norm": 5.685877799987793,
      "learning_rate": 0.00018247304270310117,
      "loss": 1.3396,
      "step": 61350
    },
    {
      "epoch": 1.8332188815573405,
      "grad_norm": 4.268862724304199,
      "learning_rate": 0.00018245875690718234,
      "loss": 1.3201,
      "step": 61400
    },
    {
      "epoch": 1.8347117308094227,
      "grad_norm": 5.0206732749938965,
      "learning_rate": 0.0001824444711112635,
      "loss": 1.372,
      "step": 61450
    },
    {
      "epoch": 1.8362045800615054,
      "grad_norm": 4.32529354095459,
      "learning_rate": 0.00018243018531534467,
      "loss": 1.3,
      "step": 61500
    },
    {
      "epoch": 1.837697429313588,
      "grad_norm": 5.349672317504883,
      "learning_rate": 0.00018241589951942583,
      "loss": 1.2957,
      "step": 61550
    },
    {
      "epoch": 1.8391902785656704,
      "grad_norm": 4.735100746154785,
      "learning_rate": 0.000182401613723507,
      "loss": 1.3546,
      "step": 61600
    },
    {
      "epoch": 1.8406831278177531,
      "grad_norm": 4.489027500152588,
      "learning_rate": 0.00018238732792758816,
      "loss": 1.2695,
      "step": 61650
    },
    {
      "epoch": 1.8421759770698354,
      "grad_norm": 3.2674996852874756,
      "learning_rate": 0.00018237304213166933,
      "loss": 1.3425,
      "step": 61700
    },
    {
      "epoch": 1.843668826321918,
      "grad_norm": 3.9815094470977783,
      "learning_rate": 0.0001823587563357505,
      "loss": 1.3396,
      "step": 61750
    },
    {
      "epoch": 1.8451616755740006,
      "grad_norm": 3.9253129959106445,
      "learning_rate": 0.00018234447053983165,
      "loss": 1.3641,
      "step": 61800
    },
    {
      "epoch": 1.846654524826083,
      "grad_norm": 6.536600112915039,
      "learning_rate": 0.00018233018474391285,
      "loss": 1.2578,
      "step": 61850
    },
    {
      "epoch": 1.8481473740781655,
      "grad_norm": 4.420536518096924,
      "learning_rate": 0.00018231589894799398,
      "loss": 1.3635,
      "step": 61900
    },
    {
      "epoch": 1.849640223330248,
      "grad_norm": 3.770279884338379,
      "learning_rate": 0.00018230161315207518,
      "loss": 1.2869,
      "step": 61950
    },
    {
      "epoch": 1.8511330725823307,
      "grad_norm": 4.154125213623047,
      "learning_rate": 0.0001822873273561563,
      "loss": 1.3246,
      "step": 62000
    },
    {
      "epoch": 1.8526259218344132,
      "grad_norm": 3.884587049484253,
      "learning_rate": 0.0001822730415602375,
      "loss": 1.329,
      "step": 62050
    },
    {
      "epoch": 1.8541187710864957,
      "grad_norm": 5.329457759857178,
      "learning_rate": 0.00018225875576431864,
      "loss": 1.3787,
      "step": 62100
    },
    {
      "epoch": 1.8556116203385782,
      "grad_norm": 5.966410160064697,
      "learning_rate": 0.00018224446996839983,
      "loss": 1.3574,
      "step": 62150
    },
    {
      "epoch": 1.8571044695906607,
      "grad_norm": 5.356328010559082,
      "learning_rate": 0.000182230184172481,
      "loss": 1.3113,
      "step": 62200
    },
    {
      "epoch": 1.8585973188427434,
      "grad_norm": 3.6018612384796143,
      "learning_rate": 0.00018221589837656216,
      "loss": 1.3296,
      "step": 62250
    },
    {
      "epoch": 1.8600901680948256,
      "grad_norm": 5.247616767883301,
      "learning_rate": 0.00018220161258064333,
      "loss": 1.3518,
      "step": 62300
    },
    {
      "epoch": 1.8615830173469083,
      "grad_norm": 5.828862190246582,
      "learning_rate": 0.0001821873267847245,
      "loss": 1.3027,
      "step": 62350
    },
    {
      "epoch": 1.8630758665989908,
      "grad_norm": 5.451898574829102,
      "learning_rate": 0.00018217304098880566,
      "loss": 1.3647,
      "step": 62400
    },
    {
      "epoch": 1.8645687158510733,
      "grad_norm": 3.905299663543701,
      "learning_rate": 0.00018215875519288682,
      "loss": 1.3528,
      "step": 62450
    },
    {
      "epoch": 1.866061565103156,
      "grad_norm": 4.312819004058838,
      "learning_rate": 0.000182144469396968,
      "loss": 1.3487,
      "step": 62500
    },
    {
      "epoch": 1.8675544143552383,
      "grad_norm": 4.3358073234558105,
      "learning_rate": 0.00018213018360104915,
      "loss": 1.2925,
      "step": 62550
    },
    {
      "epoch": 1.869047263607321,
      "grad_norm": 5.664048194885254,
      "learning_rate": 0.00018211589780513032,
      "loss": 1.3822,
      "step": 62600
    },
    {
      "epoch": 1.8705401128594035,
      "grad_norm": 4.992645263671875,
      "learning_rate": 0.0001821016120092115,
      "loss": 1.3819,
      "step": 62650
    },
    {
      "epoch": 1.872032962111486,
      "grad_norm": 4.036514759063721,
      "learning_rate": 0.00018208732621329265,
      "loss": 1.3193,
      "step": 62700
    },
    {
      "epoch": 1.8735258113635687,
      "grad_norm": 4.094419956207275,
      "learning_rate": 0.00018207304041737384,
      "loss": 1.3649,
      "step": 62750
    },
    {
      "epoch": 1.875018660615651,
      "grad_norm": 5.437168121337891,
      "learning_rate": 0.00018205875462145497,
      "loss": 1.3445,
      "step": 62800
    },
    {
      "epoch": 1.8765115098677336,
      "grad_norm": 4.937995433807373,
      "learning_rate": 0.00018204446882553617,
      "loss": 1.3588,
      "step": 62850
    },
    {
      "epoch": 1.8780043591198161,
      "grad_norm": 3.918588876724243,
      "learning_rate": 0.00018203018302961733,
      "loss": 1.3195,
      "step": 62900
    },
    {
      "epoch": 1.8794972083718986,
      "grad_norm": 4.650214672088623,
      "learning_rate": 0.0001820158972336985,
      "loss": 1.3812,
      "step": 62950
    },
    {
      "epoch": 1.880990057623981,
      "grad_norm": 3.844249725341797,
      "learning_rate": 0.00018200161143777966,
      "loss": 1.2877,
      "step": 63000
    },
    {
      "epoch": 1.8824829068760636,
      "grad_norm": 4.988340854644775,
      "learning_rate": 0.00018198732564186083,
      "loss": 1.2705,
      "step": 63050
    },
    {
      "epoch": 1.8839757561281463,
      "grad_norm": 4.229214668273926,
      "learning_rate": 0.000181973039845942,
      "loss": 1.3314,
      "step": 63100
    },
    {
      "epoch": 1.8854686053802288,
      "grad_norm": 4.095324993133545,
      "learning_rate": 0.00018195875405002315,
      "loss": 1.3779,
      "step": 63150
    },
    {
      "epoch": 1.8869614546323112,
      "grad_norm": 5.325993537902832,
      "learning_rate": 0.00018194446825410432,
      "loss": 1.3163,
      "step": 63200
    },
    {
      "epoch": 1.8884543038843937,
      "grad_norm": 5.301263809204102,
      "learning_rate": 0.00018193018245818548,
      "loss": 1.4235,
      "step": 63250
    },
    {
      "epoch": 1.8899471531364762,
      "grad_norm": 5.347028732299805,
      "learning_rate": 0.00018191589666226665,
      "loss": 1.356,
      "step": 63300
    },
    {
      "epoch": 1.891440002388559,
      "grad_norm": 4.957263469696045,
      "learning_rate": 0.00018190161086634784,
      "loss": 1.2915,
      "step": 63350
    },
    {
      "epoch": 1.8929328516406412,
      "grad_norm": 3.9531173706054688,
      "learning_rate": 0.00018188732507042898,
      "loss": 1.2905,
      "step": 63400
    },
    {
      "epoch": 1.8944257008927239,
      "grad_norm": 5.983827590942383,
      "learning_rate": 0.00018187303927451017,
      "loss": 1.3129,
      "step": 63450
    },
    {
      "epoch": 1.8959185501448064,
      "grad_norm": 4.4479522705078125,
      "learning_rate": 0.0001818587534785913,
      "loss": 1.3528,
      "step": 63500
    },
    {
      "epoch": 1.8974113993968889,
      "grad_norm": 4.956698417663574,
      "learning_rate": 0.00018184446768267247,
      "loss": 1.2675,
      "step": 63550
    },
    {
      "epoch": 1.8989042486489716,
      "grad_norm": 3.7053678035736084,
      "learning_rate": 0.00018183018188675364,
      "loss": 1.3316,
      "step": 63600
    },
    {
      "epoch": 1.9003970979010538,
      "grad_norm": 5.076517105102539,
      "learning_rate": 0.0001818158960908348,
      "loss": 1.3603,
      "step": 63650
    },
    {
      "epoch": 1.9018899471531365,
      "grad_norm": 4.377649307250977,
      "learning_rate": 0.000181801610294916,
      "loss": 1.3491,
      "step": 63700
    },
    {
      "epoch": 1.903382796405219,
      "grad_norm": 5.296698570251465,
      "learning_rate": 0.00018178732449899713,
      "loss": 1.3768,
      "step": 63750
    },
    {
      "epoch": 1.9048756456573015,
      "grad_norm": 4.674185752868652,
      "learning_rate": 0.00018177303870307832,
      "loss": 1.3599,
      "step": 63800
    },
    {
      "epoch": 1.9063684949093842,
      "grad_norm": 5.758728981018066,
      "learning_rate": 0.00018175875290715946,
      "loss": 1.3239,
      "step": 63850
    },
    {
      "epoch": 1.9078613441614665,
      "grad_norm": 4.195470333099365,
      "learning_rate": 0.00018174446711124065,
      "loss": 1.3132,
      "step": 63900
    },
    {
      "epoch": 1.9093541934135492,
      "grad_norm": 4.118239402770996,
      "learning_rate": 0.0001817301813153218,
      "loss": 1.2887,
      "step": 63950
    },
    {
      "epoch": 1.9108470426656317,
      "grad_norm": 3.598353147506714,
      "learning_rate": 0.00018171589551940298,
      "loss": 1.3341,
      "step": 64000
    },
    {
      "epoch": 1.9123398919177141,
      "grad_norm": 3.8190245628356934,
      "learning_rate": 0.00018170160972348415,
      "loss": 1.3011,
      "step": 64050
    },
    {
      "epoch": 1.9138327411697966,
      "grad_norm": 6.505800724029541,
      "learning_rate": 0.0001816873239275653,
      "loss": 1.3583,
      "step": 64100
    },
    {
      "epoch": 1.915325590421879,
      "grad_norm": 4.134885311126709,
      "learning_rate": 0.00018167303813164647,
      "loss": 1.3363,
      "step": 64150
    },
    {
      "epoch": 1.9168184396739618,
      "grad_norm": 3.753544807434082,
      "learning_rate": 0.00018165875233572764,
      "loss": 1.3415,
      "step": 64200
    },
    {
      "epoch": 1.9183112889260443,
      "grad_norm": 3.716143846511841,
      "learning_rate": 0.0001816444665398088,
      "loss": 1.3276,
      "step": 64250
    },
    {
      "epoch": 1.9198041381781268,
      "grad_norm": 4.619287490844727,
      "learning_rate": 0.00018163018074388997,
      "loss": 1.3624,
      "step": 64300
    },
    {
      "epoch": 1.9212969874302093,
      "grad_norm": 3.8095598220825195,
      "learning_rate": 0.00018161589494797113,
      "loss": 1.3488,
      "step": 64350
    },
    {
      "epoch": 1.9227898366822918,
      "grad_norm": 6.268500328063965,
      "learning_rate": 0.0001816016091520523,
      "loss": 1.3481,
      "step": 64400
    },
    {
      "epoch": 1.9242826859343745,
      "grad_norm": 6.642594814300537,
      "learning_rate": 0.00018158732335613346,
      "loss": 1.3392,
      "step": 64450
    },
    {
      "epoch": 1.9257755351864567,
      "grad_norm": 5.921909332275391,
      "learning_rate": 0.00018157303756021465,
      "loss": 1.3218,
      "step": 64500
    },
    {
      "epoch": 1.9272683844385394,
      "grad_norm": 4.467897415161133,
      "learning_rate": 0.0001815587517642958,
      "loss": 1.3087,
      "step": 64550
    },
    {
      "epoch": 1.928761233690622,
      "grad_norm": 6.775788307189941,
      "learning_rate": 0.00018154446596837698,
      "loss": 1.3712,
      "step": 64600
    },
    {
      "epoch": 1.9302540829427044,
      "grad_norm": 4.561954975128174,
      "learning_rate": 0.00018153018017245812,
      "loss": 1.3214,
      "step": 64650
    },
    {
      "epoch": 1.931746932194787,
      "grad_norm": 5.288097381591797,
      "learning_rate": 0.0001815158943765393,
      "loss": 1.3926,
      "step": 64700
    },
    {
      "epoch": 1.9332397814468694,
      "grad_norm": 4.883842945098877,
      "learning_rate": 0.00018150160858062045,
      "loss": 1.2545,
      "step": 64750
    },
    {
      "epoch": 1.934732630698952,
      "grad_norm": 4.528720378875732,
      "learning_rate": 0.00018148732278470164,
      "loss": 1.3733,
      "step": 64800
    },
    {
      "epoch": 1.9362254799510346,
      "grad_norm": 3.9458446502685547,
      "learning_rate": 0.0001814730369887828,
      "loss": 1.3685,
      "step": 64850
    },
    {
      "epoch": 1.937718329203117,
      "grad_norm": 5.117423057556152,
      "learning_rate": 0.00018145875119286397,
      "loss": 1.3712,
      "step": 64900
    },
    {
      "epoch": 1.9392111784551997,
      "grad_norm": 4.276913166046143,
      "learning_rate": 0.00018144446539694514,
      "loss": 1.2996,
      "step": 64950
    },
    {
      "epoch": 1.940704027707282,
      "grad_norm": 3.9713099002838135,
      "learning_rate": 0.0001814301796010263,
      "loss": 1.301,
      "step": 65000
    },
    {
      "epoch": 1.9421968769593647,
      "grad_norm": 4.240658760070801,
      "learning_rate": 0.00018141589380510747,
      "loss": 1.4013,
      "step": 65050
    },
    {
      "epoch": 1.9436897262114472,
      "grad_norm": 4.406320571899414,
      "learning_rate": 0.00018140160800918863,
      "loss": 1.3481,
      "step": 65100
    },
    {
      "epoch": 1.9451825754635297,
      "grad_norm": 4.139687538146973,
      "learning_rate": 0.0001813873222132698,
      "loss": 1.3179,
      "step": 65150
    },
    {
      "epoch": 1.9466754247156122,
      "grad_norm": 3.8167126178741455,
      "learning_rate": 0.00018137303641735096,
      "loss": 1.2714,
      "step": 65200
    },
    {
      "epoch": 1.9481682739676947,
      "grad_norm": 4.680548191070557,
      "learning_rate": 0.00018135875062143212,
      "loss": 1.3571,
      "step": 65250
    },
    {
      "epoch": 1.9496611232197774,
      "grad_norm": 8.530779838562012,
      "learning_rate": 0.00018134446482551332,
      "loss": 1.323,
      "step": 65300
    },
    {
      "epoch": 1.9511539724718598,
      "grad_norm": 4.576430320739746,
      "learning_rate": 0.00018133017902959445,
      "loss": 1.3709,
      "step": 65350
    },
    {
      "epoch": 1.9526468217239423,
      "grad_norm": 5.320089817047119,
      "learning_rate": 0.00018131589323367565,
      "loss": 1.3857,
      "step": 65400
    },
    {
      "epoch": 1.9541396709760248,
      "grad_norm": 4.247587203979492,
      "learning_rate": 0.00018130160743775678,
      "loss": 1.3637,
      "step": 65450
    },
    {
      "epoch": 1.9556325202281073,
      "grad_norm": 4.2958903312683105,
      "learning_rate": 0.00018128732164183797,
      "loss": 1.2861,
      "step": 65500
    },
    {
      "epoch": 1.95712536948019,
      "grad_norm": 4.818902969360352,
      "learning_rate": 0.00018127303584591914,
      "loss": 1.3264,
      "step": 65550
    },
    {
      "epoch": 1.9586182187322723,
      "grad_norm": 4.755589962005615,
      "learning_rate": 0.0001812587500500003,
      "loss": 1.3821,
      "step": 65600
    },
    {
      "epoch": 1.960111067984355,
      "grad_norm": 3.589580774307251,
      "learning_rate": 0.00018124446425408147,
      "loss": 1.2799,
      "step": 65650
    },
    {
      "epoch": 1.9616039172364375,
      "grad_norm": 5.7323431968688965,
      "learning_rate": 0.00018123017845816263,
      "loss": 1.2616,
      "step": 65700
    },
    {
      "epoch": 1.96309676648852,
      "grad_norm": 6.535925388336182,
      "learning_rate": 0.0001812158926622438,
      "loss": 1.3391,
      "step": 65750
    },
    {
      "epoch": 1.9645896157406026,
      "grad_norm": 3.783618688583374,
      "learning_rate": 0.00018120160686632496,
      "loss": 1.2743,
      "step": 65800
    },
    {
      "epoch": 1.966082464992685,
      "grad_norm": 4.055028438568115,
      "learning_rate": 0.00018118732107040613,
      "loss": 1.3319,
      "step": 65850
    },
    {
      "epoch": 1.9675753142447676,
      "grad_norm": 4.827310562133789,
      "learning_rate": 0.0001811730352744873,
      "loss": 1.3518,
      "step": 65900
    },
    {
      "epoch": 1.96906816349685,
      "grad_norm": 4.298334121704102,
      "learning_rate": 0.00018115874947856846,
      "loss": 1.3755,
      "step": 65950
    },
    {
      "epoch": 1.9705610127489326,
      "grad_norm": 4.718268871307373,
      "learning_rate": 0.00018114446368264962,
      "loss": 1.2615,
      "step": 66000
    },
    {
      "epoch": 1.9720538620010153,
      "grad_norm": 4.387974262237549,
      "learning_rate": 0.00018113017788673079,
      "loss": 1.3475,
      "step": 66050
    },
    {
      "epoch": 1.9735467112530976,
      "grad_norm": 4.896945476531982,
      "learning_rate": 0.00018111589209081198,
      "loss": 1.3145,
      "step": 66100
    },
    {
      "epoch": 1.9750395605051803,
      "grad_norm": 5.453362941741943,
      "learning_rate": 0.00018110160629489312,
      "loss": 1.3377,
      "step": 66150
    },
    {
      "epoch": 1.9765324097572627,
      "grad_norm": 6.6892499923706055,
      "learning_rate": 0.00018108732049897428,
      "loss": 1.2734,
      "step": 66200
    },
    {
      "epoch": 1.9780252590093452,
      "grad_norm": 4.2804765701293945,
      "learning_rate": 0.00018107303470305544,
      "loss": 1.3742,
      "step": 66250
    },
    {
      "epoch": 1.9795181082614277,
      "grad_norm": 4.235034465789795,
      "learning_rate": 0.0001810587489071366,
      "loss": 1.2921,
      "step": 66300
    },
    {
      "epoch": 1.9810109575135102,
      "grad_norm": 4.113801956176758,
      "learning_rate": 0.0001810444631112178,
      "loss": 1.3475,
      "step": 66350
    },
    {
      "epoch": 1.982503806765593,
      "grad_norm": 4.276202201843262,
      "learning_rate": 0.00018103017731529894,
      "loss": 1.2955,
      "step": 66400
    },
    {
      "epoch": 1.9839966560176754,
      "grad_norm": 4.72022008895874,
      "learning_rate": 0.00018101589151938013,
      "loss": 1.3676,
      "step": 66450
    },
    {
      "epoch": 1.9854895052697579,
      "grad_norm": 4.182514190673828,
      "learning_rate": 0.00018100160572346127,
      "loss": 1.3706,
      "step": 66500
    },
    {
      "epoch": 1.9869823545218404,
      "grad_norm": 3.8737592697143555,
      "learning_rate": 0.00018098731992754246,
      "loss": 1.3103,
      "step": 66550
    },
    {
      "epoch": 1.9884752037739228,
      "grad_norm": 6.266764163970947,
      "learning_rate": 0.0001809730341316236,
      "loss": 1.3175,
      "step": 66600
    },
    {
      "epoch": 1.9899680530260055,
      "grad_norm": 5.243070125579834,
      "learning_rate": 0.0001809587483357048,
      "loss": 1.3484,
      "step": 66650
    },
    {
      "epoch": 1.9914609022780878,
      "grad_norm": 4.289467811584473,
      "learning_rate": 0.00018094446253978595,
      "loss": 1.3806,
      "step": 66700
    },
    {
      "epoch": 1.9929537515301705,
      "grad_norm": 5.450198173522949,
      "learning_rate": 0.00018093017674386712,
      "loss": 1.3155,
      "step": 66750
    },
    {
      "epoch": 1.994446600782253,
      "grad_norm": 4.389066219329834,
      "learning_rate": 0.00018091589094794828,
      "loss": 1.293,
      "step": 66800
    },
    {
      "epoch": 1.9959394500343355,
      "grad_norm": 4.606175422668457,
      "learning_rate": 0.00018090160515202945,
      "loss": 1.3558,
      "step": 66850
    },
    {
      "epoch": 1.9974322992864182,
      "grad_norm": 4.8083319664001465,
      "learning_rate": 0.0001808873193561106,
      "loss": 1.2621,
      "step": 66900
    },
    {
      "epoch": 1.9989251485385005,
      "grad_norm": 3.760852813720703,
      "learning_rate": 0.00018087303356019178,
      "loss": 1.3539,
      "step": 66950
    },
    {
      "epoch": 2.000417997790583,
      "grad_norm": 3.6912026405334473,
      "learning_rate": 0.00018085874776427294,
      "loss": 1.2619,
      "step": 67000
    },
    {
      "epoch": 2.0019108470426654,
      "grad_norm": 4.681028842926025,
      "learning_rate": 0.0001808444619683541,
      "loss": 1.2457,
      "step": 67050
    },
    {
      "epoch": 2.003403696294748,
      "grad_norm": 4.194242477416992,
      "learning_rate": 0.00018083017617243527,
      "loss": 1.297,
      "step": 67100
    },
    {
      "epoch": 2.004896545546831,
      "grad_norm": 3.831331729888916,
      "learning_rate": 0.00018081589037651646,
      "loss": 1.2989,
      "step": 67150
    },
    {
      "epoch": 2.006389394798913,
      "grad_norm": 4.249265193939209,
      "learning_rate": 0.0001808016045805976,
      "loss": 1.2078,
      "step": 67200
    },
    {
      "epoch": 2.007882244050996,
      "grad_norm": 7.441315650939941,
      "learning_rate": 0.0001807873187846788,
      "loss": 1.2945,
      "step": 67250
    },
    {
      "epoch": 2.009375093303078,
      "grad_norm": 4.2568745613098145,
      "learning_rate": 0.00018077303298875993,
      "loss": 1.2295,
      "step": 67300
    },
    {
      "epoch": 2.0108679425551608,
      "grad_norm": 5.394452095031738,
      "learning_rate": 0.00018075874719284112,
      "loss": 1.2135,
      "step": 67350
    },
    {
      "epoch": 2.0123607918072435,
      "grad_norm": 3.199411630630493,
      "learning_rate": 0.00018074446139692226,
      "loss": 1.2314,
      "step": 67400
    },
    {
      "epoch": 2.0138536410593257,
      "grad_norm": 3.7905383110046387,
      "learning_rate": 0.00018073017560100345,
      "loss": 1.2227,
      "step": 67450
    },
    {
      "epoch": 2.0153464903114084,
      "grad_norm": 7.900835990905762,
      "learning_rate": 0.00018071588980508462,
      "loss": 1.2067,
      "step": 67500
    },
    {
      "epoch": 2.0168393395634907,
      "grad_norm": 4.803965091705322,
      "learning_rate": 0.00018070160400916578,
      "loss": 1.2894,
      "step": 67550
    },
    {
      "epoch": 2.0183321888155734,
      "grad_norm": 4.046397686004639,
      "learning_rate": 0.00018068731821324694,
      "loss": 1.2627,
      "step": 67600
    },
    {
      "epoch": 2.019825038067656,
      "grad_norm": 4.9588518142700195,
      "learning_rate": 0.0001806730324173281,
      "loss": 1.1931,
      "step": 67650
    },
    {
      "epoch": 2.0213178873197384,
      "grad_norm": 3.5710065364837646,
      "learning_rate": 0.00018065874662140927,
      "loss": 1.1937,
      "step": 67700
    },
    {
      "epoch": 2.022810736571821,
      "grad_norm": 4.107451915740967,
      "learning_rate": 0.00018064446082549044,
      "loss": 1.2236,
      "step": 67750
    },
    {
      "epoch": 2.0243035858239034,
      "grad_norm": 4.8303422927856445,
      "learning_rate": 0.0001806301750295716,
      "loss": 1.2223,
      "step": 67800
    },
    {
      "epoch": 2.025796435075986,
      "grad_norm": 6.539272308349609,
      "learning_rate": 0.00018061588923365277,
      "loss": 1.3199,
      "step": 67850
    },
    {
      "epoch": 2.0272892843280688,
      "grad_norm": 4.415231227874756,
      "learning_rate": 0.00018060160343773393,
      "loss": 1.2988,
      "step": 67900
    },
    {
      "epoch": 2.028782133580151,
      "grad_norm": 4.396175861358643,
      "learning_rate": 0.00018058731764181512,
      "loss": 1.1952,
      "step": 67950
    },
    {
      "epoch": 2.0302749828322337,
      "grad_norm": 3.2673914432525635,
      "learning_rate": 0.00018057303184589626,
      "loss": 1.2285,
      "step": 68000
    },
    {
      "epoch": 2.031767832084316,
      "grad_norm": 3.6252634525299072,
      "learning_rate": 0.00018055874604997745,
      "loss": 1.314,
      "step": 68050
    },
    {
      "epoch": 2.0332606813363987,
      "grad_norm": 4.550121307373047,
      "learning_rate": 0.0001805444602540586,
      "loss": 1.2846,
      "step": 68100
    },
    {
      "epoch": 2.034753530588481,
      "grad_norm": 4.809107303619385,
      "learning_rate": 0.00018053017445813978,
      "loss": 1.3165,
      "step": 68150
    },
    {
      "epoch": 2.0362463798405637,
      "grad_norm": 4.466585159301758,
      "learning_rate": 0.00018051588866222092,
      "loss": 1.2543,
      "step": 68200
    },
    {
      "epoch": 2.0377392290926464,
      "grad_norm": 4.243835926055908,
      "learning_rate": 0.0001805016028663021,
      "loss": 1.2908,
      "step": 68250
    },
    {
      "epoch": 2.0392320783447286,
      "grad_norm": 4.369659423828125,
      "learning_rate": 0.00018048731707038328,
      "loss": 1.2167,
      "step": 68300
    },
    {
      "epoch": 2.0407249275968113,
      "grad_norm": 4.6665496826171875,
      "learning_rate": 0.00018047303127446444,
      "loss": 1.2453,
      "step": 68350
    },
    {
      "epoch": 2.0422177768488936,
      "grad_norm": 4.3825507164001465,
      "learning_rate": 0.0001804587454785456,
      "loss": 1.2272,
      "step": 68400
    },
    {
      "epoch": 2.0437106261009763,
      "grad_norm": 4.299841403961182,
      "learning_rate": 0.00018044445968262677,
      "loss": 1.2198,
      "step": 68450
    },
    {
      "epoch": 2.045203475353059,
      "grad_norm": 3.937131881713867,
      "learning_rate": 0.00018043017388670794,
      "loss": 1.2763,
      "step": 68500
    },
    {
      "epoch": 2.0466963246051413,
      "grad_norm": 6.704258918762207,
      "learning_rate": 0.0001804158880907891,
      "loss": 1.3289,
      "step": 68550
    },
    {
      "epoch": 2.048189173857224,
      "grad_norm": 3.8973612785339355,
      "learning_rate": 0.00018040160229487026,
      "loss": 1.2907,
      "step": 68600
    },
    {
      "epoch": 2.0496820231093063,
      "grad_norm": 8.648911476135254,
      "learning_rate": 0.00018038731649895143,
      "loss": 1.2488,
      "step": 68650
    },
    {
      "epoch": 2.051174872361389,
      "grad_norm": 4.553196430206299,
      "learning_rate": 0.0001803730307030326,
      "loss": 1.2489,
      "step": 68700
    },
    {
      "epoch": 2.0526677216134717,
      "grad_norm": 4.379563808441162,
      "learning_rate": 0.00018035874490711376,
      "loss": 1.253,
      "step": 68750
    },
    {
      "epoch": 2.054160570865554,
      "grad_norm": 6.393334865570068,
      "learning_rate": 0.00018034445911119492,
      "loss": 1.1783,
      "step": 68800
    },
    {
      "epoch": 2.0556534201176366,
      "grad_norm": 4.846323013305664,
      "learning_rate": 0.0001803301733152761,
      "loss": 1.2583,
      "step": 68850
    },
    {
      "epoch": 2.057146269369719,
      "grad_norm": 5.24686861038208,
      "learning_rate": 0.00018031588751935725,
      "loss": 1.333,
      "step": 68900
    },
    {
      "epoch": 2.0586391186218016,
      "grad_norm": 4.824219226837158,
      "learning_rate": 0.00018030160172343842,
      "loss": 1.2931,
      "step": 68950
    },
    {
      "epoch": 2.0601319678738843,
      "grad_norm": 5.001037120819092,
      "learning_rate": 0.0001802873159275196,
      "loss": 1.2625,
      "step": 69000
    },
    {
      "epoch": 2.0616248171259666,
      "grad_norm": 4.036411762237549,
      "learning_rate": 0.00018027303013160075,
      "loss": 1.2585,
      "step": 69050
    },
    {
      "epoch": 2.0631176663780493,
      "grad_norm": 4.597596645355225,
      "learning_rate": 0.00018025874433568194,
      "loss": 1.2812,
      "step": 69100
    },
    {
      "epoch": 2.0646105156301315,
      "grad_norm": 7.785471439361572,
      "learning_rate": 0.00018024445853976308,
      "loss": 1.2749,
      "step": 69150
    },
    {
      "epoch": 2.0661033648822142,
      "grad_norm": 4.905806541442871,
      "learning_rate": 0.00018023017274384427,
      "loss": 1.2619,
      "step": 69200
    },
    {
      "epoch": 2.0675962141342965,
      "grad_norm": 5.271987438201904,
      "learning_rate": 0.0001802158869479254,
      "loss": 1.2769,
      "step": 69250
    },
    {
      "epoch": 2.069089063386379,
      "grad_norm": 4.129353046417236,
      "learning_rate": 0.0001802016011520066,
      "loss": 1.2552,
      "step": 69300
    },
    {
      "epoch": 2.070581912638462,
      "grad_norm": 4.855686187744141,
      "learning_rate": 0.00018018731535608776,
      "loss": 1.2452,
      "step": 69350
    },
    {
      "epoch": 2.072074761890544,
      "grad_norm": 4.6511383056640625,
      "learning_rate": 0.00018017302956016893,
      "loss": 1.2529,
      "step": 69400
    },
    {
      "epoch": 2.073567611142627,
      "grad_norm": 4.4711995124816895,
      "learning_rate": 0.0001801587437642501,
      "loss": 1.2721,
      "step": 69450
    },
    {
      "epoch": 2.075060460394709,
      "grad_norm": 4.466591835021973,
      "learning_rate": 0.00018014445796833126,
      "loss": 1.2654,
      "step": 69500
    },
    {
      "epoch": 2.076553309646792,
      "grad_norm": 5.165157318115234,
      "learning_rate": 0.00018013017217241242,
      "loss": 1.2522,
      "step": 69550
    },
    {
      "epoch": 2.0780461588988746,
      "grad_norm": 4.762765407562256,
      "learning_rate": 0.00018011588637649358,
      "loss": 1.2318,
      "step": 69600
    },
    {
      "epoch": 2.079539008150957,
      "grad_norm": 5.024138927459717,
      "learning_rate": 0.00018010160058057475,
      "loss": 1.2713,
      "step": 69650
    },
    {
      "epoch": 2.0810318574030395,
      "grad_norm": 5.054821014404297,
      "learning_rate": 0.00018008731478465591,
      "loss": 1.2714,
      "step": 69700
    },
    {
      "epoch": 2.082524706655122,
      "grad_norm": 5.052605152130127,
      "learning_rate": 0.00018007302898873708,
      "loss": 1.2001,
      "step": 69750
    },
    {
      "epoch": 2.0840175559072045,
      "grad_norm": 6.121365547180176,
      "learning_rate": 0.00018005874319281827,
      "loss": 1.2372,
      "step": 69800
    },
    {
      "epoch": 2.085510405159287,
      "grad_norm": 4.6751627922058105,
      "learning_rate": 0.0001800444573968994,
      "loss": 1.2366,
      "step": 69850
    },
    {
      "epoch": 2.0870032544113695,
      "grad_norm": 4.396819591522217,
      "learning_rate": 0.0001800301716009806,
      "loss": 1.2878,
      "step": 69900
    },
    {
      "epoch": 2.088496103663452,
      "grad_norm": 5.096658229827881,
      "learning_rate": 0.00018001588580506174,
      "loss": 1.2265,
      "step": 69950
    },
    {
      "epoch": 2.0899889529155344,
      "grad_norm": 4.45212459564209,
      "learning_rate": 0.00018000160000914293,
      "loss": 1.2261,
      "step": 70000
    },
    {
      "epoch": 2.091481802167617,
      "grad_norm": 4.837035179138184,
      "learning_rate": 0.00017998731421322407,
      "loss": 1.2857,
      "step": 70050
    },
    {
      "epoch": 2.0929746514197,
      "grad_norm": 4.402126789093018,
      "learning_rate": 0.00017997302841730526,
      "loss": 1.2614,
      "step": 70100
    },
    {
      "epoch": 2.094467500671782,
      "grad_norm": 3.9073615074157715,
      "learning_rate": 0.00017995874262138642,
      "loss": 1.2485,
      "step": 70150
    },
    {
      "epoch": 2.095960349923865,
      "grad_norm": 5.109444618225098,
      "learning_rate": 0.0001799444568254676,
      "loss": 1.2674,
      "step": 70200
    },
    {
      "epoch": 2.097453199175947,
      "grad_norm": 6.542872428894043,
      "learning_rate": 0.00017993017102954875,
      "loss": 1.2715,
      "step": 70250
    },
    {
      "epoch": 2.09894604842803,
      "grad_norm": 5.090799808502197,
      "learning_rate": 0.00017991588523362992,
      "loss": 1.2821,
      "step": 70300
    },
    {
      "epoch": 2.100438897680112,
      "grad_norm": 5.429018497467041,
      "learning_rate": 0.00017990159943771108,
      "loss": 1.2994,
      "step": 70350
    },
    {
      "epoch": 2.1019317469321948,
      "grad_norm": 8.873945236206055,
      "learning_rate": 0.00017988731364179225,
      "loss": 1.2912,
      "step": 70400
    },
    {
      "epoch": 2.1034245961842775,
      "grad_norm": 5.144947528839111,
      "learning_rate": 0.0001798730278458734,
      "loss": 1.2744,
      "step": 70450
    },
    {
      "epoch": 2.1049174454363597,
      "grad_norm": 5.346638202667236,
      "learning_rate": 0.00017985874204995458,
      "loss": 1.311,
      "step": 70500
    },
    {
      "epoch": 2.1064102946884424,
      "grad_norm": 5.384441375732422,
      "learning_rate": 0.00017984445625403574,
      "loss": 1.2662,
      "step": 70550
    },
    {
      "epoch": 2.1079031439405247,
      "grad_norm": 4.645060062408447,
      "learning_rate": 0.00017983017045811693,
      "loss": 1.2366,
      "step": 70600
    },
    {
      "epoch": 2.1093959931926074,
      "grad_norm": 4.9864349365234375,
      "learning_rate": 0.00017981588466219807,
      "loss": 1.2531,
      "step": 70650
    },
    {
      "epoch": 2.11088884244469,
      "grad_norm": 4.057093143463135,
      "learning_rate": 0.00017980159886627926,
      "loss": 1.2477,
      "step": 70700
    },
    {
      "epoch": 2.1123816916967724,
      "grad_norm": 4.121668815612793,
      "learning_rate": 0.0001797873130703604,
      "loss": 1.2533,
      "step": 70750
    },
    {
      "epoch": 2.113874540948855,
      "grad_norm": 5.140614032745361,
      "learning_rate": 0.0001797730272744416,
      "loss": 1.1833,
      "step": 70800
    },
    {
      "epoch": 2.1153673902009373,
      "grad_norm": 5.469926834106445,
      "learning_rate": 0.00017975874147852273,
      "loss": 1.2746,
      "step": 70850
    },
    {
      "epoch": 2.11686023945302,
      "grad_norm": 3.0967893600463867,
      "learning_rate": 0.00017974445568260392,
      "loss": 1.2449,
      "step": 70900
    },
    {
      "epoch": 2.1183530887051027,
      "grad_norm": 4.644883155822754,
      "learning_rate": 0.00017973016988668508,
      "loss": 1.2304,
      "step": 70950
    },
    {
      "epoch": 2.119845937957185,
      "grad_norm": 4.095651626586914,
      "learning_rate": 0.00017971588409076625,
      "loss": 1.3378,
      "step": 71000
    },
    {
      "epoch": 2.1213387872092677,
      "grad_norm": 5.656352519989014,
      "learning_rate": 0.00017970159829484741,
      "loss": 1.268,
      "step": 71050
    },
    {
      "epoch": 2.12283163646135,
      "grad_norm": 5.86411190032959,
      "learning_rate": 0.00017968731249892858,
      "loss": 1.3146,
      "step": 71100
    },
    {
      "epoch": 2.1243244857134327,
      "grad_norm": 4.751001358032227,
      "learning_rate": 0.00017967302670300974,
      "loss": 1.2861,
      "step": 71150
    },
    {
      "epoch": 2.1258173349655154,
      "grad_norm": 4.90724515914917,
      "learning_rate": 0.0001796587409070909,
      "loss": 1.2307,
      "step": 71200
    },
    {
      "epoch": 2.1273101842175977,
      "grad_norm": 4.715629577636719,
      "learning_rate": 0.00017964445511117207,
      "loss": 1.2602,
      "step": 71250
    },
    {
      "epoch": 2.1288030334696804,
      "grad_norm": 3.9297828674316406,
      "learning_rate": 0.00017963016931525324,
      "loss": 1.2705,
      "step": 71300
    },
    {
      "epoch": 2.1302958827217626,
      "grad_norm": 4.373486518859863,
      "learning_rate": 0.0001796158835193344,
      "loss": 1.264,
      "step": 71350
    },
    {
      "epoch": 2.1317887319738453,
      "grad_norm": 5.277180194854736,
      "learning_rate": 0.00017960159772341557,
      "loss": 1.2703,
      "step": 71400
    },
    {
      "epoch": 2.133281581225928,
      "grad_norm": 3.8668055534362793,
      "learning_rate": 0.00017958731192749673,
      "loss": 1.2655,
      "step": 71450
    },
    {
      "epoch": 2.1347744304780103,
      "grad_norm": 4.491888046264648,
      "learning_rate": 0.0001795730261315779,
      "loss": 1.2659,
      "step": 71500
    },
    {
      "epoch": 2.136267279730093,
      "grad_norm": 4.461415767669678,
      "learning_rate": 0.00017955874033565906,
      "loss": 1.3075,
      "step": 71550
    },
    {
      "epoch": 2.1377601289821753,
      "grad_norm": 5.062393665313721,
      "learning_rate": 0.00017954445453974023,
      "loss": 1.2328,
      "step": 71600
    },
    {
      "epoch": 2.139252978234258,
      "grad_norm": 5.79408073425293,
      "learning_rate": 0.0001795301687438214,
      "loss": 1.2528,
      "step": 71650
    },
    {
      "epoch": 2.1407458274863402,
      "grad_norm": 5.496243476867676,
      "learning_rate": 0.00017951588294790255,
      "loss": 1.2409,
      "step": 71700
    },
    {
      "epoch": 2.142238676738423,
      "grad_norm": 5.3767805099487305,
      "learning_rate": 0.00017950159715198375,
      "loss": 1.2383,
      "step": 71750
    },
    {
      "epoch": 2.1437315259905056,
      "grad_norm": 6.0737199783325195,
      "learning_rate": 0.00017948731135606488,
      "loss": 1.2936,
      "step": 71800
    },
    {
      "epoch": 2.145224375242588,
      "grad_norm": 5.363690376281738,
      "learning_rate": 0.00017947302556014608,
      "loss": 1.2774,
      "step": 71850
    },
    {
      "epoch": 2.1467172244946706,
      "grad_norm": 4.2048773765563965,
      "learning_rate": 0.0001794587397642272,
      "loss": 1.2476,
      "step": 71900
    },
    {
      "epoch": 2.148210073746753,
      "grad_norm": 5.775049686431885,
      "learning_rate": 0.0001794444539683084,
      "loss": 1.2646,
      "step": 71950
    },
    {
      "epoch": 2.1497029229988356,
      "grad_norm": 4.7285475730896,
      "learning_rate": 0.00017943016817238957,
      "loss": 1.3038,
      "step": 72000
    },
    {
      "epoch": 2.1511957722509183,
      "grad_norm": 4.310266971588135,
      "learning_rate": 0.00017941588237647073,
      "loss": 1.2478,
      "step": 72050
    },
    {
      "epoch": 2.1526886215030006,
      "grad_norm": 5.618819713592529,
      "learning_rate": 0.0001794015965805519,
      "loss": 1.2788,
      "step": 72100
    },
    {
      "epoch": 2.1541814707550833,
      "grad_norm": 5.624669075012207,
      "learning_rate": 0.00017938731078463306,
      "loss": 1.3053,
      "step": 72150
    },
    {
      "epoch": 2.1556743200071655,
      "grad_norm": 3.442650079727173,
      "learning_rate": 0.00017937302498871423,
      "loss": 1.2774,
      "step": 72200
    },
    {
      "epoch": 2.1571671692592482,
      "grad_norm": 5.232537746429443,
      "learning_rate": 0.0001793587391927954,
      "loss": 1.3006,
      "step": 72250
    },
    {
      "epoch": 2.1586600185113305,
      "grad_norm": 4.750761985778809,
      "learning_rate": 0.00017934445339687656,
      "loss": 1.2696,
      "step": 72300
    },
    {
      "epoch": 2.160152867763413,
      "grad_norm": 5.479654788970947,
      "learning_rate": 0.00017933016760095772,
      "loss": 1.2896,
      "step": 72350
    },
    {
      "epoch": 2.161645717015496,
      "grad_norm": 4.499126434326172,
      "learning_rate": 0.0001793158818050389,
      "loss": 1.2962,
      "step": 72400
    },
    {
      "epoch": 2.163138566267578,
      "grad_norm": 5.156326770782471,
      "learning_rate": 0.00017930159600912008,
      "loss": 1.2852,
      "step": 72450
    },
    {
      "epoch": 2.164631415519661,
      "grad_norm": 4.520088195800781,
      "learning_rate": 0.00017928731021320122,
      "loss": 1.2527,
      "step": 72500
    },
    {
      "epoch": 2.166124264771743,
      "grad_norm": 4.284939289093018,
      "learning_rate": 0.0001792730244172824,
      "loss": 1.3213,
      "step": 72550
    },
    {
      "epoch": 2.167617114023826,
      "grad_norm": 4.373524188995361,
      "learning_rate": 0.00017925873862136355,
      "loss": 1.2853,
      "step": 72600
    },
    {
      "epoch": 2.1691099632759085,
      "grad_norm": 3.9682133197784424,
      "learning_rate": 0.00017924445282544474,
      "loss": 1.2707,
      "step": 72650
    },
    {
      "epoch": 2.170602812527991,
      "grad_norm": 5.218033313751221,
      "learning_rate": 0.00017923016702952587,
      "loss": 1.2492,
      "step": 72700
    },
    {
      "epoch": 2.1720956617800735,
      "grad_norm": 4.900534152984619,
      "learning_rate": 0.00017921588123360707,
      "loss": 1.256,
      "step": 72750
    },
    {
      "epoch": 2.173588511032156,
      "grad_norm": 4.519747257232666,
      "learning_rate": 0.00017920159543768823,
      "loss": 1.2367,
      "step": 72800
    },
    {
      "epoch": 2.1750813602842385,
      "grad_norm": 4.465813159942627,
      "learning_rate": 0.0001791873096417694,
      "loss": 1.286,
      "step": 72850
    },
    {
      "epoch": 2.176574209536321,
      "grad_norm": 6.142160892486572,
      "learning_rate": 0.00017917302384585056,
      "loss": 1.2959,
      "step": 72900
    },
    {
      "epoch": 2.1780670587884035,
      "grad_norm": 3.851306200027466,
      "learning_rate": 0.00017915873804993173,
      "loss": 1.2511,
      "step": 72950
    },
    {
      "epoch": 2.179559908040486,
      "grad_norm": 5.0324788093566895,
      "learning_rate": 0.0001791444522540129,
      "loss": 1.2652,
      "step": 73000
    },
    {
      "epoch": 2.1810527572925684,
      "grad_norm": 4.664618492126465,
      "learning_rate": 0.00017913016645809405,
      "loss": 1.2639,
      "step": 73050
    },
    {
      "epoch": 2.182545606544651,
      "grad_norm": 4.418179988861084,
      "learning_rate": 0.00017911588066217522,
      "loss": 1.2948,
      "step": 73100
    },
    {
      "epoch": 2.184038455796734,
      "grad_norm": 4.895113945007324,
      "learning_rate": 0.00017910159486625638,
      "loss": 1.2283,
      "step": 73150
    },
    {
      "epoch": 2.185531305048816,
      "grad_norm": 4.2365946769714355,
      "learning_rate": 0.00017908730907033755,
      "loss": 1.2772,
      "step": 73200
    },
    {
      "epoch": 2.187024154300899,
      "grad_norm": 6.705420017242432,
      "learning_rate": 0.00017907302327441874,
      "loss": 1.273,
      "step": 73250
    },
    {
      "epoch": 2.188517003552981,
      "grad_norm": 5.094226360321045,
      "learning_rate": 0.00017905873747849988,
      "loss": 1.2391,
      "step": 73300
    },
    {
      "epoch": 2.1900098528050638,
      "grad_norm": 4.424404621124268,
      "learning_rate": 0.00017904445168258107,
      "loss": 1.2198,
      "step": 73350
    },
    {
      "epoch": 2.1915027020571465,
      "grad_norm": 5.0152201652526855,
      "learning_rate": 0.0001790301658866622,
      "loss": 1.3167,
      "step": 73400
    },
    {
      "epoch": 2.1929955513092287,
      "grad_norm": 4.916745662689209,
      "learning_rate": 0.0001790158800907434,
      "loss": 1.2482,
      "step": 73450
    },
    {
      "epoch": 2.1944884005613114,
      "grad_norm": 4.396121978759766,
      "learning_rate": 0.00017900159429482454,
      "loss": 1.2425,
      "step": 73500
    },
    {
      "epoch": 2.1959812498133937,
      "grad_norm": 4.289492607116699,
      "learning_rate": 0.00017898730849890573,
      "loss": 1.2647,
      "step": 73550
    },
    {
      "epoch": 2.1974740990654764,
      "grad_norm": 6.336133003234863,
      "learning_rate": 0.0001789730227029869,
      "loss": 1.2513,
      "step": 73600
    },
    {
      "epoch": 2.198966948317559,
      "grad_norm": 4.202422618865967,
      "learning_rate": 0.00017895873690706806,
      "loss": 1.2657,
      "step": 73650
    },
    {
      "epoch": 2.2004597975696414,
      "grad_norm": 4.372437953948975,
      "learning_rate": 0.00017894445111114922,
      "loss": 1.2432,
      "step": 73700
    },
    {
      "epoch": 2.201952646821724,
      "grad_norm": 5.008980751037598,
      "learning_rate": 0.0001789301653152304,
      "loss": 1.2848,
      "step": 73750
    },
    {
      "epoch": 2.2034454960738064,
      "grad_norm": 5.251125812530518,
      "learning_rate": 0.00017891587951931155,
      "loss": 1.3296,
      "step": 73800
    },
    {
      "epoch": 2.204938345325889,
      "grad_norm": 3.898746967315674,
      "learning_rate": 0.00017890159372339272,
      "loss": 1.2639,
      "step": 73850
    },
    {
      "epoch": 2.2064311945779713,
      "grad_norm": 5.178562164306641,
      "learning_rate": 0.00017888730792747388,
      "loss": 1.3065,
      "step": 73900
    },
    {
      "epoch": 2.207924043830054,
      "grad_norm": 5.667030334472656,
      "learning_rate": 0.00017887302213155505,
      "loss": 1.2954,
      "step": 73950
    },
    {
      "epoch": 2.2094168930821367,
      "grad_norm": 7.237538814544678,
      "learning_rate": 0.0001788587363356362,
      "loss": 1.2558,
      "step": 74000
    },
    {
      "epoch": 2.210909742334219,
      "grad_norm": 4.029843330383301,
      "learning_rate": 0.00017884445053971737,
      "loss": 1.3004,
      "step": 74050
    },
    {
      "epoch": 2.2124025915863017,
      "grad_norm": 5.686726093292236,
      "learning_rate": 0.00017883016474379854,
      "loss": 1.278,
      "step": 74100
    },
    {
      "epoch": 2.213895440838384,
      "grad_norm": 5.373737812042236,
      "learning_rate": 0.0001788158789478797,
      "loss": 1.3089,
      "step": 74150
    },
    {
      "epoch": 2.2153882900904667,
      "grad_norm": 4.493228912353516,
      "learning_rate": 0.00017880159315196087,
      "loss": 1.2856,
      "step": 74200
    },
    {
      "epoch": 2.2168811393425494,
      "grad_norm": 5.083597183227539,
      "learning_rate": 0.00017878730735604203,
      "loss": 1.2792,
      "step": 74250
    },
    {
      "epoch": 2.2183739885946316,
      "grad_norm": 5.1655144691467285,
      "learning_rate": 0.0001787730215601232,
      "loss": 1.2852,
      "step": 74300
    },
    {
      "epoch": 2.2198668378467143,
      "grad_norm": 3.2413179874420166,
      "learning_rate": 0.00017875873576420436,
      "loss": 1.2617,
      "step": 74350
    },
    {
      "epoch": 2.2213596870987966,
      "grad_norm": 4.30172061920166,
      "learning_rate": 0.00017874444996828555,
      "loss": 1.2949,
      "step": 74400
    },
    {
      "epoch": 2.2228525363508793,
      "grad_norm": 4.27219295501709,
      "learning_rate": 0.0001787301641723667,
      "loss": 1.2293,
      "step": 74450
    },
    {
      "epoch": 2.2243453856029616,
      "grad_norm": 5.430578708648682,
      "learning_rate": 0.00017871587837644788,
      "loss": 1.2662,
      "step": 74500
    },
    {
      "epoch": 2.2258382348550443,
      "grad_norm": 5.0630292892456055,
      "learning_rate": 0.00017870159258052902,
      "loss": 1.308,
      "step": 74550
    },
    {
      "epoch": 2.227331084107127,
      "grad_norm": 4.326575756072998,
      "learning_rate": 0.0001786873067846102,
      "loss": 1.3244,
      "step": 74600
    },
    {
      "epoch": 2.2288239333592093,
      "grad_norm": 4.2859206199646,
      "learning_rate": 0.00017867302098869138,
      "loss": 1.2879,
      "step": 74650
    },
    {
      "epoch": 2.230316782611292,
      "grad_norm": 3.931136131286621,
      "learning_rate": 0.00017865873519277254,
      "loss": 1.3046,
      "step": 74700
    },
    {
      "epoch": 2.231809631863374,
      "grad_norm": 3.9645185470581055,
      "learning_rate": 0.0001786444493968537,
      "loss": 1.2712,
      "step": 74750
    },
    {
      "epoch": 2.233302481115457,
      "grad_norm": 5.313785076141357,
      "learning_rate": 0.00017863016360093487,
      "loss": 1.2514,
      "step": 74800
    },
    {
      "epoch": 2.2347953303675396,
      "grad_norm": 4.994843482971191,
      "learning_rate": 0.00017861587780501604,
      "loss": 1.2803,
      "step": 74850
    },
    {
      "epoch": 2.236288179619622,
      "grad_norm": 6.394142150878906,
      "learning_rate": 0.0001786015920090972,
      "loss": 1.2424,
      "step": 74900
    },
    {
      "epoch": 2.2377810288717046,
      "grad_norm": 4.67982816696167,
      "learning_rate": 0.00017858730621317837,
      "loss": 1.2775,
      "step": 74950
    },
    {
      "epoch": 2.239273878123787,
      "grad_norm": 5.697408199310303,
      "learning_rate": 0.00017857302041725953,
      "loss": 1.2665,
      "step": 75000
    },
    {
      "epoch": 2.2407667273758696,
      "grad_norm": 6.522719383239746,
      "learning_rate": 0.0001785587346213407,
      "loss": 1.3017,
      "step": 75050
    },
    {
      "epoch": 2.2422595766279523,
      "grad_norm": 3.7727861404418945,
      "learning_rate": 0.00017854444882542186,
      "loss": 1.3516,
      "step": 75100
    },
    {
      "epoch": 2.2437524258800345,
      "grad_norm": 4.810925483703613,
      "learning_rate": 0.00017853016302950302,
      "loss": 1.277,
      "step": 75150
    },
    {
      "epoch": 2.2452452751321172,
      "grad_norm": 3.6915650367736816,
      "learning_rate": 0.00017851587723358422,
      "loss": 1.3159,
      "step": 75200
    },
    {
      "epoch": 2.2467381243841995,
      "grad_norm": 4.427685260772705,
      "learning_rate": 0.00017850159143766535,
      "loss": 1.2725,
      "step": 75250
    },
    {
      "epoch": 2.248230973636282,
      "grad_norm": 7.222316265106201,
      "learning_rate": 0.00017848730564174655,
      "loss": 1.3272,
      "step": 75300
    },
    {
      "epoch": 2.249723822888365,
      "grad_norm": 4.113325595855713,
      "learning_rate": 0.00017847301984582768,
      "loss": 1.2861,
      "step": 75350
    },
    {
      "epoch": 2.251216672140447,
      "grad_norm": 3.8266818523406982,
      "learning_rate": 0.00017845873404990887,
      "loss": 1.2688,
      "step": 75400
    },
    {
      "epoch": 2.25270952139253,
      "grad_norm": 5.281492710113525,
      "learning_rate": 0.00017844444825399004,
      "loss": 1.2743,
      "step": 75450
    },
    {
      "epoch": 2.254202370644612,
      "grad_norm": 5.222156047821045,
      "learning_rate": 0.0001784301624580712,
      "loss": 1.2324,
      "step": 75500
    },
    {
      "epoch": 2.255695219896695,
      "grad_norm": 4.961234092712402,
      "learning_rate": 0.00017841587666215237,
      "loss": 1.3248,
      "step": 75550
    },
    {
      "epoch": 2.2571880691487776,
      "grad_norm": 5.501556396484375,
      "learning_rate": 0.00017840159086623353,
      "loss": 1.2868,
      "step": 75600
    },
    {
      "epoch": 2.25868091840086,
      "grad_norm": 4.440338611602783,
      "learning_rate": 0.0001783873050703147,
      "loss": 1.237,
      "step": 75650
    },
    {
      "epoch": 2.2601737676529425,
      "grad_norm": 7.373712539672852,
      "learning_rate": 0.00017837301927439586,
      "loss": 1.2931,
      "step": 75700
    },
    {
      "epoch": 2.261666616905025,
      "grad_norm": 4.324766635894775,
      "learning_rate": 0.00017835873347847703,
      "loss": 1.2762,
      "step": 75750
    },
    {
      "epoch": 2.2631594661571075,
      "grad_norm": 7.707190990447998,
      "learning_rate": 0.0001783444476825582,
      "loss": 1.2565,
      "step": 75800
    },
    {
      "epoch": 2.26465231540919,
      "grad_norm": 3.9111223220825195,
      "learning_rate": 0.00017833016188663936,
      "loss": 1.2134,
      "step": 75850
    },
    {
      "epoch": 2.2661451646612725,
      "grad_norm": 4.735013961791992,
      "learning_rate": 0.00017831587609072055,
      "loss": 1.1908,
      "step": 75900
    },
    {
      "epoch": 2.267638013913355,
      "grad_norm": 5.743906021118164,
      "learning_rate": 0.00017830159029480169,
      "loss": 1.294,
      "step": 75950
    },
    {
      "epoch": 2.2691308631654374,
      "grad_norm": 3.792663097381592,
      "learning_rate": 0.00017828730449888288,
      "loss": 1.3725,
      "step": 76000
    },
    {
      "epoch": 2.27062371241752,
      "grad_norm": 4.77357292175293,
      "learning_rate": 0.00017827301870296402,
      "loss": 1.2511,
      "step": 76050
    },
    {
      "epoch": 2.272116561669603,
      "grad_norm": 5.1196675300598145,
      "learning_rate": 0.0001782587329070452,
      "loss": 1.3099,
      "step": 76100
    },
    {
      "epoch": 2.273609410921685,
      "grad_norm": 5.5518574714660645,
      "learning_rate": 0.00017824444711112634,
      "loss": 1.2841,
      "step": 76150
    },
    {
      "epoch": 2.275102260173768,
      "grad_norm": 4.075642108917236,
      "learning_rate": 0.00017823016131520754,
      "loss": 1.2857,
      "step": 76200
    },
    {
      "epoch": 2.27659510942585,
      "grad_norm": 6.330162048339844,
      "learning_rate": 0.0001782158755192887,
      "loss": 1.253,
      "step": 76250
    },
    {
      "epoch": 2.278087958677933,
      "grad_norm": 3.8504278659820557,
      "learning_rate": 0.00017820158972336987,
      "loss": 1.3167,
      "step": 76300
    },
    {
      "epoch": 2.279580807930015,
      "grad_norm": 5.074968338012695,
      "learning_rate": 0.00017818730392745103,
      "loss": 1.3109,
      "step": 76350
    },
    {
      "epoch": 2.2810736571820978,
      "grad_norm": 4.519435405731201,
      "learning_rate": 0.0001781730181315322,
      "loss": 1.2721,
      "step": 76400
    },
    {
      "epoch": 2.2825665064341805,
      "grad_norm": 5.285332202911377,
      "learning_rate": 0.00017815873233561336,
      "loss": 1.3091,
      "step": 76450
    },
    {
      "epoch": 2.2840593556862627,
      "grad_norm": 6.193671226501465,
      "learning_rate": 0.00017814444653969452,
      "loss": 1.2676,
      "step": 76500
    },
    {
      "epoch": 2.2855522049383454,
      "grad_norm": 4.4518961906433105,
      "learning_rate": 0.0001781301607437757,
      "loss": 1.2637,
      "step": 76550
    },
    {
      "epoch": 2.2870450541904277,
      "grad_norm": 5.584630966186523,
      "learning_rate": 0.00017811587494785685,
      "loss": 1.2906,
      "step": 76600
    },
    {
      "epoch": 2.2885379034425104,
      "grad_norm": 5.528042793273926,
      "learning_rate": 0.00017810158915193802,
      "loss": 1.2701,
      "step": 76650
    },
    {
      "epoch": 2.2900307526945927,
      "grad_norm": 5.087693214416504,
      "learning_rate": 0.00017808730335601918,
      "loss": 1.33,
      "step": 76700
    },
    {
      "epoch": 2.2915236019466754,
      "grad_norm": 4.278517723083496,
      "learning_rate": 0.00017807301756010035,
      "loss": 1.2854,
      "step": 76750
    },
    {
      "epoch": 2.293016451198758,
      "grad_norm": 4.385780334472656,
      "learning_rate": 0.0001780587317641815,
      "loss": 1.2567,
      "step": 76800
    },
    {
      "epoch": 2.2945093004508403,
      "grad_norm": 4.630320072174072,
      "learning_rate": 0.00017804444596826268,
      "loss": 1.2706,
      "step": 76850
    },
    {
      "epoch": 2.296002149702923,
      "grad_norm": 6.519657135009766,
      "learning_rate": 0.00017803016017234384,
      "loss": 1.2892,
      "step": 76900
    },
    {
      "epoch": 2.2974949989550053,
      "grad_norm": 5.163790702819824,
      "learning_rate": 0.000178015874376425,
      "loss": 1.3251,
      "step": 76950
    },
    {
      "epoch": 2.298987848207088,
      "grad_norm": 4.371760368347168,
      "learning_rate": 0.00017800158858050617,
      "loss": 1.3014,
      "step": 77000
    },
    {
      "epoch": 2.3004806974591707,
      "grad_norm": 4.700384616851807,
      "learning_rate": 0.00017798730278458736,
      "loss": 1.2648,
      "step": 77050
    },
    {
      "epoch": 2.301973546711253,
      "grad_norm": 4.602656364440918,
      "learning_rate": 0.0001779730169886685,
      "loss": 1.3043,
      "step": 77100
    },
    {
      "epoch": 2.3034663959633357,
      "grad_norm": 4.436239719390869,
      "learning_rate": 0.0001779587311927497,
      "loss": 1.2702,
      "step": 77150
    },
    {
      "epoch": 2.304959245215418,
      "grad_norm": 5.805758476257324,
      "learning_rate": 0.00017794444539683083,
      "loss": 1.2556,
      "step": 77200
    },
    {
      "epoch": 2.3064520944675007,
      "grad_norm": 5.682462215423584,
      "learning_rate": 0.00017793015960091202,
      "loss": 1.2422,
      "step": 77250
    },
    {
      "epoch": 2.3079449437195834,
      "grad_norm": 5.270015716552734,
      "learning_rate": 0.00017791587380499316,
      "loss": 1.2666,
      "step": 77300
    },
    {
      "epoch": 2.3094377929716656,
      "grad_norm": 5.771697998046875,
      "learning_rate": 0.00017790158800907435,
      "loss": 1.2735,
      "step": 77350
    },
    {
      "epoch": 2.3109306422237483,
      "grad_norm": 4.319996356964111,
      "learning_rate": 0.00017788730221315551,
      "loss": 1.2362,
      "step": 77400
    },
    {
      "epoch": 2.3124234914758306,
      "grad_norm": 7.646838188171387,
      "learning_rate": 0.00017787301641723668,
      "loss": 1.2998,
      "step": 77450
    },
    {
      "epoch": 2.3139163407279133,
      "grad_norm": 4.850773334503174,
      "learning_rate": 0.00017785873062131784,
      "loss": 1.2505,
      "step": 77500
    },
    {
      "epoch": 2.315409189979996,
      "grad_norm": 8.903724670410156,
      "learning_rate": 0.000177844444825399,
      "loss": 1.2483,
      "step": 77550
    },
    {
      "epoch": 2.3169020392320783,
      "grad_norm": 4.112529754638672,
      "learning_rate": 0.00017783015902948017,
      "loss": 1.2622,
      "step": 77600
    },
    {
      "epoch": 2.318394888484161,
      "grad_norm": 5.173572063446045,
      "learning_rate": 0.00017781587323356134,
      "loss": 1.2783,
      "step": 77650
    },
    {
      "epoch": 2.3198877377362432,
      "grad_norm": 3.9499335289001465,
      "learning_rate": 0.0001778015874376425,
      "loss": 1.3203,
      "step": 77700
    },
    {
      "epoch": 2.321380586988326,
      "grad_norm": 4.4116926193237305,
      "learning_rate": 0.00017778730164172367,
      "loss": 1.2819,
      "step": 77750
    },
    {
      "epoch": 2.3228734362404087,
      "grad_norm": 5.434197902679443,
      "learning_rate": 0.00017777301584580483,
      "loss": 1.2591,
      "step": 77800
    },
    {
      "epoch": 2.324366285492491,
      "grad_norm": 3.9472904205322266,
      "learning_rate": 0.00017775873004988602,
      "loss": 1.2609,
      "step": 77850
    },
    {
      "epoch": 2.3258591347445736,
      "grad_norm": 4.981082439422607,
      "learning_rate": 0.00017774444425396716,
      "loss": 1.3558,
      "step": 77900
    },
    {
      "epoch": 2.327351983996656,
      "grad_norm": 4.96250057220459,
      "learning_rate": 0.00017773015845804835,
      "loss": 1.2478,
      "step": 77950
    },
    {
      "epoch": 2.3288448332487386,
      "grad_norm": 7.093445301055908,
      "learning_rate": 0.0001777158726621295,
      "loss": 1.2659,
      "step": 78000
    },
    {
      "epoch": 2.3303376825008213,
      "grad_norm": 4.415938854217529,
      "learning_rate": 0.00017770158686621068,
      "loss": 1.2994,
      "step": 78050
    },
    {
      "epoch": 2.3318305317529036,
      "grad_norm": 6.983063697814941,
      "learning_rate": 0.00017768730107029185,
      "loss": 1.2994,
      "step": 78100
    },
    {
      "epoch": 2.3333233810049863,
      "grad_norm": 6.34420919418335,
      "learning_rate": 0.000177673015274373,
      "loss": 1.3172,
      "step": 78150
    },
    {
      "epoch": 2.3348162302570685,
      "grad_norm": 4.156967639923096,
      "learning_rate": 0.00017765872947845418,
      "loss": 1.2425,
      "step": 78200
    },
    {
      "epoch": 2.3363090795091512,
      "grad_norm": 4.001344680786133,
      "learning_rate": 0.00017764444368253534,
      "loss": 1.3034,
      "step": 78250
    },
    {
      "epoch": 2.337801928761234,
      "grad_norm": 5.008638381958008,
      "learning_rate": 0.0001776301578866165,
      "loss": 1.2912,
      "step": 78300
    },
    {
      "epoch": 2.339294778013316,
      "grad_norm": 3.917297840118408,
      "learning_rate": 0.00017761587209069767,
      "loss": 1.2807,
      "step": 78350
    },
    {
      "epoch": 2.340787627265399,
      "grad_norm": 4.353655815124512,
      "learning_rate": 0.00017760158629477884,
      "loss": 1.3237,
      "step": 78400
    },
    {
      "epoch": 2.342280476517481,
      "grad_norm": 5.5648698806762695,
      "learning_rate": 0.00017758730049886,
      "loss": 1.3145,
      "step": 78450
    },
    {
      "epoch": 2.343773325769564,
      "grad_norm": 3.928065538406372,
      "learning_rate": 0.00017757301470294116,
      "loss": 1.297,
      "step": 78500
    },
    {
      "epoch": 2.345266175021646,
      "grad_norm": 4.029613494873047,
      "learning_rate": 0.00017755872890702233,
      "loss": 1.2477,
      "step": 78550
    },
    {
      "epoch": 2.346759024273729,
      "grad_norm": 3.9108121395111084,
      "learning_rate": 0.0001775444431111035,
      "loss": 1.2537,
      "step": 78600
    },
    {
      "epoch": 2.3482518735258116,
      "grad_norm": 4.727762699127197,
      "learning_rate": 0.00017753015731518469,
      "loss": 1.2956,
      "step": 78650
    },
    {
      "epoch": 2.349744722777894,
      "grad_norm": 4.287507057189941,
      "learning_rate": 0.00017751587151926582,
      "loss": 1.2689,
      "step": 78700
    },
    {
      "epoch": 2.3512375720299765,
      "grad_norm": 4.162285804748535,
      "learning_rate": 0.00017750158572334701,
      "loss": 1.2803,
      "step": 78750
    },
    {
      "epoch": 2.352730421282059,
      "grad_norm": 5.458852291107178,
      "learning_rate": 0.00017748729992742815,
      "loss": 1.2789,
      "step": 78800
    },
    {
      "epoch": 2.3542232705341415,
      "grad_norm": 4.103569984436035,
      "learning_rate": 0.00017747301413150934,
      "loss": 1.3299,
      "step": 78850
    },
    {
      "epoch": 2.3557161197862238,
      "grad_norm": 4.286417007446289,
      "learning_rate": 0.0001774587283355905,
      "loss": 1.3333,
      "step": 78900
    },
    {
      "epoch": 2.3572089690383065,
      "grad_norm": 4.811729431152344,
      "learning_rate": 0.00017744444253967167,
      "loss": 1.3414,
      "step": 78950
    },
    {
      "epoch": 2.358701818290389,
      "grad_norm": 3.960171699523926,
      "learning_rate": 0.00017743015674375284,
      "loss": 1.2661,
      "step": 79000
    },
    {
      "epoch": 2.3601946675424714,
      "grad_norm": 5.161068439483643,
      "learning_rate": 0.000177415870947834,
      "loss": 1.2622,
      "step": 79050
    },
    {
      "epoch": 2.361687516794554,
      "grad_norm": 5.81963586807251,
      "learning_rate": 0.00017740158515191517,
      "loss": 1.3336,
      "step": 79100
    },
    {
      "epoch": 2.3631803660466364,
      "grad_norm": 5.815190315246582,
      "learning_rate": 0.00017738729935599633,
      "loss": 1.3078,
      "step": 79150
    },
    {
      "epoch": 2.364673215298719,
      "grad_norm": 3.994666814804077,
      "learning_rate": 0.0001773730135600775,
      "loss": 1.2237,
      "step": 79200
    },
    {
      "epoch": 2.366166064550802,
      "grad_norm": 5.232024669647217,
      "learning_rate": 0.00017735872776415866,
      "loss": 1.3267,
      "step": 79250
    },
    {
      "epoch": 2.367658913802884,
      "grad_norm": 4.517995357513428,
      "learning_rate": 0.00017734444196823983,
      "loss": 1.3168,
      "step": 79300
    },
    {
      "epoch": 2.3691517630549668,
      "grad_norm": 4.49736213684082,
      "learning_rate": 0.000177330156172321,
      "loss": 1.3403,
      "step": 79350
    },
    {
      "epoch": 2.370644612307049,
      "grad_norm": 5.295841693878174,
      "learning_rate": 0.00017731587037640216,
      "loss": 1.2639,
      "step": 79400
    },
    {
      "epoch": 2.3721374615591317,
      "grad_norm": 6.010612487792969,
      "learning_rate": 0.00017730158458048332,
      "loss": 1.273,
      "step": 79450
    },
    {
      "epoch": 2.3736303108112144,
      "grad_norm": 4.014612674713135,
      "learning_rate": 0.00017728729878456448,
      "loss": 1.3027,
      "step": 79500
    },
    {
      "epoch": 2.3751231600632967,
      "grad_norm": 5.472930908203125,
      "learning_rate": 0.00017727301298864565,
      "loss": 1.2927,
      "step": 79550
    },
    {
      "epoch": 2.3766160093153794,
      "grad_norm": 5.090315341949463,
      "learning_rate": 0.00017725872719272681,
      "loss": 1.2994,
      "step": 79600
    },
    {
      "epoch": 2.3781088585674617,
      "grad_norm": 4.583268165588379,
      "learning_rate": 0.00017724444139680798,
      "loss": 1.3249,
      "step": 79650
    },
    {
      "epoch": 2.3796017078195444,
      "grad_norm": 5.765147686004639,
      "learning_rate": 0.00017723015560088917,
      "loss": 1.2582,
      "step": 79700
    },
    {
      "epoch": 2.381094557071627,
      "grad_norm": 4.751801490783691,
      "learning_rate": 0.0001772158698049703,
      "loss": 1.2497,
      "step": 79750
    },
    {
      "epoch": 2.3825874063237094,
      "grad_norm": 6.316953182220459,
      "learning_rate": 0.0001772015840090515,
      "loss": 1.2949,
      "step": 79800
    },
    {
      "epoch": 2.384080255575792,
      "grad_norm": 4.6090779304504395,
      "learning_rate": 0.00017718729821313264,
      "loss": 1.2922,
      "step": 79850
    },
    {
      "epoch": 2.3855731048278743,
      "grad_norm": 5.004429340362549,
      "learning_rate": 0.00017717301241721383,
      "loss": 1.336,
      "step": 79900
    },
    {
      "epoch": 2.387065954079957,
      "grad_norm": 5.7317352294921875,
      "learning_rate": 0.00017715872662129497,
      "loss": 1.2296,
      "step": 79950
    },
    {
      "epoch": 2.3885588033320397,
      "grad_norm": 4.938337326049805,
      "learning_rate": 0.00017714444082537616,
      "loss": 1.3118,
      "step": 80000
    },
    {
      "epoch": 2.390051652584122,
      "grad_norm": 7.3811869621276855,
      "learning_rate": 0.00017713015502945732,
      "loss": 1.2989,
      "step": 80050
    },
    {
      "epoch": 2.3915445018362047,
      "grad_norm": 4.3925604820251465,
      "learning_rate": 0.0001771158692335385,
      "loss": 1.3183,
      "step": 80100
    },
    {
      "epoch": 2.393037351088287,
      "grad_norm": 3.382279396057129,
      "learning_rate": 0.00017710158343761965,
      "loss": 1.2749,
      "step": 80150
    },
    {
      "epoch": 2.3945302003403697,
      "grad_norm": 6.085289478302002,
      "learning_rate": 0.00017708729764170082,
      "loss": 1.2881,
      "step": 80200
    },
    {
      "epoch": 2.3960230495924524,
      "grad_norm": 3.96476149559021,
      "learning_rate": 0.00017707301184578198,
      "loss": 1.2656,
      "step": 80250
    },
    {
      "epoch": 2.3975158988445346,
      "grad_norm": 4.456584930419922,
      "learning_rate": 0.00017705872604986315,
      "loss": 1.2969,
      "step": 80300
    },
    {
      "epoch": 2.3990087480966173,
      "grad_norm": 5.09492301940918,
      "learning_rate": 0.0001770444402539443,
      "loss": 1.2679,
      "step": 80350
    },
    {
      "epoch": 2.4005015973486996,
      "grad_norm": 4.211208820343018,
      "learning_rate": 0.00017703015445802548,
      "loss": 1.2235,
      "step": 80400
    },
    {
      "epoch": 2.4019944466007823,
      "grad_norm": 4.683793067932129,
      "learning_rate": 0.00017701586866210664,
      "loss": 1.2791,
      "step": 80450
    },
    {
      "epoch": 2.403487295852865,
      "grad_norm": 4.003592014312744,
      "learning_rate": 0.00017700158286618783,
      "loss": 1.2498,
      "step": 80500
    },
    {
      "epoch": 2.4049801451049473,
      "grad_norm": 5.065978527069092,
      "learning_rate": 0.00017698729707026897,
      "loss": 1.2681,
      "step": 80550
    },
    {
      "epoch": 2.40647299435703,
      "grad_norm": 4.453160762786865,
      "learning_rate": 0.00017697301127435016,
      "loss": 1.3567,
      "step": 80600
    },
    {
      "epoch": 2.4079658436091123,
      "grad_norm": 4.852217197418213,
      "learning_rate": 0.0001769587254784313,
      "loss": 1.2808,
      "step": 80650
    },
    {
      "epoch": 2.409458692861195,
      "grad_norm": 4.629857063293457,
      "learning_rate": 0.0001769444396825125,
      "loss": 1.2994,
      "step": 80700
    },
    {
      "epoch": 2.4109515421132772,
      "grad_norm": 4.611933708190918,
      "learning_rate": 0.00017693015388659363,
      "loss": 1.2477,
      "step": 80750
    },
    {
      "epoch": 2.41244439136536,
      "grad_norm": 4.3425822257995605,
      "learning_rate": 0.00017691586809067482,
      "loss": 1.2861,
      "step": 80800
    },
    {
      "epoch": 2.4139372406174426,
      "grad_norm": 5.061252593994141,
      "learning_rate": 0.00017690158229475598,
      "loss": 1.345,
      "step": 80850
    },
    {
      "epoch": 2.415430089869525,
      "grad_norm": 5.230067253112793,
      "learning_rate": 0.00017688729649883715,
      "loss": 1.3097,
      "step": 80900
    },
    {
      "epoch": 2.4169229391216076,
      "grad_norm": 4.092568874359131,
      "learning_rate": 0.00017687301070291831,
      "loss": 1.2609,
      "step": 80950
    },
    {
      "epoch": 2.41841578837369,
      "grad_norm": 7.115472316741943,
      "learning_rate": 0.00017685872490699948,
      "loss": 1.3306,
      "step": 81000
    },
    {
      "epoch": 2.4199086376257726,
      "grad_norm": 4.142023086547852,
      "learning_rate": 0.00017684443911108064,
      "loss": 1.2715,
      "step": 81050
    },
    {
      "epoch": 2.421401486877855,
      "grad_norm": 4.5259575843811035,
      "learning_rate": 0.0001768301533151618,
      "loss": 1.3097,
      "step": 81100
    },
    {
      "epoch": 2.4228943361299375,
      "grad_norm": 4.749493598937988,
      "learning_rate": 0.00017681586751924297,
      "loss": 1.2413,
      "step": 81150
    },
    {
      "epoch": 2.4243871853820202,
      "grad_norm": 3.844097137451172,
      "learning_rate": 0.00017680158172332414,
      "loss": 1.3365,
      "step": 81200
    },
    {
      "epoch": 2.4258800346341025,
      "grad_norm": 4.504172325134277,
      "learning_rate": 0.0001767872959274053,
      "loss": 1.2889,
      "step": 81250
    },
    {
      "epoch": 2.427372883886185,
      "grad_norm": 4.036706447601318,
      "learning_rate": 0.0001767730101314865,
      "loss": 1.3041,
      "step": 81300
    },
    {
      "epoch": 2.4288657331382675,
      "grad_norm": 4.100889205932617,
      "learning_rate": 0.00017675872433556763,
      "loss": 1.2886,
      "step": 81350
    },
    {
      "epoch": 2.43035858239035,
      "grad_norm": 5.082368850708008,
      "learning_rate": 0.00017674443853964882,
      "loss": 1.333,
      "step": 81400
    },
    {
      "epoch": 2.431851431642433,
      "grad_norm": 4.644407749176025,
      "learning_rate": 0.00017673015274372996,
      "loss": 1.3191,
      "step": 81450
    },
    {
      "epoch": 2.433344280894515,
      "grad_norm": 4.445605754852295,
      "learning_rate": 0.00017671586694781115,
      "loss": 1.2961,
      "step": 81500
    },
    {
      "epoch": 2.434837130146598,
      "grad_norm": 4.809011936187744,
      "learning_rate": 0.00017670158115189232,
      "loss": 1.2496,
      "step": 81550
    },
    {
      "epoch": 2.43632997939868,
      "grad_norm": 5.294357776641846,
      "learning_rate": 0.00017668729535597348,
      "loss": 1.2885,
      "step": 81600
    },
    {
      "epoch": 2.437822828650763,
      "grad_norm": 4.103408336639404,
      "learning_rate": 0.00017667300956005465,
      "loss": 1.2612,
      "step": 81650
    },
    {
      "epoch": 2.4393156779028455,
      "grad_norm": 4.427671909332275,
      "learning_rate": 0.0001766587237641358,
      "loss": 1.2849,
      "step": 81700
    },
    {
      "epoch": 2.440808527154928,
      "grad_norm": 5.336637496948242,
      "learning_rate": 0.00017664443796821698,
      "loss": 1.2923,
      "step": 81750
    },
    {
      "epoch": 2.4423013764070105,
      "grad_norm": 5.235950946807861,
      "learning_rate": 0.00017663015217229814,
      "loss": 1.3086,
      "step": 81800
    },
    {
      "epoch": 2.4437942256590928,
      "grad_norm": 5.3591766357421875,
      "learning_rate": 0.0001766158663763793,
      "loss": 1.3326,
      "step": 81850
    },
    {
      "epoch": 2.4452870749111755,
      "grad_norm": 5.432180881500244,
      "learning_rate": 0.00017660158058046047,
      "loss": 1.2918,
      "step": 81900
    },
    {
      "epoch": 2.446779924163258,
      "grad_norm": 5.184196949005127,
      "learning_rate": 0.00017658729478454163,
      "loss": 1.2652,
      "step": 81950
    },
    {
      "epoch": 2.4482727734153404,
      "grad_norm": 4.962531566619873,
      "learning_rate": 0.0001765730089886228,
      "loss": 1.2579,
      "step": 82000
    },
    {
      "epoch": 2.449765622667423,
      "grad_norm": 3.7370457649230957,
      "learning_rate": 0.00017655872319270396,
      "loss": 1.3204,
      "step": 82050
    },
    {
      "epoch": 2.4512584719195054,
      "grad_norm": 4.21932315826416,
      "learning_rate": 0.00017654443739678513,
      "loss": 1.333,
      "step": 82100
    },
    {
      "epoch": 2.452751321171588,
      "grad_norm": 4.366473197937012,
      "learning_rate": 0.0001765301516008663,
      "loss": 1.2594,
      "step": 82150
    },
    {
      "epoch": 2.454244170423671,
      "grad_norm": 5.760409832000732,
      "learning_rate": 0.00017651586580494746,
      "loss": 1.3004,
      "step": 82200
    },
    {
      "epoch": 2.455737019675753,
      "grad_norm": 4.652368545532227,
      "learning_rate": 0.00017650158000902862,
      "loss": 1.3092,
      "step": 82250
    },
    {
      "epoch": 2.457229868927836,
      "grad_norm": 3.5619237422943115,
      "learning_rate": 0.0001764872942131098,
      "loss": 1.3315,
      "step": 82300
    },
    {
      "epoch": 2.458722718179918,
      "grad_norm": 3.285752534866333,
      "learning_rate": 0.00017647300841719098,
      "loss": 1.2906,
      "step": 82350
    },
    {
      "epoch": 2.4602155674320008,
      "grad_norm": 4.153928279876709,
      "learning_rate": 0.00017645872262127212,
      "loss": 1.3055,
      "step": 82400
    },
    {
      "epoch": 2.4617084166840835,
      "grad_norm": 3.767779588699341,
      "learning_rate": 0.0001764444368253533,
      "loss": 1.264,
      "step": 82450
    },
    {
      "epoch": 2.4632012659361657,
      "grad_norm": 4.4513983726501465,
      "learning_rate": 0.00017643015102943445,
      "loss": 1.2749,
      "step": 82500
    },
    {
      "epoch": 2.4646941151882484,
      "grad_norm": 4.600104808807373,
      "learning_rate": 0.00017641586523351564,
      "loss": 1.3194,
      "step": 82550
    },
    {
      "epoch": 2.4661869644403307,
      "grad_norm": 5.654660701751709,
      "learning_rate": 0.00017640157943759677,
      "loss": 1.2525,
      "step": 82600
    },
    {
      "epoch": 2.4676798136924134,
      "grad_norm": 5.852449893951416,
      "learning_rate": 0.00017638729364167797,
      "loss": 1.2609,
      "step": 82650
    },
    {
      "epoch": 2.469172662944496,
      "grad_norm": 4.8410139083862305,
      "learning_rate": 0.00017637300784575913,
      "loss": 1.2892,
      "step": 82700
    },
    {
      "epoch": 2.4706655121965784,
      "grad_norm": 4.150029182434082,
      "learning_rate": 0.0001763587220498403,
      "loss": 1.2893,
      "step": 82750
    },
    {
      "epoch": 2.472158361448661,
      "grad_norm": 4.355432033538818,
      "learning_rate": 0.00017634443625392146,
      "loss": 1.294,
      "step": 82800
    },
    {
      "epoch": 2.4736512107007433,
      "grad_norm": 4.689866542816162,
      "learning_rate": 0.00017633015045800262,
      "loss": 1.3327,
      "step": 82850
    },
    {
      "epoch": 2.475144059952826,
      "grad_norm": 5.386442184448242,
      "learning_rate": 0.0001763158646620838,
      "loss": 1.3027,
      "step": 82900
    },
    {
      "epoch": 2.4766369092049083,
      "grad_norm": 4.756774425506592,
      "learning_rate": 0.00017630157886616495,
      "loss": 1.2994,
      "step": 82950
    },
    {
      "epoch": 2.478129758456991,
      "grad_norm": 5.8331451416015625,
      "learning_rate": 0.00017628729307024612,
      "loss": 1.298,
      "step": 83000
    },
    {
      "epoch": 2.4796226077090737,
      "grad_norm": 4.216109275817871,
      "learning_rate": 0.00017627300727432728,
      "loss": 1.3288,
      "step": 83050
    },
    {
      "epoch": 2.481115456961156,
      "grad_norm": 5.123735427856445,
      "learning_rate": 0.00017625872147840845,
      "loss": 1.3103,
      "step": 83100
    },
    {
      "epoch": 2.4826083062132387,
      "grad_norm": 3.7272541522979736,
      "learning_rate": 0.00017624443568248964,
      "loss": 1.2753,
      "step": 83150
    },
    {
      "epoch": 2.484101155465321,
      "grad_norm": 5.656236171722412,
      "learning_rate": 0.00017623014988657078,
      "loss": 1.3013,
      "step": 83200
    },
    {
      "epoch": 2.4855940047174037,
      "grad_norm": 4.702991962432861,
      "learning_rate": 0.00017621586409065197,
      "loss": 1.3073,
      "step": 83250
    },
    {
      "epoch": 2.487086853969486,
      "grad_norm": 4.512374401092529,
      "learning_rate": 0.0001762015782947331,
      "loss": 1.2951,
      "step": 83300
    },
    {
      "epoch": 2.4885797032215686,
      "grad_norm": 3.560082197189331,
      "learning_rate": 0.0001761872924988143,
      "loss": 1.3083,
      "step": 83350
    },
    {
      "epoch": 2.4900725524736513,
      "grad_norm": 5.189292907714844,
      "learning_rate": 0.00017617300670289544,
      "loss": 1.3153,
      "step": 83400
    },
    {
      "epoch": 2.4915654017257336,
      "grad_norm": 4.624311447143555,
      "learning_rate": 0.00017615872090697663,
      "loss": 1.2606,
      "step": 83450
    },
    {
      "epoch": 2.4930582509778163,
      "grad_norm": 4.030190944671631,
      "learning_rate": 0.0001761444351110578,
      "loss": 1.2552,
      "step": 83500
    },
    {
      "epoch": 2.4945511002298986,
      "grad_norm": 5.9866042137146,
      "learning_rate": 0.00017613014931513896,
      "loss": 1.2818,
      "step": 83550
    },
    {
      "epoch": 2.4960439494819813,
      "grad_norm": 4.025645732879639,
      "learning_rate": 0.00017611586351922012,
      "loss": 1.325,
      "step": 83600
    },
    {
      "epoch": 2.497536798734064,
      "grad_norm": 5.12217903137207,
      "learning_rate": 0.0001761015777233013,
      "loss": 1.2656,
      "step": 83650
    },
    {
      "epoch": 2.4990296479861462,
      "grad_norm": 3.5725717544555664,
      "learning_rate": 0.00017608729192738245,
      "loss": 1.2605,
      "step": 83700
    },
    {
      "epoch": 2.500522497238229,
      "grad_norm": 6.810322284698486,
      "learning_rate": 0.00017607300613146362,
      "loss": 1.3275,
      "step": 83750
    },
    {
      "epoch": 2.502015346490311,
      "grad_norm": 6.723217487335205,
      "learning_rate": 0.00017605872033554478,
      "loss": 1.3038,
      "step": 83800
    },
    {
      "epoch": 2.503508195742394,
      "grad_norm": 5.044365882873535,
      "learning_rate": 0.00017604443453962595,
      "loss": 1.2821,
      "step": 83850
    },
    {
      "epoch": 2.5050010449944766,
      "grad_norm": 4.397944450378418,
      "learning_rate": 0.0001760301487437071,
      "loss": 1.3281,
      "step": 83900
    },
    {
      "epoch": 2.506493894246559,
      "grad_norm": 3.9087088108062744,
      "learning_rate": 0.0001760158629477883,
      "loss": 1.2674,
      "step": 83950
    },
    {
      "epoch": 2.5079867434986416,
      "grad_norm": 4.3125128746032715,
      "learning_rate": 0.00017600157715186944,
      "loss": 1.2654,
      "step": 84000
    },
    {
      "epoch": 2.509479592750724,
      "grad_norm": 5.792150020599365,
      "learning_rate": 0.00017598729135595063,
      "loss": 1.3846,
      "step": 84050
    },
    {
      "epoch": 2.5109724420028066,
      "grad_norm": 4.449563026428223,
      "learning_rate": 0.00017597300556003177,
      "loss": 1.2402,
      "step": 84100
    },
    {
      "epoch": 2.5124652912548893,
      "grad_norm": 3.286771774291992,
      "learning_rate": 0.00017595871976411296,
      "loss": 1.2547,
      "step": 84150
    },
    {
      "epoch": 2.5139581405069715,
      "grad_norm": 4.84338903427124,
      "learning_rate": 0.0001759444339681941,
      "loss": 1.2127,
      "step": 84200
    },
    {
      "epoch": 2.5154509897590542,
      "grad_norm": 4.913907051086426,
      "learning_rate": 0.0001759301481722753,
      "loss": 1.3453,
      "step": 84250
    },
    {
      "epoch": 2.5169438390111365,
      "grad_norm": 3.828268527984619,
      "learning_rate": 0.00017591586237635645,
      "loss": 1.3056,
      "step": 84300
    },
    {
      "epoch": 2.518436688263219,
      "grad_norm": 4.86277961730957,
      "learning_rate": 0.00017590157658043762,
      "loss": 1.2613,
      "step": 84350
    },
    {
      "epoch": 2.519929537515302,
      "grad_norm": 3.3477306365966797,
      "learning_rate": 0.00017588729078451878,
      "loss": 1.2784,
      "step": 84400
    },
    {
      "epoch": 2.521422386767384,
      "grad_norm": 3.982935905456543,
      "learning_rate": 0.00017587300498859995,
      "loss": 1.279,
      "step": 84450
    },
    {
      "epoch": 2.522915236019467,
      "grad_norm": 4.321762561798096,
      "learning_rate": 0.0001758587191926811,
      "loss": 1.2706,
      "step": 84500
    },
    {
      "epoch": 2.524408085271549,
      "grad_norm": 3.989497423171997,
      "learning_rate": 0.00017584443339676228,
      "loss": 1.2431,
      "step": 84550
    },
    {
      "epoch": 2.525900934523632,
      "grad_norm": 4.457726955413818,
      "learning_rate": 0.00017583014760084344,
      "loss": 1.2686,
      "step": 84600
    },
    {
      "epoch": 2.5273937837757146,
      "grad_norm": 4.100541114807129,
      "learning_rate": 0.0001758158618049246,
      "loss": 1.2379,
      "step": 84650
    },
    {
      "epoch": 2.528886633027797,
      "grad_norm": 5.386853218078613,
      "learning_rate": 0.00017580157600900577,
      "loss": 1.2767,
      "step": 84700
    },
    {
      "epoch": 2.5303794822798795,
      "grad_norm": 4.51826286315918,
      "learning_rate": 0.00017578729021308694,
      "loss": 1.2072,
      "step": 84750
    },
    {
      "epoch": 2.531872331531962,
      "grad_norm": 7.255303382873535,
      "learning_rate": 0.0001757730044171681,
      "loss": 1.2324,
      "step": 84800
    },
    {
      "epoch": 2.5333651807840445,
      "grad_norm": 4.7532572746276855,
      "learning_rate": 0.00017575871862124927,
      "loss": 1.2534,
      "step": 84850
    },
    {
      "epoch": 2.534858030036127,
      "grad_norm": 4.200450897216797,
      "learning_rate": 0.00017574443282533043,
      "loss": 1.2732,
      "step": 84900
    },
    {
      "epoch": 2.5363508792882095,
      "grad_norm": 4.062733173370361,
      "learning_rate": 0.0001757301470294116,
      "loss": 1.2987,
      "step": 84950
    },
    {
      "epoch": 2.537843728540292,
      "grad_norm": 4.786027908325195,
      "learning_rate": 0.00017571586123349279,
      "loss": 1.2593,
      "step": 85000
    },
    {
      "epoch": 2.5393365777923744,
      "grad_norm": 6.346192359924316,
      "learning_rate": 0.00017570157543757392,
      "loss": 1.2854,
      "step": 85050
    },
    {
      "epoch": 2.540829427044457,
      "grad_norm": 5.030992031097412,
      "learning_rate": 0.00017568728964165512,
      "loss": 1.2498,
      "step": 85100
    },
    {
      "epoch": 2.54232227629654,
      "grad_norm": 3.7949612140655518,
      "learning_rate": 0.00017567300384573625,
      "loss": 1.275,
      "step": 85150
    },
    {
      "epoch": 2.543815125548622,
      "grad_norm": 4.442857265472412,
      "learning_rate": 0.00017565871804981745,
      "loss": 1.2156,
      "step": 85200
    },
    {
      "epoch": 2.5453079748007044,
      "grad_norm": 5.143223285675049,
      "learning_rate": 0.00017564443225389858,
      "loss": 1.3554,
      "step": 85250
    },
    {
      "epoch": 2.546800824052787,
      "grad_norm": 4.240771770477295,
      "learning_rate": 0.00017563014645797977,
      "loss": 1.3288,
      "step": 85300
    },
    {
      "epoch": 2.5482936733048698,
      "grad_norm": 5.060192108154297,
      "learning_rate": 0.00017561586066206094,
      "loss": 1.3284,
      "step": 85350
    },
    {
      "epoch": 2.549786522556952,
      "grad_norm": 4.155733585357666,
      "learning_rate": 0.0001756015748661421,
      "loss": 1.2422,
      "step": 85400
    },
    {
      "epoch": 2.5512793718090347,
      "grad_norm": 5.4249749183654785,
      "learning_rate": 0.00017558728907022327,
      "loss": 1.262,
      "step": 85450
    },
    {
      "epoch": 2.552772221061117,
      "grad_norm": 5.704741954803467,
      "learning_rate": 0.00017557300327430443,
      "loss": 1.308,
      "step": 85500
    },
    {
      "epoch": 2.5542650703131997,
      "grad_norm": 4.013110637664795,
      "learning_rate": 0.0001755587174783856,
      "loss": 1.2667,
      "step": 85550
    },
    {
      "epoch": 2.5557579195652824,
      "grad_norm": 4.5984673500061035,
      "learning_rate": 0.00017554443168246676,
      "loss": 1.2756,
      "step": 85600
    },
    {
      "epoch": 2.5572507688173647,
      "grad_norm": 4.716318130493164,
      "learning_rate": 0.00017553014588654793,
      "loss": 1.2896,
      "step": 85650
    },
    {
      "epoch": 2.5587436180694474,
      "grad_norm": 3.9019134044647217,
      "learning_rate": 0.0001755158600906291,
      "loss": 1.3419,
      "step": 85700
    },
    {
      "epoch": 2.5602364673215297,
      "grad_norm": 4.762269020080566,
      "learning_rate": 0.00017550157429471026,
      "loss": 1.2856,
      "step": 85750
    },
    {
      "epoch": 2.5617293165736124,
      "grad_norm": 4.8936567306518555,
      "learning_rate": 0.00017548728849879145,
      "loss": 1.3033,
      "step": 85800
    },
    {
      "epoch": 2.563222165825695,
      "grad_norm": 5.605986595153809,
      "learning_rate": 0.00017547300270287259,
      "loss": 1.2928,
      "step": 85850
    },
    {
      "epoch": 2.5647150150777773,
      "grad_norm": 5.456994533538818,
      "learning_rate": 0.00017545871690695378,
      "loss": 1.2646,
      "step": 85900
    },
    {
      "epoch": 2.56620786432986,
      "grad_norm": 5.125797748565674,
      "learning_rate": 0.00017544443111103491,
      "loss": 1.3337,
      "step": 85950
    },
    {
      "epoch": 2.5677007135819423,
      "grad_norm": 6.698281288146973,
      "learning_rate": 0.0001754301453151161,
      "loss": 1.231,
      "step": 86000
    },
    {
      "epoch": 2.569193562834025,
      "grad_norm": 4.8110175132751465,
      "learning_rate": 0.00017541585951919724,
      "loss": 1.2829,
      "step": 86050
    },
    {
      "epoch": 2.5706864120861077,
      "grad_norm": 4.389436721801758,
      "learning_rate": 0.00017540157372327844,
      "loss": 1.3394,
      "step": 86100
    },
    {
      "epoch": 2.57217926133819,
      "grad_norm": 4.567478179931641,
      "learning_rate": 0.0001753872879273596,
      "loss": 1.3179,
      "step": 86150
    },
    {
      "epoch": 2.5736721105902727,
      "grad_norm": 6.0036444664001465,
      "learning_rate": 0.00017537300213144077,
      "loss": 1.2837,
      "step": 86200
    },
    {
      "epoch": 2.575164959842355,
      "grad_norm": 4.954610824584961,
      "learning_rate": 0.00017535871633552193,
      "loss": 1.3057,
      "step": 86250
    },
    {
      "epoch": 2.5766578090944376,
      "grad_norm": 3.623244524002075,
      "learning_rate": 0.0001753444305396031,
      "loss": 1.3203,
      "step": 86300
    },
    {
      "epoch": 2.5781506583465204,
      "grad_norm": 3.969407081604004,
      "learning_rate": 0.00017533014474368426,
      "loss": 1.2465,
      "step": 86350
    },
    {
      "epoch": 2.5796435075986026,
      "grad_norm": 4.700910568237305,
      "learning_rate": 0.00017531585894776542,
      "loss": 1.2673,
      "step": 86400
    },
    {
      "epoch": 2.5811363568506853,
      "grad_norm": 4.05939245223999,
      "learning_rate": 0.0001753015731518466,
      "loss": 1.2177,
      "step": 86450
    },
    {
      "epoch": 2.5826292061027676,
      "grad_norm": 4.144320011138916,
      "learning_rate": 0.00017528728735592775,
      "loss": 1.2198,
      "step": 86500
    },
    {
      "epoch": 2.5841220553548503,
      "grad_norm": 4.483496189117432,
      "learning_rate": 0.00017527300156000892,
      "loss": 1.3403,
      "step": 86550
    },
    {
      "epoch": 2.585614904606933,
      "grad_norm": 6.164941787719727,
      "learning_rate": 0.0001752587157640901,
      "loss": 1.3121,
      "step": 86600
    },
    {
      "epoch": 2.5871077538590153,
      "grad_norm": 5.265034198760986,
      "learning_rate": 0.00017524442996817125,
      "loss": 1.2484,
      "step": 86650
    },
    {
      "epoch": 2.588600603111098,
      "grad_norm": 4.718167304992676,
      "learning_rate": 0.00017523014417225244,
      "loss": 1.3256,
      "step": 86700
    },
    {
      "epoch": 2.5900934523631802,
      "grad_norm": 3.9518258571624756,
      "learning_rate": 0.00017521585837633358,
      "loss": 1.271,
      "step": 86750
    },
    {
      "epoch": 2.591586301615263,
      "grad_norm": 4.604781627655029,
      "learning_rate": 0.00017520157258041477,
      "loss": 1.2821,
      "step": 86800
    },
    {
      "epoch": 2.5930791508673456,
      "grad_norm": 3.5907840728759766,
      "learning_rate": 0.0001751872867844959,
      "loss": 1.2958,
      "step": 86850
    },
    {
      "epoch": 2.594572000119428,
      "grad_norm": 5.262489318847656,
      "learning_rate": 0.0001751730009885771,
      "loss": 1.3186,
      "step": 86900
    },
    {
      "epoch": 2.5960648493715106,
      "grad_norm": 4.812371253967285,
      "learning_rate": 0.00017515871519265826,
      "loss": 1.2586,
      "step": 86950
    },
    {
      "epoch": 2.597557698623593,
      "grad_norm": 4.645204544067383,
      "learning_rate": 0.00017514442939673943,
      "loss": 1.2656,
      "step": 87000
    },
    {
      "epoch": 2.5990505478756756,
      "grad_norm": 5.977488040924072,
      "learning_rate": 0.0001751301436008206,
      "loss": 1.3167,
      "step": 87050
    },
    {
      "epoch": 2.6005433971277583,
      "grad_norm": 4.534546375274658,
      "learning_rate": 0.00017511585780490176,
      "loss": 1.3263,
      "step": 87100
    },
    {
      "epoch": 2.6020362463798405,
      "grad_norm": 4.562127113342285,
      "learning_rate": 0.00017510157200898292,
      "loss": 1.3122,
      "step": 87150
    },
    {
      "epoch": 2.6035290956319233,
      "grad_norm": 3.588327646255493,
      "learning_rate": 0.00017508728621306409,
      "loss": 1.2953,
      "step": 87200
    },
    {
      "epoch": 2.6050219448840055,
      "grad_norm": 5.136357307434082,
      "learning_rate": 0.00017507300041714525,
      "loss": 1.2831,
      "step": 87250
    },
    {
      "epoch": 2.606514794136088,
      "grad_norm": 4.8900299072265625,
      "learning_rate": 0.00017505871462122641,
      "loss": 1.2847,
      "step": 87300
    },
    {
      "epoch": 2.608007643388171,
      "grad_norm": 3.7218592166900635,
      "learning_rate": 0.00017504442882530758,
      "loss": 1.2828,
      "step": 87350
    },
    {
      "epoch": 2.609500492640253,
      "grad_norm": 3.4648168087005615,
      "learning_rate": 0.00017503014302938874,
      "loss": 1.2576,
      "step": 87400
    },
    {
      "epoch": 2.6109933418923355,
      "grad_norm": 3.6955080032348633,
      "learning_rate": 0.0001750158572334699,
      "loss": 1.3208,
      "step": 87450
    },
    {
      "epoch": 2.612486191144418,
      "grad_norm": 4.296437740325928,
      "learning_rate": 0.00017500157143755107,
      "loss": 1.2455,
      "step": 87500
    },
    {
      "epoch": 2.613979040396501,
      "grad_norm": 4.3054518699646,
      "learning_rate": 0.00017498728564163224,
      "loss": 1.2722,
      "step": 87550
    },
    {
      "epoch": 2.615471889648583,
      "grad_norm": 6.258134841918945,
      "learning_rate": 0.0001749729998457134,
      "loss": 1.2777,
      "step": 87600
    },
    {
      "epoch": 2.616964738900666,
      "grad_norm": 5.092291355133057,
      "learning_rate": 0.0001749587140497946,
      "loss": 1.2906,
      "step": 87650
    },
    {
      "epoch": 2.618457588152748,
      "grad_norm": 3.986844778060913,
      "learning_rate": 0.00017494442825387573,
      "loss": 1.3326,
      "step": 87700
    },
    {
      "epoch": 2.619950437404831,
      "grad_norm": 5.269211292266846,
      "learning_rate": 0.00017493014245795692,
      "loss": 1.3253,
      "step": 87750
    },
    {
      "epoch": 2.6214432866569135,
      "grad_norm": 3.6939330101013184,
      "learning_rate": 0.00017491585666203806,
      "loss": 1.318,
      "step": 87800
    },
    {
      "epoch": 2.6229361359089958,
      "grad_norm": 5.028196334838867,
      "learning_rate": 0.00017490157086611925,
      "loss": 1.3089,
      "step": 87850
    },
    {
      "epoch": 2.6244289851610785,
      "grad_norm": 5.179739475250244,
      "learning_rate": 0.0001748872850702004,
      "loss": 1.2471,
      "step": 87900
    },
    {
      "epoch": 2.6259218344131607,
      "grad_norm": 4.581385135650635,
      "learning_rate": 0.00017487299927428158,
      "loss": 1.2703,
      "step": 87950
    },
    {
      "epoch": 2.6274146836652434,
      "grad_norm": 4.185626983642578,
      "learning_rate": 0.00017485871347836275,
      "loss": 1.2967,
      "step": 88000
    },
    {
      "epoch": 2.628907532917326,
      "grad_norm": 4.166527271270752,
      "learning_rate": 0.0001748444276824439,
      "loss": 1.282,
      "step": 88050
    },
    {
      "epoch": 2.6304003821694084,
      "grad_norm": 5.894084930419922,
      "learning_rate": 0.00017483014188652508,
      "loss": 1.3337,
      "step": 88100
    },
    {
      "epoch": 2.631893231421491,
      "grad_norm": 5.775843620300293,
      "learning_rate": 0.00017481585609060624,
      "loss": 1.3186,
      "step": 88150
    },
    {
      "epoch": 2.6333860806735734,
      "grad_norm": 4.601424217224121,
      "learning_rate": 0.0001748015702946874,
      "loss": 1.2961,
      "step": 88200
    },
    {
      "epoch": 2.634878929925656,
      "grad_norm": 4.345365524291992,
      "learning_rate": 0.00017478728449876857,
      "loss": 1.2776,
      "step": 88250
    },
    {
      "epoch": 2.636371779177739,
      "grad_norm": 4.751996040344238,
      "learning_rate": 0.00017477299870284973,
      "loss": 1.2847,
      "step": 88300
    },
    {
      "epoch": 2.637864628429821,
      "grad_norm": 4.37894344329834,
      "learning_rate": 0.0001747587129069309,
      "loss": 1.2631,
      "step": 88350
    },
    {
      "epoch": 2.6393574776819038,
      "grad_norm": 4.729623794555664,
      "learning_rate": 0.00017474442711101206,
      "loss": 1.3288,
      "step": 88400
    },
    {
      "epoch": 2.640850326933986,
      "grad_norm": 4.7951788902282715,
      "learning_rate": 0.00017473014131509326,
      "loss": 1.3207,
      "step": 88450
    },
    {
      "epoch": 2.6423431761860687,
      "grad_norm": 3.692270040512085,
      "learning_rate": 0.0001747158555191744,
      "loss": 1.3038,
      "step": 88500
    },
    {
      "epoch": 2.6438360254381514,
      "grad_norm": 4.1009297370910645,
      "learning_rate": 0.00017470156972325559,
      "loss": 1.2874,
      "step": 88550
    },
    {
      "epoch": 2.6453288746902337,
      "grad_norm": 4.919840335845947,
      "learning_rate": 0.00017468728392733672,
      "loss": 1.294,
      "step": 88600
    },
    {
      "epoch": 2.6468217239423164,
      "grad_norm": 3.7501938343048096,
      "learning_rate": 0.00017467299813141791,
      "loss": 1.3196,
      "step": 88650
    },
    {
      "epoch": 2.6483145731943987,
      "grad_norm": 4.218129634857178,
      "learning_rate": 0.00017465871233549905,
      "loss": 1.2823,
      "step": 88700
    },
    {
      "epoch": 2.6498074224464814,
      "grad_norm": 5.627344131469727,
      "learning_rate": 0.00017464442653958024,
      "loss": 1.2796,
      "step": 88750
    },
    {
      "epoch": 2.651300271698564,
      "grad_norm": 4.543562412261963,
      "learning_rate": 0.0001746301407436614,
      "loss": 1.2605,
      "step": 88800
    },
    {
      "epoch": 2.6527931209506463,
      "grad_norm": 4.705561637878418,
      "learning_rate": 0.00017461585494774257,
      "loss": 1.2822,
      "step": 88850
    },
    {
      "epoch": 2.654285970202729,
      "grad_norm": 4.356922149658203,
      "learning_rate": 0.00017460156915182374,
      "loss": 1.3247,
      "step": 88900
    },
    {
      "epoch": 2.6557788194548113,
      "grad_norm": 4.276957988739014,
      "learning_rate": 0.0001745872833559049,
      "loss": 1.3103,
      "step": 88950
    },
    {
      "epoch": 2.657271668706894,
      "grad_norm": 4.122916221618652,
      "learning_rate": 0.00017457299755998607,
      "loss": 1.302,
      "step": 89000
    },
    {
      "epoch": 2.6587645179589767,
      "grad_norm": 4.473240852355957,
      "learning_rate": 0.00017455871176406723,
      "loss": 1.274,
      "step": 89050
    },
    {
      "epoch": 2.660257367211059,
      "grad_norm": 5.299698829650879,
      "learning_rate": 0.0001745444259681484,
      "loss": 1.2879,
      "step": 89100
    },
    {
      "epoch": 2.6617502164631417,
      "grad_norm": 5.3960347175598145,
      "learning_rate": 0.00017453014017222956,
      "loss": 1.3083,
      "step": 89150
    },
    {
      "epoch": 2.663243065715224,
      "grad_norm": 9.475055694580078,
      "learning_rate": 0.00017451585437631073,
      "loss": 1.3123,
      "step": 89200
    },
    {
      "epoch": 2.6647359149673067,
      "grad_norm": 4.01664924621582,
      "learning_rate": 0.00017450156858039192,
      "loss": 1.2673,
      "step": 89250
    },
    {
      "epoch": 2.6662287642193894,
      "grad_norm": 4.74378776550293,
      "learning_rate": 0.00017448728278447306,
      "loss": 1.2704,
      "step": 89300
    },
    {
      "epoch": 2.6677216134714716,
      "grad_norm": 3.8527019023895264,
      "learning_rate": 0.00017447299698855425,
      "loss": 1.2638,
      "step": 89350
    },
    {
      "epoch": 2.6692144627235543,
      "grad_norm": 5.14741849899292,
      "learning_rate": 0.00017445871119263538,
      "loss": 1.2843,
      "step": 89400
    },
    {
      "epoch": 2.6707073119756366,
      "grad_norm": 4.814597129821777,
      "learning_rate": 0.00017444442539671658,
      "loss": 1.2955,
      "step": 89450
    },
    {
      "epoch": 2.6722001612277193,
      "grad_norm": 3.8238747119903564,
      "learning_rate": 0.00017443013960079771,
      "loss": 1.2921,
      "step": 89500
    },
    {
      "epoch": 2.673693010479802,
      "grad_norm": 5.118923187255859,
      "learning_rate": 0.0001744158538048789,
      "loss": 1.3114,
      "step": 89550
    },
    {
      "epoch": 2.6751858597318843,
      "grad_norm": 4.3184380531311035,
      "learning_rate": 0.00017440156800896007,
      "loss": 1.2387,
      "step": 89600
    },
    {
      "epoch": 2.6766787089839665,
      "grad_norm": 4.6351728439331055,
      "learning_rate": 0.00017438728221304123,
      "loss": 1.2524,
      "step": 89650
    },
    {
      "epoch": 2.6781715582360492,
      "grad_norm": 4.391300678253174,
      "learning_rate": 0.0001743729964171224,
      "loss": 1.2591,
      "step": 89700
    },
    {
      "epoch": 2.679664407488132,
      "grad_norm": 3.749375104904175,
      "learning_rate": 0.00017435871062120354,
      "loss": 1.247,
      "step": 89750
    },
    {
      "epoch": 2.681157256740214,
      "grad_norm": 4.260756492614746,
      "learning_rate": 0.00017434442482528473,
      "loss": 1.2672,
      "step": 89800
    },
    {
      "epoch": 2.682650105992297,
      "grad_norm": 3.6358819007873535,
      "learning_rate": 0.00017433013902936587,
      "loss": 1.2941,
      "step": 89850
    },
    {
      "epoch": 2.684142955244379,
      "grad_norm": 4.653028964996338,
      "learning_rate": 0.00017431585323344706,
      "loss": 1.3237,
      "step": 89900
    },
    {
      "epoch": 2.685635804496462,
      "grad_norm": 4.692080497741699,
      "learning_rate": 0.00017430156743752822,
      "loss": 1.325,
      "step": 89950
    },
    {
      "epoch": 2.6871286537485446,
      "grad_norm": 4.317993640899658,
      "learning_rate": 0.0001742872816416094,
      "loss": 1.2491,
      "step": 90000
    }
  ],
  "logging_steps": 50,
  "max_steps": 700001,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 21,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.274296357100454e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}