{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5001406645619081,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001000281329123816,
      "grad_norm": 5.35972785949707,
      "learning_rate": 0.0001999999450590425,
      "loss": 2.2302,
      "step": 1
    },
    {
      "epoch": 0.002000562658247632,
      "grad_norm": 2.222695827484131,
      "learning_rate": 0.00019999978023623033,
      "loss": 1.5115,
      "step": 2
    },
    {
      "epoch": 0.003000843987371448,
      "grad_norm": 2.513378858566284,
      "learning_rate": 0.0001999995055317446,
      "loss": 1.3975,
      "step": 3
    },
    {
      "epoch": 0.004001125316495264,
      "grad_norm": 1.3351486921310425,
      "learning_rate": 0.00019999912094588717,
      "loss": 1.1971,
      "step": 4
    },
    {
      "epoch": 0.005001406645619081,
      "grad_norm": 2.562424659729004,
      "learning_rate": 0.00019999862647908064,
      "loss": 1.7113,
      "step": 5
    },
    {
      "epoch": 0.006001687974742896,
      "grad_norm": 2.8064351081848145,
      "learning_rate": 0.00019999802213186834,
      "loss": 1.9351,
      "step": 6
    },
    {
      "epoch": 0.007001969303866712,
      "grad_norm": 1.640201449394226,
      "learning_rate": 0.0001999973079049143,
      "loss": 1.1837,
      "step": 7
    },
    {
      "epoch": 0.008002250632990529,
      "grad_norm": 0.86270672082901,
      "learning_rate": 0.00019999648379900338,
      "loss": 0.9277,
      "step": 8
    },
    {
      "epoch": 0.009002531962114344,
      "grad_norm": 0.9115725755691528,
      "learning_rate": 0.0001999955498150411,
      "loss": 1.0205,
      "step": 9
    },
    {
      "epoch": 0.010002813291238161,
      "grad_norm": 0.8274697065353394,
      "learning_rate": 0.00019999450595405374,
      "loss": 0.959,
      "step": 10
    },
    {
      "epoch": 0.011003094620361977,
      "grad_norm": 0.8405666947364807,
      "learning_rate": 0.0001999933522171883,
      "loss": 0.7265,
      "step": 11
    },
    {
      "epoch": 0.012003375949485792,
      "grad_norm": 0.6411307454109192,
      "learning_rate": 0.00019999208860571255,
      "loss": 0.7987,
      "step": 12
    },
    {
      "epoch": 0.01300365727860961,
      "grad_norm": 0.5139630436897278,
      "learning_rate": 0.00019999071512101496,
      "loss": 0.7672,
      "step": 13
    },
    {
      "epoch": 0.014003938607733425,
      "grad_norm": 0.5102379322052002,
      "learning_rate": 0.00019998923176460474,
      "loss": 0.7856,
      "step": 14
    },
    {
      "epoch": 0.01500421993685724,
      "grad_norm": 0.41639742255210876,
      "learning_rate": 0.00019998763853811184,
      "loss": 0.6591,
      "step": 15
    },
    {
      "epoch": 0.016004501265981057,
      "grad_norm": 0.5631232857704163,
      "learning_rate": 0.00019998593544328692,
      "loss": 0.6823,
      "step": 16
    },
    {
      "epoch": 0.017004782595104875,
      "grad_norm": 0.601658046245575,
      "learning_rate": 0.00019998412248200138,
      "loss": 0.8059,
      "step": 17
    },
    {
      "epoch": 0.01800506392422869,
      "grad_norm": 0.5272611379623413,
      "learning_rate": 0.00019998219965624734,
      "loss": 0.7871,
      "step": 18
    },
    {
      "epoch": 0.019005345253352506,
      "grad_norm": 0.4801584780216217,
      "learning_rate": 0.0001999801669681376,
      "loss": 0.69,
      "step": 19
    },
    {
      "epoch": 0.020005626582476323,
      "grad_norm": 0.5749354362487793,
      "learning_rate": 0.00019997802441990573,
      "loss": 0.669,
      "step": 20
    },
    {
      "epoch": 0.021005907911600136,
      "grad_norm": 0.5993685126304626,
      "learning_rate": 0.00019997577201390606,
      "loss": 0.7249,
      "step": 21
    },
    {
      "epoch": 0.022006189240723954,
      "grad_norm": 0.5108559727668762,
      "learning_rate": 0.00019997340975261353,
      "loss": 0.6593,
      "step": 22
    },
    {
      "epoch": 0.02300647056984777,
      "grad_norm": 0.5030386447906494,
      "learning_rate": 0.00019997093763862383,
      "loss": 0.6637,
      "step": 23
    },
    {
      "epoch": 0.024006751898971584,
      "grad_norm": 0.4953747093677521,
      "learning_rate": 0.0001999683556746534,
      "loss": 0.6865,
      "step": 24
    },
    {
      "epoch": 0.0250070332280954,
      "grad_norm": 0.44772374629974365,
      "learning_rate": 0.0001999656638635393,
      "loss": 0.6848,
      "step": 25
    },
    {
      "epoch": 0.02600731455721922,
      "grad_norm": 0.5174048542976379,
      "learning_rate": 0.0001999628622082394,
      "loss": 0.6173,
      "step": 26
    },
    {
      "epoch": 0.027007595886343033,
      "grad_norm": 0.525750994682312,
      "learning_rate": 0.0001999599507118322,
      "loss": 0.7309,
      "step": 27
    },
    {
      "epoch": 0.02800787721546685,
      "grad_norm": 0.5904054641723633,
      "learning_rate": 0.00019995692937751683,
      "loss": 0.7459,
      "step": 28
    },
    {
      "epoch": 0.029008158544590667,
      "grad_norm": 0.4662536084651947,
      "learning_rate": 0.0001999537982086133,
      "loss": 0.6309,
      "step": 29
    },
    {
      "epoch": 0.03000843987371448,
      "grad_norm": 0.5540624856948853,
      "learning_rate": 0.00019995055720856218,
      "loss": 0.6604,
      "step": 30
    },
    {
      "epoch": 0.031008721202838298,
      "grad_norm": 0.4457445442676544,
      "learning_rate": 0.00019994720638092468,
      "loss": 0.636,
      "step": 31
    },
    {
      "epoch": 0.032009002531962115,
      "grad_norm": 0.4571971893310547,
      "learning_rate": 0.00019994374572938277,
      "loss": 0.6506,
      "step": 32
    },
    {
      "epoch": 0.03300928386108593,
      "grad_norm": 0.5759263634681702,
      "learning_rate": 0.00019994017525773913,
      "loss": 0.5958,
      "step": 33
    },
    {
      "epoch": 0.03400956519020975,
      "grad_norm": 0.4859147071838379,
      "learning_rate": 0.00019993649496991705,
      "loss": 0.6042,
      "step": 34
    },
    {
      "epoch": 0.03500984651933356,
      "grad_norm": 0.3346123695373535,
      "learning_rate": 0.00019993270486996046,
      "loss": 0.5615,
      "step": 35
    },
    {
      "epoch": 0.03601012784845738,
      "grad_norm": 0.4086807668209076,
      "learning_rate": 0.000199928804962034,
      "loss": 0.5048,
      "step": 36
    },
    {
      "epoch": 0.037010409177581194,
      "grad_norm": 0.45720070600509644,
      "learning_rate": 0.00019992479525042303,
      "loss": 0.6318,
      "step": 37
    },
    {
      "epoch": 0.03801069050670501,
      "grad_norm": 0.3767400085926056,
      "learning_rate": 0.00019992067573953342,
      "loss": 0.6242,
      "step": 38
    },
    {
      "epoch": 0.03901097183582883,
      "grad_norm": 0.4942374527454376,
      "learning_rate": 0.0001999164464338918,
      "loss": 0.5696,
      "step": 39
    },
    {
      "epoch": 0.040011253164952645,
      "grad_norm": 0.6337136626243591,
      "learning_rate": 0.0001999121073381454,
      "loss": 0.6565,
      "step": 40
    },
    {
      "epoch": 0.041011534494076456,
      "grad_norm": 0.5405489802360535,
      "learning_rate": 0.0001999076584570621,
      "loss": 0.6096,
      "step": 41
    },
    {
      "epoch": 0.04201181582320027,
      "grad_norm": 0.7466319799423218,
      "learning_rate": 0.00019990309979553045,
      "loss": 0.5782,
      "step": 42
    },
    {
      "epoch": 0.04301209715232409,
      "grad_norm": 0.691272497177124,
      "learning_rate": 0.00019989843135855958,
      "loss": 0.622,
      "step": 43
    },
    {
      "epoch": 0.04401237848144791,
      "grad_norm": 0.563149094581604,
      "learning_rate": 0.00019989365315127922,
      "loss": 0.6198,
      "step": 44
    },
    {
      "epoch": 0.045012659810571724,
      "grad_norm": 0.47361648082733154,
      "learning_rate": 0.0001998887651789398,
      "loss": 0.5823,
      "step": 45
    },
    {
      "epoch": 0.04601294113969554,
      "grad_norm": 0.5451771020889282,
      "learning_rate": 0.0001998837674469123,
      "loss": 0.6963,
      "step": 46
    },
    {
      "epoch": 0.04701322246881936,
      "grad_norm": 0.4364232122898102,
      "learning_rate": 0.00019987865996068833,
      "loss": 0.5565,
      "step": 47
    },
    {
      "epoch": 0.04801350379794317,
      "grad_norm": 0.4986230134963989,
      "learning_rate": 0.00019987344272588006,
      "loss": 0.5707,
      "step": 48
    },
    {
      "epoch": 0.049013785127066986,
      "grad_norm": 0.46246567368507385,
      "learning_rate": 0.00019986811574822033,
      "loss": 0.5864,
      "step": 49
    },
    {
      "epoch": 0.0500140664561908,
      "grad_norm": 0.41016271710395813,
      "learning_rate": 0.00019986267903356254,
      "loss": 0.552,
      "step": 50
    },
    {
      "epoch": 0.05101434778531462,
      "grad_norm": 0.3554680049419403,
      "learning_rate": 0.0001998571325878806,
      "loss": 0.5582,
      "step": 51
    },
    {
      "epoch": 0.05201462911443844,
      "grad_norm": 0.4468945562839508,
      "learning_rate": 0.0001998514764172691,
      "loss": 0.5108,
      "step": 52
    },
    {
      "epoch": 0.053014910443562255,
      "grad_norm": 0.4482102692127228,
      "learning_rate": 0.00019984571052794313,
      "loss": 0.564,
      "step": 53
    },
    {
      "epoch": 0.054015191772686065,
      "grad_norm": 0.5015258193016052,
      "learning_rate": 0.00019983983492623833,
      "loss": 0.6323,
      "step": 54
    },
    {
      "epoch": 0.05501547310180988,
      "grad_norm": 0.4177694618701935,
      "learning_rate": 0.00019983384961861096,
      "loss": 0.565,
      "step": 55
    },
    {
      "epoch": 0.0560157544309337,
      "grad_norm": 0.3855839669704437,
      "learning_rate": 0.0001998277546116378,
      "loss": 0.4947,
      "step": 56
    },
    {
      "epoch": 0.05701603576005752,
      "grad_norm": 0.40609753131866455,
      "learning_rate": 0.00019982154991201608,
      "loss": 0.5428,
      "step": 57
    },
    {
      "epoch": 0.058016317089181334,
      "grad_norm": 0.41122737526893616,
      "learning_rate": 0.00019981523552656377,
      "loss": 0.5374,
      "step": 58
    },
    {
      "epoch": 0.05901659841830515,
      "grad_norm": 0.42212188243865967,
      "learning_rate": 0.00019980881146221914,
      "loss": 0.5192,
      "step": 59
    },
    {
      "epoch": 0.06001687974742896,
      "grad_norm": 0.40538090467453003,
      "learning_rate": 0.00019980227772604112,
      "loss": 0.568,
      "step": 60
    },
    {
      "epoch": 0.06101716107655278,
      "grad_norm": 0.3973814845085144,
      "learning_rate": 0.0001997956343252091,
      "loss": 0.5156,
      "step": 61
    },
    {
      "epoch": 0.062017442405676595,
      "grad_norm": 0.36684533953666687,
      "learning_rate": 0.00019978888126702296,
      "loss": 0.4978,
      "step": 62
    },
    {
      "epoch": 0.06301772373480041,
      "grad_norm": 0.45237308740615845,
      "learning_rate": 0.00019978201855890308,
      "loss": 0.5377,
      "step": 63
    },
    {
      "epoch": 0.06401800506392423,
      "grad_norm": 0.4670639634132385,
      "learning_rate": 0.00019977504620839035,
      "loss": 0.5208,
      "step": 64
    },
    {
      "epoch": 0.06501828639304805,
      "grad_norm": 0.4872264564037323,
      "learning_rate": 0.00019976796422314615,
      "loss": 0.5248,
      "step": 65
    },
    {
      "epoch": 0.06601856772217186,
      "grad_norm": 0.5238344073295593,
      "learning_rate": 0.00019976077261095226,
      "loss": 0.6493,
      "step": 66
    },
    {
      "epoch": 0.06701884905129568,
      "grad_norm": 0.6553749442100525,
      "learning_rate": 0.00019975347137971098,
      "loss": 0.5854,
      "step": 67
    },
    {
      "epoch": 0.0680191303804195,
      "grad_norm": 0.49785664677619934,
      "learning_rate": 0.00019974606053744503,
      "loss": 0.551,
      "step": 68
    },
    {
      "epoch": 0.06901941170954332,
      "grad_norm": 0.4122195839881897,
      "learning_rate": 0.00019973854009229763,
      "loss": 0.491,
      "step": 69
    },
    {
      "epoch": 0.07001969303866712,
      "grad_norm": 0.4016016721725464,
      "learning_rate": 0.00019973091005253232,
      "loss": 0.5082,
      "step": 70
    },
    {
      "epoch": 0.07101997436779094,
      "grad_norm": 0.3816310465335846,
      "learning_rate": 0.0001997231704265332,
      "loss": 0.5183,
      "step": 71
    },
    {
      "epoch": 0.07202025569691475,
      "grad_norm": 0.3194684684276581,
      "learning_rate": 0.00019971532122280464,
      "loss": 0.4754,
      "step": 72
    },
    {
      "epoch": 0.07302053702603857,
      "grad_norm": 0.33635130524635315,
      "learning_rate": 0.0001997073624499716,
      "loss": 0.5171,
      "step": 73
    },
    {
      "epoch": 0.07402081835516239,
      "grad_norm": 0.4141254723072052,
      "learning_rate": 0.0001996992941167792,
      "loss": 0.5794,
      "step": 74
    },
    {
      "epoch": 0.0750210996842862,
      "grad_norm": 0.381892591714859,
      "learning_rate": 0.00019969111623209323,
      "loss": 0.4623,
      "step": 75
    },
    {
      "epoch": 0.07602138101341002,
      "grad_norm": 0.5439775586128235,
      "learning_rate": 0.00019968282880489957,
      "loss": 0.5866,
      "step": 76
    },
    {
      "epoch": 0.07702166234253384,
      "grad_norm": 0.418949693441391,
      "learning_rate": 0.00019967443184430467,
      "loss": 0.4915,
      "step": 77
    },
    {
      "epoch": 0.07802194367165766,
      "grad_norm": 0.41112858057022095,
      "learning_rate": 0.0001996659253595353,
      "loss": 0.5748,
      "step": 78
    },
    {
      "epoch": 0.07902222500078147,
      "grad_norm": 0.44345736503601074,
      "learning_rate": 0.0001996573093599385,
      "loss": 0.5249,
      "step": 79
    },
    {
      "epoch": 0.08002250632990529,
      "grad_norm": 0.37021568417549133,
      "learning_rate": 0.00019964858385498172,
      "loss": 0.4544,
      "step": 80
    },
    {
      "epoch": 0.08102278765902911,
      "grad_norm": 0.38031843304634094,
      "learning_rate": 0.00019963974885425266,
      "loss": 0.4866,
      "step": 81
    },
    {
      "epoch": 0.08202306898815291,
      "grad_norm": 0.35023489594459534,
      "learning_rate": 0.00019963080436745945,
      "loss": 0.5048,
      "step": 82
    },
    {
      "epoch": 0.08302335031727673,
      "grad_norm": 0.4405301511287689,
      "learning_rate": 0.00019962175040443044,
      "loss": 0.4983,
      "step": 83
    },
    {
      "epoch": 0.08402363164640055,
      "grad_norm": 0.28912487626075745,
      "learning_rate": 0.0001996125869751143,
      "loss": 0.4442,
      "step": 84
    },
    {
      "epoch": 0.08502391297552436,
      "grad_norm": 0.3885224759578705,
      "learning_rate": 0.00019960331408957997,
      "loss": 0.4661,
      "step": 85
    },
    {
      "epoch": 0.08602419430464818,
      "grad_norm": 0.34638479351997375,
      "learning_rate": 0.00019959393175801671,
      "loss": 0.5045,
      "step": 86
    },
    {
      "epoch": 0.087024475633772,
      "grad_norm": 0.36377573013305664,
      "learning_rate": 0.00019958443999073397,
      "loss": 0.4827,
      "step": 87
    },
    {
      "epoch": 0.08802475696289581,
      "grad_norm": 0.3212088942527771,
      "learning_rate": 0.00019957483879816151,
      "loss": 0.5012,
      "step": 88
    },
    {
      "epoch": 0.08902503829201963,
      "grad_norm": 0.37304019927978516,
      "learning_rate": 0.00019956512819084928,
      "loss": 0.5184,
      "step": 89
    },
    {
      "epoch": 0.09002531962114345,
      "grad_norm": 0.4289352297782898,
      "learning_rate": 0.00019955530817946748,
      "loss": 0.4852,
      "step": 90
    },
    {
      "epoch": 0.09102560095026727,
      "grad_norm": 0.37521710991859436,
      "learning_rate": 0.00019954537877480655,
      "loss": 0.4492,
      "step": 91
    },
    {
      "epoch": 0.09202588227939108,
      "grad_norm": 0.40025514364242554,
      "learning_rate": 0.00019953533998777706,
      "loss": 0.4724,
      "step": 92
    },
    {
      "epoch": 0.0930261636085149,
      "grad_norm": 0.3497840464115143,
      "learning_rate": 0.00019952519182940993,
      "loss": 0.5133,
      "step": 93
    },
    {
      "epoch": 0.09402644493763872,
      "grad_norm": 0.3495926558971405,
      "learning_rate": 0.00019951493431085603,
      "loss": 0.4623,
      "step": 94
    },
    {
      "epoch": 0.09502672626676252,
      "grad_norm": 0.40398499369621277,
      "learning_rate": 0.00019950456744338658,
      "loss": 0.475,
      "step": 95
    },
    {
      "epoch": 0.09602700759588634,
      "grad_norm": 0.37891319394111633,
      "learning_rate": 0.00019949409123839288,
      "loss": 0.5273,
      "step": 96
    },
    {
      "epoch": 0.09702728892501016,
      "grad_norm": 0.37730953097343445,
      "learning_rate": 0.00019948350570738642,
      "loss": 0.5177,
      "step": 97
    },
    {
      "epoch": 0.09802757025413397,
      "grad_norm": 0.3501202464103699,
      "learning_rate": 0.0001994728108619987,
      "loss": 0.5032,
      "step": 98
    },
    {
      "epoch": 0.09902785158325779,
      "grad_norm": 0.35575389862060547,
      "learning_rate": 0.0001994620067139815,
      "loss": 0.4533,
      "step": 99
    },
    {
      "epoch": 0.1000281329123816,
      "grad_norm": 0.3112006187438965,
      "learning_rate": 0.00019945109327520658,
      "loss": 0.461,
      "step": 100
    },
    {
      "epoch": 0.10102841424150542,
      "grad_norm": 0.37702956795692444,
      "learning_rate": 0.00019944007055766586,
      "loss": 0.4929,
      "step": 101
    },
    {
      "epoch": 0.10202869557062924,
      "grad_norm": 0.3292798101902008,
      "learning_rate": 0.00019942893857347128,
      "loss": 0.5031,
      "step": 102
    },
    {
      "epoch": 0.10302897689975306,
      "grad_norm": 0.36839979887008667,
      "learning_rate": 0.00019941769733485494,
      "loss": 0.5204,
      "step": 103
    },
    {
      "epoch": 0.10402925822887688,
      "grad_norm": 0.4038218557834625,
      "learning_rate": 0.00019940634685416888,
      "loss": 0.5036,
      "step": 104
    },
    {
      "epoch": 0.10502953955800069,
      "grad_norm": 0.3477931022644043,
      "learning_rate": 0.00019939488714388524,
      "loss": 0.4547,
      "step": 105
    },
    {
      "epoch": 0.10602982088712451,
      "grad_norm": 0.3681749105453491,
      "learning_rate": 0.00019938331821659614,
      "loss": 0.4588,
      "step": 106
    },
    {
      "epoch": 0.10703010221624833,
      "grad_norm": 0.295214980840683,
      "learning_rate": 0.0001993716400850138,
      "loss": 0.4626,
      "step": 107
    },
    {
      "epoch": 0.10803038354537213,
      "grad_norm": 0.35674479603767395,
      "learning_rate": 0.0001993598527619703,
      "loss": 0.4726,
      "step": 108
    },
    {
      "epoch": 0.10903066487449595,
      "grad_norm": 0.3253282606601715,
      "learning_rate": 0.00019934795626041783,
      "loss": 0.4847,
      "step": 109
    },
    {
      "epoch": 0.11003094620361976,
      "grad_norm": 0.3049859404563904,
      "learning_rate": 0.0001993359505934285,
      "loss": 0.4037,
      "step": 110
    },
    {
      "epoch": 0.11103122753274358,
      "grad_norm": 0.3004118502140045,
      "learning_rate": 0.00019932383577419432,
      "loss": 0.4445,
      "step": 111
    },
    {
      "epoch": 0.1120315088618674,
      "grad_norm": 0.32134872674942017,
      "learning_rate": 0.0001993116118160273,
      "loss": 0.4729,
      "step": 112
    },
    {
      "epoch": 0.11303179019099122,
      "grad_norm": 0.4950062930583954,
      "learning_rate": 0.00019929927873235938,
      "loss": 0.5163,
      "step": 113
    },
    {
      "epoch": 0.11403207152011503,
      "grad_norm": 0.34567245841026306,
      "learning_rate": 0.00019928683653674237,
      "loss": 0.4865,
      "step": 114
    },
    {
      "epoch": 0.11503235284923885,
      "grad_norm": 0.4753624200820923,
      "learning_rate": 0.00019927428524284805,
      "loss": 0.5264,
      "step": 115
    },
    {
      "epoch": 0.11603263417836267,
      "grad_norm": 0.3215334117412567,
      "learning_rate": 0.00019926162486446792,
      "loss": 0.4876,
      "step": 116
    },
    {
      "epoch": 0.11703291550748648,
      "grad_norm": 0.3329266905784607,
      "learning_rate": 0.0001992488554155135,
      "loss": 0.4813,
      "step": 117
    },
    {
      "epoch": 0.1180331968366103,
      "grad_norm": 0.3493446707725525,
      "learning_rate": 0.00019923597691001615,
      "loss": 0.4675,
      "step": 118
    },
    {
      "epoch": 0.11903347816573412,
      "grad_norm": 0.336968332529068,
      "learning_rate": 0.0001992229893621269,
      "loss": 0.4712,
      "step": 119
    },
    {
      "epoch": 0.12003375949485792,
      "grad_norm": 0.32647812366485596,
      "learning_rate": 0.00019920989278611687,
      "loss": 0.475,
      "step": 120
    },
    {
      "epoch": 0.12103404082398174,
      "grad_norm": 0.36252933740615845,
      "learning_rate": 0.0001991966871963767,
      "loss": 0.5218,
      "step": 121
    },
    {
      "epoch": 0.12203432215310556,
      "grad_norm": 0.31632480025291443,
      "learning_rate": 0.000199183372607417,
      "loss": 0.4798,
      "step": 122
    },
    {
      "epoch": 0.12303460348222937,
      "grad_norm": 0.38675904273986816,
      "learning_rate": 0.0001991699490338681,
      "loss": 0.4993,
      "step": 123
    },
    {
      "epoch": 0.12403488481135319,
      "grad_norm": 0.3446194529533386,
      "learning_rate": 0.00019915641649048005,
      "loss": 0.4299,
      "step": 124
    },
    {
      "epoch": 0.12503516614047702,
      "grad_norm": 0.35425180196762085,
      "learning_rate": 0.0001991427749921227,
      "loss": 0.4235,
      "step": 125
    },
    {
      "epoch": 0.12603544746960083,
      "grad_norm": 0.3712017834186554,
      "learning_rate": 0.00019912902455378556,
      "loss": 0.4756,
      "step": 126
    },
    {
      "epoch": 0.12703572879872463,
      "grad_norm": 0.37233901023864746,
      "learning_rate": 0.00019911516519057788,
      "loss": 0.4519,
      "step": 127
    },
    {
      "epoch": 0.12803601012784846,
      "grad_norm": 0.3591810464859009,
      "learning_rate": 0.00019910119691772863,
      "loss": 0.4598,
      "step": 128
    },
    {
      "epoch": 0.12903629145697226,
      "grad_norm": 0.3153529465198517,
      "learning_rate": 0.00019908711975058637,
      "loss": 0.4534,
      "step": 129
    },
    {
      "epoch": 0.1300365727860961,
      "grad_norm": 0.2840005159378052,
      "learning_rate": 0.0001990729337046194,
      "loss": 0.3998,
      "step": 130
    },
    {
      "epoch": 0.1310368541152199,
      "grad_norm": 0.31850433349609375,
      "learning_rate": 0.0001990586387954156,
      "loss": 0.4448,
      "step": 131
    },
    {
      "epoch": 0.13203713544434373,
      "grad_norm": 0.35457924008369446,
      "learning_rate": 0.00019904423503868247,
      "loss": 0.4629,
      "step": 132
    },
    {
      "epoch": 0.13303741677346753,
      "grad_norm": 0.30500829219818115,
      "learning_rate": 0.00019902972245024715,
      "loss": 0.4671,
      "step": 133
    },
    {
      "epoch": 0.13403769810259136,
      "grad_norm": 0.28144538402557373,
      "learning_rate": 0.00019901510104605637,
      "loss": 0.4292,
      "step": 134
    },
    {
      "epoch": 0.13503797943171517,
      "grad_norm": 0.32505279779434204,
      "learning_rate": 0.00019900037084217637,
      "loss": 0.4862,
      "step": 135
    },
    {
      "epoch": 0.136038260760839,
      "grad_norm": 0.30766561627388,
      "learning_rate": 0.00019898553185479303,
      "loss": 0.4031,
      "step": 136
    },
    {
      "epoch": 0.1370385420899628,
      "grad_norm": 0.3485235273838043,
      "learning_rate": 0.00019897058410021167,
      "loss": 0.4298,
      "step": 137
    },
    {
      "epoch": 0.13803882341908663,
      "grad_norm": 0.3085039556026459,
      "learning_rate": 0.00019895552759485722,
      "loss": 0.4799,
      "step": 138
    },
    {
      "epoch": 0.13903910474821043,
      "grad_norm": 0.32276931405067444,
      "learning_rate": 0.00019894036235527395,
      "loss": 0.4488,
      "step": 139
    },
    {
      "epoch": 0.14003938607733424,
      "grad_norm": 0.3475525677204132,
      "learning_rate": 0.00019892508839812584,
      "loss": 0.4838,
      "step": 140
    },
    {
      "epoch": 0.14103966740645807,
      "grad_norm": 0.3139823079109192,
      "learning_rate": 0.00019890970574019617,
      "loss": 0.4428,
      "step": 141
    },
    {
      "epoch": 0.14203994873558187,
      "grad_norm": 0.3611343801021576,
      "learning_rate": 0.00019889421439838763,
      "loss": 0.4276,
      "step": 142
    },
    {
      "epoch": 0.1430402300647057,
      "grad_norm": 0.25845348834991455,
      "learning_rate": 0.00019887861438972246,
      "loss": 0.4118,
      "step": 143
    },
    {
      "epoch": 0.1440405113938295,
      "grad_norm": 0.3170836567878723,
      "learning_rate": 0.00019886290573134228,
      "loss": 0.4679,
      "step": 144
    },
    {
      "epoch": 0.14504079272295334,
      "grad_norm": 0.28241029381752014,
      "learning_rate": 0.000198847088440508,
      "loss": 0.415,
      "step": 145
    },
    {
      "epoch": 0.14604107405207714,
      "grad_norm": 0.3596460223197937,
      "learning_rate": 0.0001988311625346,
      "loss": 0.4619,
      "step": 146
    },
    {
      "epoch": 0.14704135538120097,
      "grad_norm": 0.2995036244392395,
      "learning_rate": 0.00019881512803111796,
      "loss": 0.4465,
      "step": 147
    },
    {
      "epoch": 0.14804163671032478,
      "grad_norm": 0.3156537413597107,
      "learning_rate": 0.00019879898494768093,
      "loss": 0.4638,
      "step": 148
    },
    {
      "epoch": 0.1490419180394486,
      "grad_norm": 0.39860132336616516,
      "learning_rate": 0.00019878273330202717,
      "loss": 0.4385,
      "step": 149
    },
    {
      "epoch": 0.1500421993685724,
      "grad_norm": 0.2727566659450531,
      "learning_rate": 0.00019876637311201433,
      "loss": 0.4114,
      "step": 150
    },
    {
      "epoch": 0.15104248069769624,
      "grad_norm": 0.2695155739784241,
      "learning_rate": 0.00019874990439561934,
      "loss": 0.4154,
      "step": 151
    },
    {
      "epoch": 0.15204276202682004,
      "grad_norm": 0.33344969153404236,
      "learning_rate": 0.0001987333271709383,
      "loss": 0.477,
      "step": 152
    },
    {
      "epoch": 0.15304304335594385,
      "grad_norm": 0.283865749835968,
      "learning_rate": 0.00019871664145618657,
      "loss": 0.4346,
      "step": 153
    },
    {
      "epoch": 0.15404332468506768,
      "grad_norm": 0.27019625902175903,
      "learning_rate": 0.00019869984726969878,
      "loss": 0.4061,
      "step": 154
    },
    {
      "epoch": 0.15504360601419148,
      "grad_norm": 0.2840537428855896,
      "learning_rate": 0.00019868294462992866,
      "loss": 0.4521,
      "step": 155
    },
    {
      "epoch": 0.1560438873433153,
      "grad_norm": 0.2851097583770752,
      "learning_rate": 0.00019866593355544922,
      "loss": 0.3906,
      "step": 156
    },
    {
      "epoch": 0.15704416867243912,
      "grad_norm": 0.3750353753566742,
      "learning_rate": 0.00019864881406495246,
      "loss": 0.475,
      "step": 157
    },
    {
      "epoch": 0.15804445000156295,
      "grad_norm": 0.32451707124710083,
      "learning_rate": 0.00019863158617724967,
      "loss": 0.4191,
      "step": 158
    },
    {
      "epoch": 0.15904473133068675,
      "grad_norm": 0.32818925380706787,
      "learning_rate": 0.00019861424991127115,
      "loss": 0.4004,
      "step": 159
    },
    {
      "epoch": 0.16004501265981058,
      "grad_norm": 0.3559636175632477,
      "learning_rate": 0.00019859680528606637,
      "loss": 0.5077,
      "step": 160
    },
    {
      "epoch": 0.16104529398893438,
      "grad_norm": 0.3427540063858032,
      "learning_rate": 0.00019857925232080373,
      "loss": 0.4617,
      "step": 161
    },
    {
      "epoch": 0.16204557531805822,
      "grad_norm": 0.3394433856010437,
      "learning_rate": 0.00019856159103477086,
      "loss": 0.4849,
      "step": 162
    },
    {
      "epoch": 0.16304585664718202,
      "grad_norm": 0.34075990319252014,
      "learning_rate": 0.00019854382144737426,
      "loss": 0.4486,
      "step": 163
    },
    {
      "epoch": 0.16404613797630582,
      "grad_norm": 0.3583144545555115,
      "learning_rate": 0.00019852594357813952,
      "loss": 0.4616,
      "step": 164
    },
    {
      "epoch": 0.16504641930542965,
      "grad_norm": 0.31794264912605286,
      "learning_rate": 0.00019850795744671116,
      "loss": 0.4964,
      "step": 165
    },
    {
      "epoch": 0.16604670063455346,
      "grad_norm": 0.3856930732727051,
      "learning_rate": 0.0001984898630728527,
      "loss": 0.4186,
      "step": 166
    },
    {
      "epoch": 0.1670469819636773,
      "grad_norm": 0.2840869724750519,
      "learning_rate": 0.0001984716604764466,
      "loss": 0.4166,
      "step": 167
    },
    {
      "epoch": 0.1680472632928011,
      "grad_norm": 0.2664308249950409,
      "learning_rate": 0.0001984533496774942,
      "loss": 0.3982,
      "step": 168
    },
    {
      "epoch": 0.16904754462192492,
      "grad_norm": 0.30538210272789,
      "learning_rate": 0.0001984349306961158,
      "loss": 0.4286,
      "step": 169
    },
    {
      "epoch": 0.17004782595104873,
      "grad_norm": 0.32451754808425903,
      "learning_rate": 0.00019841640355255043,
      "loss": 0.4558,
      "step": 170
    },
    {
      "epoch": 0.17104810728017256,
      "grad_norm": 0.29396554827690125,
      "learning_rate": 0.00019839776826715614,
      "loss": 0.3862,
      "step": 171
    },
    {
      "epoch": 0.17204838860929636,
      "grad_norm": 0.337141752243042,
      "learning_rate": 0.00019837902486040978,
      "loss": 0.4805,
      "step": 172
    },
    {
      "epoch": 0.1730486699384202,
      "grad_norm": 0.27084705233573914,
      "learning_rate": 0.0001983601733529069,
      "loss": 0.4354,
      "step": 173
    },
    {
      "epoch": 0.174048951267544,
      "grad_norm": 0.31350868940353394,
      "learning_rate": 0.00019834121376536187,
      "loss": 0.4606,
      "step": 174
    },
    {
      "epoch": 0.17504923259666783,
      "grad_norm": 0.33653292059898376,
      "learning_rate": 0.00019832214611860793,
      "loss": 0.5141,
      "step": 175
    },
    {
      "epoch": 0.17604951392579163,
      "grad_norm": 0.28151023387908936,
      "learning_rate": 0.00019830297043359692,
      "loss": 0.4386,
      "step": 176
    },
    {
      "epoch": 0.17704979525491543,
      "grad_norm": 0.3067042827606201,
      "learning_rate": 0.00019828368673139947,
      "loss": 0.4084,
      "step": 177
    },
    {
      "epoch": 0.17805007658403926,
      "grad_norm": 0.2893977463245392,
      "learning_rate": 0.0001982642950332049,
      "loss": 0.4395,
      "step": 178
    },
    {
      "epoch": 0.17905035791316307,
      "grad_norm": 0.2908482253551483,
      "learning_rate": 0.00019824479536032112,
      "loss": 0.3923,
      "step": 179
    },
    {
      "epoch": 0.1800506392422869,
      "grad_norm": 0.30002954602241516,
      "learning_rate": 0.0001982251877341748,
      "loss": 0.3975,
      "step": 180
    },
    {
      "epoch": 0.1810509205714107,
      "grad_norm": 0.31258004903793335,
      "learning_rate": 0.00019820547217631117,
      "loss": 0.4244,
      "step": 181
    },
    {
      "epoch": 0.18205120190053453,
      "grad_norm": 0.4338636100292206,
      "learning_rate": 0.00019818564870839405,
      "loss": 0.4361,
      "step": 182
    },
    {
      "epoch": 0.18305148322965833,
      "grad_norm": 0.2990376949310303,
      "learning_rate": 0.00019816571735220583,
      "loss": 0.4257,
      "step": 183
    },
    {
      "epoch": 0.18405176455878217,
      "grad_norm": 0.3014718294143677,
      "learning_rate": 0.00019814567812964748,
      "loss": 0.3985,
      "step": 184
    },
    {
      "epoch": 0.18505204588790597,
      "grad_norm": 0.2983604073524475,
      "learning_rate": 0.00019812553106273847,
      "loss": 0.413,
      "step": 185
    },
    {
      "epoch": 0.1860523272170298,
      "grad_norm": 0.26064085960388184,
      "learning_rate": 0.00019810527617361681,
      "loss": 0.3713,
      "step": 186
    },
    {
      "epoch": 0.1870526085461536,
      "grad_norm": 0.317233681678772,
      "learning_rate": 0.00019808491348453894,
      "loss": 0.4123,
      "step": 187
    },
    {
      "epoch": 0.18805288987527743,
      "grad_norm": 0.32353103160858154,
      "learning_rate": 0.00019806444301787978,
      "loss": 0.4241,
      "step": 188
    },
    {
      "epoch": 0.18905317120440124,
      "grad_norm": 0.3154282569885254,
      "learning_rate": 0.0001980438647961327,
      "loss": 0.464,
      "step": 189
    },
    {
      "epoch": 0.19005345253352504,
      "grad_norm": 0.5833514928817749,
      "learning_rate": 0.00019802317884190935,
      "loss": 0.4107,
      "step": 190
    },
    {
      "epoch": 0.19105373386264887,
      "grad_norm": 0.26230475306510925,
      "learning_rate": 0.00019800238517793996,
      "loss": 0.4036,
      "step": 191
    },
    {
      "epoch": 0.19205401519177268,
      "grad_norm": 0.2554140090942383,
      "learning_rate": 0.00019798148382707296,
      "loss": 0.3799,
      "step": 192
    },
    {
      "epoch": 0.1930542965208965,
      "grad_norm": 0.2918940484523773,
      "learning_rate": 0.00019796047481227515,
      "loss": 0.4366,
      "step": 193
    },
    {
      "epoch": 0.1940545778500203,
      "grad_norm": 0.27494698762893677,
      "learning_rate": 0.00019793935815663163,
      "loss": 0.4452,
      "step": 194
    },
    {
      "epoch": 0.19505485917914414,
      "grad_norm": 0.31667977571487427,
      "learning_rate": 0.00019791813388334581,
      "loss": 0.4058,
      "step": 195
    },
    {
      "epoch": 0.19605514050826794,
      "grad_norm": 0.8476918339729309,
      "learning_rate": 0.00019789680201573933,
      "loss": 0.4026,
      "step": 196
    },
    {
      "epoch": 0.19705542183739178,
      "grad_norm": 0.38591355085372925,
      "learning_rate": 0.00019787536257725202,
      "loss": 0.4508,
      "step": 197
    },
    {
      "epoch": 0.19805570316651558,
      "grad_norm": 0.565131664276123,
      "learning_rate": 0.00019785381559144196,
      "loss": 0.405,
      "step": 198
    },
    {
      "epoch": 0.1990559844956394,
      "grad_norm": 0.32413753867149353,
      "learning_rate": 0.00019783216108198542,
      "loss": 0.3934,
      "step": 199
    },
    {
      "epoch": 0.2000562658247632,
      "grad_norm": 0.3276815712451935,
      "learning_rate": 0.00019781039907267677,
      "loss": 0.448,
      "step": 200
    },
    {
      "epoch": 0.20105654715388704,
      "grad_norm": 0.3068343698978424,
      "learning_rate": 0.00019778852958742853,
      "loss": 0.4454,
      "step": 201
    },
    {
      "epoch": 0.20205682848301085,
      "grad_norm": 0.29425284266471863,
      "learning_rate": 0.00019776655265027127,
      "loss": 0.4473,
      "step": 202
    },
    {
      "epoch": 0.20305710981213465,
      "grad_norm": 0.26625314354896545,
      "learning_rate": 0.00019774446828535371,
      "loss": 0.3749,
      "step": 203
    },
    {
      "epoch": 0.20405739114125848,
      "grad_norm": 0.29396072030067444,
      "learning_rate": 0.00019772227651694256,
      "loss": 0.4101,
      "step": 204
    },
    {
      "epoch": 0.20505767247038229,
      "grad_norm": 0.301468163728714,
      "learning_rate": 0.00019769997736942258,
      "loss": 0.3681,
      "step": 205
    },
    {
      "epoch": 0.20605795379950612,
      "grad_norm": 0.3525398373603821,
      "learning_rate": 0.00019767757086729647,
      "loss": 0.3903,
      "step": 206
    },
    {
      "epoch": 0.20705823512862992,
      "grad_norm": 0.3096582889556885,
      "learning_rate": 0.00019765505703518496,
      "loss": 0.4027,
      "step": 207
    },
    {
      "epoch": 0.20805851645775375,
      "grad_norm": 0.47655290365219116,
      "learning_rate": 0.00019763243589782662,
      "loss": 0.4027,
      "step": 208
    },
    {
      "epoch": 0.20905879778687755,
      "grad_norm": 0.30790695548057556,
      "learning_rate": 0.00019760970748007803,
      "loss": 0.3621,
      "step": 209
    },
    {
      "epoch": 0.21005907911600138,
      "grad_norm": 0.32526037096977234,
      "learning_rate": 0.0001975868718069136,
      "loss": 0.4366,
      "step": 210
    },
    {
      "epoch": 0.2110593604451252,
      "grad_norm": 0.28599101305007935,
      "learning_rate": 0.00019756392890342563,
      "loss": 0.3801,
      "step": 211
    },
    {
      "epoch": 0.21205964177424902,
      "grad_norm": 0.3474257290363312,
      "learning_rate": 0.00019754087879482422,
      "loss": 0.4245,
      "step": 212
    },
    {
      "epoch": 0.21305992310337282,
      "grad_norm": 0.34767356514930725,
      "learning_rate": 0.00019751772150643722,
      "loss": 0.4571,
      "step": 213
    },
    {
      "epoch": 0.21406020443249665,
      "grad_norm": 0.3164922297000885,
      "learning_rate": 0.00019749445706371038,
      "loss": 0.4161,
      "step": 214
    },
    {
      "epoch": 0.21506048576162046,
      "grad_norm": 0.3492523729801178,
      "learning_rate": 0.00019747108549220702,
      "loss": 0.4551,
      "step": 215
    },
    {
      "epoch": 0.21606076709074426,
      "grad_norm": 0.31708961725234985,
      "learning_rate": 0.00019744760681760832,
      "loss": 0.4416,
      "step": 216
    },
    {
      "epoch": 0.2170610484198681,
      "grad_norm": 0.2811904847621918,
      "learning_rate": 0.00019742402106571314,
      "loss": 0.4187,
      "step": 217
    },
    {
      "epoch": 0.2180613297489919,
      "grad_norm": 0.31777599453926086,
      "learning_rate": 0.00019740032826243788,
      "loss": 0.4135,
      "step": 218
    },
    {
      "epoch": 0.21906161107811573,
      "grad_norm": 0.2507490813732147,
      "learning_rate": 0.0001973765284338167,
      "loss": 0.371,
      "step": 219
    },
    {
      "epoch": 0.22006189240723953,
      "grad_norm": 0.2680091857910156,
      "learning_rate": 0.00019735262160600127,
      "loss": 0.416,
      "step": 220
    },
    {
      "epoch": 0.22106217373636336,
      "grad_norm": 0.359147846698761,
      "learning_rate": 0.00019732860780526088,
      "loss": 0.4347,
      "step": 221
    },
    {
      "epoch": 0.22206245506548716,
      "grad_norm": 0.29516977071762085,
      "learning_rate": 0.00019730448705798239,
      "loss": 0.4553,
      "step": 222
    },
    {
      "epoch": 0.223062736394611,
      "grad_norm": 0.3992477059364319,
      "learning_rate": 0.00019728025939067008,
      "loss": 0.3778,
      "step": 223
    },
    {
      "epoch": 0.2240630177237348,
      "grad_norm": 0.25359246134757996,
      "learning_rate": 0.00019725592482994583,
      "loss": 0.3797,
      "step": 224
    },
    {
      "epoch": 0.22506329905285863,
      "grad_norm": 0.4242265522480011,
      "learning_rate": 0.00019723148340254892,
      "loss": 0.3718,
      "step": 225
    },
    {
      "epoch": 0.22606358038198243,
      "grad_norm": 0.26787999272346497,
      "learning_rate": 0.00019720693513533598,
      "loss": 0.3857,
      "step": 226
    },
    {
      "epoch": 0.22706386171110624,
      "grad_norm": 0.23111741244792938,
      "learning_rate": 0.00019718228005528122,
      "loss": 0.349,
      "step": 227
    },
    {
      "epoch": 0.22806414304023007,
      "grad_norm": 0.2645491063594818,
      "learning_rate": 0.00019715751818947603,
      "loss": 0.413,
      "step": 228
    },
    {
      "epoch": 0.22906442436935387,
      "grad_norm": 0.30736207962036133,
      "learning_rate": 0.0001971326495651293,
      "loss": 0.4129,
      "step": 229
    },
    {
      "epoch": 0.2300647056984777,
      "grad_norm": 0.30426740646362305,
      "learning_rate": 0.00019710767420956705,
      "loss": 0.3977,
      "step": 230
    },
    {
      "epoch": 0.2310649870276015,
      "grad_norm": 0.3317965865135193,
      "learning_rate": 0.0001970825921502328,
      "loss": 0.4309,
      "step": 231
    },
    {
      "epoch": 0.23206526835672533,
      "grad_norm": 0.31584450602531433,
      "learning_rate": 0.0001970574034146871,
      "loss": 0.3855,
      "step": 232
    },
    {
      "epoch": 0.23306554968584914,
      "grad_norm": 0.2980955243110657,
      "learning_rate": 0.00019703210803060782,
      "loss": 0.415,
      "step": 233
    },
    {
      "epoch": 0.23406583101497297,
      "grad_norm": 0.24319759011268616,
      "learning_rate": 0.00019700670602579008,
      "loss": 0.4051,
      "step": 234
    },
    {
      "epoch": 0.23506611234409677,
      "grad_norm": 0.3305763900279999,
      "learning_rate": 0.00019698119742814606,
      "loss": 0.4734,
      "step": 235
    },
    {
      "epoch": 0.2360663936732206,
      "grad_norm": 0.26589512825012207,
      "learning_rate": 0.00019695558226570507,
      "loss": 0.4197,
      "step": 236
    },
    {
      "epoch": 0.2370666750023444,
      "grad_norm": 0.3133488893508911,
      "learning_rate": 0.00019692986056661356,
      "loss": 0.4478,
      "step": 237
    },
    {
      "epoch": 0.23806695633146824,
      "grad_norm": 0.27163878083229065,
      "learning_rate": 0.00019690403235913504,
      "loss": 0.3991,
      "step": 238
    },
    {
      "epoch": 0.23906723766059204,
      "grad_norm": 0.256094753742218,
      "learning_rate": 0.00019687809767165,
      "loss": 0.4073,
      "step": 239
    },
    {
      "epoch": 0.24006751898971584,
      "grad_norm": 0.31229445338249207,
      "learning_rate": 0.000196852056532656,
      "loss": 0.3997,
      "step": 240
    },
    {
      "epoch": 0.24106780031883968,
      "grad_norm": 0.2771662771701813,
      "learning_rate": 0.00019682590897076752,
      "loss": 0.4268,
      "step": 241
    },
    {
      "epoch": 0.24206808164796348,
      "grad_norm": 0.25749140977859497,
      "learning_rate": 0.00019679965501471608,
      "loss": 0.3355,
      "step": 242
    },
    {
      "epoch": 0.2430683629770873,
      "grad_norm": 0.29009538888931274,
      "learning_rate": 0.0001967732946933499,
      "loss": 0.444,
      "step": 243
    },
    {
      "epoch": 0.2440686443062111,
      "grad_norm": 0.2621695101261139,
      "learning_rate": 0.00019674682803563428,
      "loss": 0.4083,
      "step": 244
    },
    {
      "epoch": 0.24506892563533494,
      "grad_norm": 0.342739462852478,
      "learning_rate": 0.00019672025507065131,
      "loss": 0.4375,
      "step": 245
    },
    {
      "epoch": 0.24606920696445875,
      "grad_norm": 0.3089183270931244,
      "learning_rate": 0.00019669357582759983,
      "loss": 0.4151,
      "step": 246
    },
    {
      "epoch": 0.24706948829358258,
      "grad_norm": 0.4531654119491577,
      "learning_rate": 0.00019666679033579552,
      "loss": 0.4266,
      "step": 247
    },
    {
      "epoch": 0.24806976962270638,
      "grad_norm": 0.27016133069992065,
      "learning_rate": 0.00019663989862467082,
      "loss": 0.4274,
      "step": 248
    },
    {
      "epoch": 0.2490700509518302,
      "grad_norm": 0.27403491735458374,
      "learning_rate": 0.00019661290072377482,
      "loss": 0.383,
      "step": 249
    },
    {
      "epoch": 0.25007033228095404,
      "grad_norm": 0.2885066866874695,
      "learning_rate": 0.00019658579666277334,
      "loss": 0.4072,
      "step": 250
    },
    {
      "epoch": 0.2510706136100778,
      "grad_norm": 0.24861767888069153,
      "learning_rate": 0.0001965585864714488,
      "loss": 0.3717,
      "step": 251
    },
    {
      "epoch": 0.25207089493920165,
      "grad_norm": 0.2628852128982544,
      "learning_rate": 0.00019653127017970034,
      "loss": 0.3876,
      "step": 252
    },
    {
      "epoch": 0.2530711762683255,
      "grad_norm": 0.2990763783454895,
      "learning_rate": 0.0001965038478175436,
      "loss": 0.4279,
      "step": 253
    },
    {
      "epoch": 0.25407145759744926,
      "grad_norm": 0.36784109473228455,
      "learning_rate": 0.00019647631941511082,
      "loss": 0.3788,
      "step": 254
    },
    {
      "epoch": 0.2550717389265731,
      "grad_norm": 0.34909719228744507,
      "learning_rate": 0.0001964486850026507,
      "loss": 0.4315,
      "step": 255
    },
    {
      "epoch": 0.2560720202556969,
      "grad_norm": 0.3366207480430603,
      "learning_rate": 0.00019642094461052852,
      "loss": 0.4038,
      "step": 256
    },
    {
      "epoch": 0.25707230158482075,
      "grad_norm": 0.27798953652381897,
      "learning_rate": 0.00019639309826922585,
      "loss": 0.4142,
      "step": 257
    },
    {
      "epoch": 0.2580725829139445,
      "grad_norm": 0.23876960575580597,
      "learning_rate": 0.0001963651460093409,
      "loss": 0.3745,
      "step": 258
    },
    {
      "epoch": 0.25907286424306836,
      "grad_norm": 0.8659366369247437,
      "learning_rate": 0.00019633708786158806,
      "loss": 0.4195,
      "step": 259
    },
    {
      "epoch": 0.2600731455721922,
      "grad_norm": 0.2651257812976837,
      "learning_rate": 0.00019630892385679818,
      "loss": 0.4289,
      "step": 260
    },
    {
      "epoch": 0.261073426901316,
      "grad_norm": 0.23504844307899475,
      "learning_rate": 0.00019628065402591845,
      "loss": 0.3659,
      "step": 261
    },
    {
      "epoch": 0.2620737082304398,
      "grad_norm": 0.2684794068336487,
      "learning_rate": 0.00019625227840001225,
      "loss": 0.3854,
      "step": 262
    },
    {
      "epoch": 0.2630739895595636,
      "grad_norm": 0.25387704372406006,
      "learning_rate": 0.0001962237970102593,
      "loss": 0.3877,
      "step": 263
    },
    {
      "epoch": 0.26407427088868746,
      "grad_norm": 0.24009455740451813,
      "learning_rate": 0.0001961952098879555,
      "loss": 0.3785,
      "step": 264
    },
    {
      "epoch": 0.26507455221781123,
      "grad_norm": 0.27629122138023376,
      "learning_rate": 0.00019616651706451287,
      "loss": 0.3474,
      "step": 265
    },
    {
      "epoch": 0.26607483354693506,
      "grad_norm": 0.2714496850967407,
      "learning_rate": 0.0001961377185714597,
      "loss": 0.3708,
      "step": 266
    },
    {
      "epoch": 0.2670751148760589,
      "grad_norm": 0.2567186951637268,
      "learning_rate": 0.0001961088144404403,
      "loss": 0.3725,
      "step": 267
    },
    {
      "epoch": 0.2680753962051827,
      "grad_norm": 0.2741433382034302,
      "learning_rate": 0.00019607980470321505,
      "loss": 0.4145,
      "step": 268
    },
    {
      "epoch": 0.2690756775343065,
      "grad_norm": 0.28943076729774475,
      "learning_rate": 0.00019605068939166045,
      "loss": 0.4202,
      "step": 269
    },
    {
      "epoch": 0.27007595886343033,
      "grad_norm": 0.2776735723018646,
      "learning_rate": 0.00019602146853776894,
      "loss": 0.3991,
      "step": 270
    },
    {
      "epoch": 0.27107624019255416,
      "grad_norm": 0.23200170695781708,
      "learning_rate": 0.000195992142173649,
      "loss": 0.3167,
      "step": 271
    },
    {
      "epoch": 0.272076521521678,
      "grad_norm": 0.44690176844596863,
      "learning_rate": 0.0001959627103315249,
      "loss": 0.3782,
      "step": 272
    },
    {
      "epoch": 0.27307680285080177,
      "grad_norm": 0.28088852763175964,
      "learning_rate": 0.00019593317304373705,
      "loss": 0.3672,
      "step": 273
    },
    {
      "epoch": 0.2740770841799256,
      "grad_norm": 0.29305198788642883,
      "learning_rate": 0.00019590353034274144,
      "loss": 0.4075,
      "step": 274
    },
    {
      "epoch": 0.27507736550904943,
      "grad_norm": 0.30940836668014526,
      "learning_rate": 0.00019587378226111014,
      "loss": 0.4669,
      "step": 275
    },
    {
      "epoch": 0.27607764683817326,
      "grad_norm": 0.2809953987598419,
      "learning_rate": 0.00019584392883153088,
      "loss": 0.3778,
      "step": 276
    },
    {
      "epoch": 0.27707792816729704,
      "grad_norm": 0.32367703318595886,
      "learning_rate": 0.00019581397008680717,
      "loss": 0.3945,
      "step": 277
    },
    {
      "epoch": 0.27807820949642087,
      "grad_norm": 0.3615182042121887,
      "learning_rate": 0.00019578390605985826,
      "loss": 0.4112,
      "step": 278
    },
    {
      "epoch": 0.2790784908255447,
      "grad_norm": 0.3386395275592804,
      "learning_rate": 0.00019575373678371909,
      "loss": 0.4026,
      "step": 279
    },
    {
      "epoch": 0.2800787721546685,
      "grad_norm": 0.30399829149246216,
      "learning_rate": 0.00019572346229154025,
      "loss": 0.3748,
      "step": 280
    },
    {
      "epoch": 0.2810790534837923,
      "grad_norm": 0.24133749306201935,
      "learning_rate": 0.00019569308261658787,
      "loss": 0.3774,
      "step": 281
    },
    {
      "epoch": 0.28207933481291614,
      "grad_norm": 0.24240224063396454,
      "learning_rate": 0.00019566259779224378,
      "loss": 0.4204,
      "step": 282
    },
    {
      "epoch": 0.28307961614203997,
      "grad_norm": 0.278439998626709,
      "learning_rate": 0.00019563200785200526,
      "loss": 0.375,
      "step": 283
    },
    {
      "epoch": 0.28407989747116374,
      "grad_norm": 0.24788439273834229,
      "learning_rate": 0.00019560131282948516,
      "loss": 0.3577,
      "step": 284
    },
    {
      "epoch": 0.2850801788002876,
      "grad_norm": 0.31011322140693665,
      "learning_rate": 0.0001955705127584117,
      "loss": 0.4538,
      "step": 285
    },
    {
      "epoch": 0.2860804601294114,
      "grad_norm": 0.257071316242218,
      "learning_rate": 0.00019553960767262863,
      "loss": 0.3881,
      "step": 286
    },
    {
      "epoch": 0.28708074145853524,
      "grad_norm": 0.5129604339599609,
      "learning_rate": 0.00019550859760609503,
      "loss": 0.3833,
      "step": 287
    },
    {
      "epoch": 0.288081022787659,
      "grad_norm": 0.234080508351326,
      "learning_rate": 0.00019547748259288536,
      "loss": 0.3843,
      "step": 288
    },
    {
      "epoch": 0.28908130411678284,
      "grad_norm": 0.4176006615161896,
      "learning_rate": 0.0001954462626671894,
      "loss": 0.431,
      "step": 289
    },
    {
      "epoch": 0.2900815854459067,
      "grad_norm": 0.30611085891723633,
      "learning_rate": 0.0001954149378633122,
      "loss": 0.4091,
      "step": 290
    },
    {
      "epoch": 0.29108186677503045,
      "grad_norm": 0.26238131523132324,
      "learning_rate": 0.00019538350821567404,
      "loss": 0.346,
      "step": 291
    },
    {
      "epoch": 0.2920821481041543,
      "grad_norm": 0.28092989325523376,
      "learning_rate": 0.00019535197375881045,
      "loss": 0.4251,
      "step": 292
    },
    {
      "epoch": 0.2930824294332781,
      "grad_norm": 0.27506351470947266,
      "learning_rate": 0.00019532033452737205,
      "loss": 0.3944,
      "step": 293
    },
    {
      "epoch": 0.29408271076240194,
      "grad_norm": 0.2865377366542816,
      "learning_rate": 0.00019528859055612468,
      "loss": 0.4214,
      "step": 294
    },
    {
      "epoch": 0.2950829920915257,
      "grad_norm": 0.2901553809642792,
      "learning_rate": 0.0001952567418799492,
      "loss": 0.4395,
      "step": 295
    },
    {
      "epoch": 0.29608327342064955,
      "grad_norm": 0.22749190032482147,
      "learning_rate": 0.00019522478853384155,
      "loss": 0.36,
      "step": 296
    },
    {
      "epoch": 0.2970835547497734,
      "grad_norm": 0.26204293966293335,
      "learning_rate": 0.00019519273055291266,
      "loss": 0.4006,
      "step": 297
    },
    {
      "epoch": 0.2980838360788972,
      "grad_norm": 0.25596141815185547,
      "learning_rate": 0.00019516056797238846,
      "loss": 0.3789,
      "step": 298
    },
    {
      "epoch": 0.299084117408021,
      "grad_norm": 0.2562818229198456,
      "learning_rate": 0.00019512830082760987,
      "loss": 0.404,
      "step": 299
    },
    {
      "epoch": 0.3000843987371448,
      "grad_norm": 0.3109433352947235,
      "learning_rate": 0.00019509592915403255,
      "loss": 0.4498,
      "step": 300
    },
    {
      "epoch": 0.30108468006626865,
      "grad_norm": 0.27133166790008545,
      "learning_rate": 0.00019506345298722717,
      "loss": 0.3634,
      "step": 301
    },
    {
      "epoch": 0.3020849613953925,
      "grad_norm": 0.29989707469940186,
      "learning_rate": 0.00019503087236287913,
      "loss": 0.4089,
      "step": 302
    },
    {
      "epoch": 0.30308524272451626,
      "grad_norm": 0.31667646765708923,
      "learning_rate": 0.00019499818731678873,
      "loss": 0.3368,
      "step": 303
    },
    {
      "epoch": 0.3040855240536401,
      "grad_norm": 0.2644802927970886,
      "learning_rate": 0.00019496539788487082,
      "loss": 0.39,
      "step": 304
    },
    {
      "epoch": 0.3050858053827639,
      "grad_norm": 0.36831945180892944,
      "learning_rate": 0.0001949325041031551,
      "loss": 0.4179,
      "step": 305
    },
    {
      "epoch": 0.3060860867118877,
      "grad_norm": 0.2724269926548004,
      "learning_rate": 0.0001948995060077859,
      "loss": 0.4387,
      "step": 306
    },
    {
      "epoch": 0.3070863680410115,
      "grad_norm": 0.24974443018436432,
      "learning_rate": 0.0001948664036350221,
      "loss": 0.3933,
      "step": 307
    },
    {
      "epoch": 0.30808664937013536,
      "grad_norm": 0.26349300146102905,
      "learning_rate": 0.00019483319702123732,
      "loss": 0.3895,
      "step": 308
    },
    {
      "epoch": 0.3090869306992592,
      "grad_norm": 0.3007661998271942,
      "learning_rate": 0.00019479988620291956,
      "loss": 0.3507,
      "step": 309
    },
    {
      "epoch": 0.31008721202838296,
      "grad_norm": 0.2490624487400055,
      "learning_rate": 0.00019476647121667137,
      "loss": 0.3334,
      "step": 310
    },
    {
      "epoch": 0.3110874933575068,
      "grad_norm": 0.2903260886669159,
      "learning_rate": 0.00019473295209920983,
      "loss": 0.4624,
      "step": 311
    },
    {
      "epoch": 0.3120877746866306,
      "grad_norm": 0.2873629033565521,
      "learning_rate": 0.00019469932888736632,
      "loss": 0.4099,
      "step": 312
    },
    {
      "epoch": 0.31308805601575446,
      "grad_norm": 0.2918339967727661,
      "learning_rate": 0.00019466560161808674,
      "loss": 0.3617,
      "step": 313
    },
    {
      "epoch": 0.31408833734487823,
      "grad_norm": 0.2637473940849304,
      "learning_rate": 0.00019463177032843124,
      "loss": 0.3906,
      "step": 314
    },
    {
      "epoch": 0.31508861867400206,
      "grad_norm": 0.364161878824234,
      "learning_rate": 0.00019459783505557424,
      "loss": 0.402,
      "step": 315
    },
    {
      "epoch": 0.3160889000031259,
      "grad_norm": 0.27097612619400024,
      "learning_rate": 0.00019456379583680452,
      "loss": 0.3748,
      "step": 316
    },
    {
      "epoch": 0.31708918133224967,
      "grad_norm": 0.29231584072113037,
      "learning_rate": 0.000194529652709525,
      "loss": 0.4232,
      "step": 317
    },
    {
      "epoch": 0.3180894626613735,
      "grad_norm": 0.2763817310333252,
      "learning_rate": 0.00019449540571125286,
      "loss": 0.3895,
      "step": 318
    },
    {
      "epoch": 0.31908974399049733,
      "grad_norm": 0.27088895440101624,
      "learning_rate": 0.00019446105487961926,
      "loss": 0.369,
      "step": 319
    },
    {
      "epoch": 0.32009002531962116,
      "grad_norm": 0.37634366750717163,
      "learning_rate": 0.0001944266002523696,
      "loss": 0.3474,
      "step": 320
    },
    {
      "epoch": 0.32109030664874494,
      "grad_norm": 0.26676034927368164,
      "learning_rate": 0.0001943920418673633,
      "loss": 0.3793,
      "step": 321
    },
    {
      "epoch": 0.32209058797786877,
      "grad_norm": 0.27637040615081787,
      "learning_rate": 0.00019435737976257377,
      "loss": 0.3914,
      "step": 322
    },
    {
      "epoch": 0.3230908693069926,
      "grad_norm": 0.2805480360984802,
      "learning_rate": 0.00019432261397608834,
      "loss": 0.3985,
      "step": 323
    },
    {
      "epoch": 0.32409115063611643,
      "grad_norm": 0.23319090902805328,
      "learning_rate": 0.00019428774454610843,
      "loss": 0.3664,
      "step": 324
    },
    {
      "epoch": 0.3250914319652402,
      "grad_norm": 0.23325973749160767,
      "learning_rate": 0.00019425277151094913,
      "loss": 0.3393,
      "step": 325
    },
    {
      "epoch": 0.32609171329436404,
      "grad_norm": 0.26448819041252136,
      "learning_rate": 0.00019421769490903957,
      "loss": 0.3972,
      "step": 326
    },
    {
      "epoch": 0.32709199462348787,
      "grad_norm": 0.2579127848148346,
      "learning_rate": 0.0001941825147789225,
      "loss": 0.317,
      "step": 327
    },
    {
      "epoch": 0.32809227595261165,
      "grad_norm": 0.2921576499938965,
      "learning_rate": 0.00019414723115925456,
      "loss": 0.3781,
      "step": 328
    },
    {
      "epoch": 0.3290925572817355,
      "grad_norm": 0.2583731710910797,
      "learning_rate": 0.0001941118440888061,
      "loss": 0.4428,
      "step": 329
    },
    {
      "epoch": 0.3300928386108593,
      "grad_norm": 0.30170106887817383,
      "learning_rate": 0.0001940763536064611,
      "loss": 0.3839,
      "step": 330
    },
    {
      "epoch": 0.33109311993998314,
      "grad_norm": 0.35121557116508484,
      "learning_rate": 0.00019404075975121716,
      "loss": 0.4176,
      "step": 331
    },
    {
      "epoch": 0.3320934012691069,
      "grad_norm": 0.29237017035484314,
      "learning_rate": 0.0001940050625621855,
      "loss": 0.3906,
      "step": 332
    },
    {
      "epoch": 0.33309368259823074,
      "grad_norm": 0.22922085225582123,
      "learning_rate": 0.00019396926207859084,
      "loss": 0.36,
      "step": 333
    },
    {
      "epoch": 0.3340939639273546,
      "grad_norm": 0.2649683952331543,
      "learning_rate": 0.0001939333583397715,
      "loss": 0.3781,
      "step": 334
    },
    {
      "epoch": 0.3350942452564784,
      "grad_norm": 0.22827917337417603,
      "learning_rate": 0.00019389735138517915,
      "loss": 0.343,
      "step": 335
    },
    {
      "epoch": 0.3360945265856022,
      "grad_norm": 0.26274073123931885,
      "learning_rate": 0.00019386124125437895,
      "loss": 0.3464,
      "step": 336
    },
    {
      "epoch": 0.337094807914726,
      "grad_norm": 0.27798980474472046,
      "learning_rate": 0.00019382502798704935,
      "loss": 0.3553,
      "step": 337
    },
    {
      "epoch": 0.33809508924384984,
      "grad_norm": 0.3262801170349121,
      "learning_rate": 0.00019378871162298227,
      "loss": 0.4019,
      "step": 338
    },
    {
      "epoch": 0.3390953705729737,
      "grad_norm": 0.27483808994293213,
      "learning_rate": 0.00019375229220208276,
      "loss": 0.3487,
      "step": 339
    },
    {
      "epoch": 0.34009565190209745,
      "grad_norm": 0.38662320375442505,
      "learning_rate": 0.00019371576976436917,
      "loss": 0.4287,
      "step": 340
    },
    {
      "epoch": 0.3410959332312213,
      "grad_norm": 0.2590675354003906,
      "learning_rate": 0.00019367914434997312,
      "loss": 0.3743,
      "step": 341
    },
    {
      "epoch": 0.3420962145603451,
      "grad_norm": 0.2873009741306305,
      "learning_rate": 0.00019364241599913924,
      "loss": 0.3902,
      "step": 342
    },
    {
      "epoch": 0.3430964958894689,
      "grad_norm": 0.3668614327907562,
      "learning_rate": 0.0001936055847522254,
      "loss": 0.3645,
      "step": 343
    },
    {
      "epoch": 0.3440967772185927,
      "grad_norm": 0.306484580039978,
      "learning_rate": 0.00019356865064970244,
      "loss": 0.3919,
      "step": 344
    },
    {
      "epoch": 0.34509705854771655,
      "grad_norm": 0.3255596458911896,
      "learning_rate": 0.0001935316137321543,
      "loss": 0.4359,
      "step": 345
    },
    {
      "epoch": 0.3460973398768404,
      "grad_norm": 0.4163992702960968,
      "learning_rate": 0.00019349447404027782,
      "loss": 0.3574,
      "step": 346
    },
    {
      "epoch": 0.34709762120596416,
      "grad_norm": 0.344198614358902,
      "learning_rate": 0.00019345723161488283,
      "loss": 0.4011,
      "step": 347
    },
    {
      "epoch": 0.348097902535088,
      "grad_norm": 0.2813580334186554,
      "learning_rate": 0.000193419886496892,
      "loss": 0.3766,
      "step": 348
    },
    {
      "epoch": 0.3490981838642118,
      "grad_norm": 0.27424079179763794,
      "learning_rate": 0.00019338243872734086,
      "loss": 0.3241,
      "step": 349
    },
    {
      "epoch": 0.35009846519333565,
      "grad_norm": 0.2421925663948059,
      "learning_rate": 0.00019334488834737775,
      "loss": 0.3852,
      "step": 350
    },
    {
      "epoch": 0.3510987465224594,
      "grad_norm": 0.22247883677482605,
      "learning_rate": 0.00019330723539826375,
      "loss": 0.3595,
      "step": 351
    },
    {
      "epoch": 0.35209902785158326,
      "grad_norm": 0.26099297404289246,
      "learning_rate": 0.00019326947992137262,
      "loss": 0.3739,
      "step": 352
    },
    {
      "epoch": 0.3530993091807071,
      "grad_norm": 0.28967341780662537,
      "learning_rate": 0.00019323162195819082,
      "loss": 0.3831,
      "step": 353
    },
    {
      "epoch": 0.35409959050983086,
      "grad_norm": 0.2714492678642273,
      "learning_rate": 0.0001931936615503174,
      "loss": 0.3413,
      "step": 354
    },
    {
      "epoch": 0.3550998718389547,
      "grad_norm": 0.27229174971580505,
      "learning_rate": 0.000193155598739464,
      "loss": 0.3825,
      "step": 355
    },
    {
      "epoch": 0.3561001531680785,
      "grad_norm": 0.23672465980052948,
      "learning_rate": 0.0001931174335674547,
      "loss": 0.3694,
      "step": 356
    },
    {
      "epoch": 0.35710043449720236,
      "grad_norm": 0.23558713495731354,
      "learning_rate": 0.0001930791660762262,
      "loss": 0.3629,
      "step": 357
    },
    {
      "epoch": 0.35810071582632613,
      "grad_norm": 0.46677660942077637,
      "learning_rate": 0.00019304079630782752,
      "loss": 0.3946,
      "step": 358
    },
    {
      "epoch": 0.35910099715544996,
      "grad_norm": 0.24028021097183228,
      "learning_rate": 0.0001930023243044201,
      "loss": 0.3605,
      "step": 359
    },
    {
      "epoch": 0.3601012784845738,
      "grad_norm": 0.25813183188438416,
      "learning_rate": 0.00019296375010827773,
      "loss": 0.3983,
      "step": 360
    },
    {
      "epoch": 0.3611015598136976,
      "grad_norm": 0.23249755799770355,
      "learning_rate": 0.00019292507376178643,
      "loss": 0.3963,
      "step": 361
    },
    {
      "epoch": 0.3621018411428214,
      "grad_norm": 0.31035348773002625,
      "learning_rate": 0.00019288629530744454,
      "loss": 0.4103,
      "step": 362
    },
    {
      "epoch": 0.36310212247194523,
      "grad_norm": 0.2520258128643036,
      "learning_rate": 0.0001928474147878626,
      "loss": 0.3408,
      "step": 363
    },
    {
      "epoch": 0.36410240380106906,
      "grad_norm": 0.25639891624450684,
      "learning_rate": 0.0001928084322457632,
      "loss": 0.3771,
      "step": 364
    },
    {
      "epoch": 0.3651026851301929,
      "grad_norm": 0.2763039171695709,
      "learning_rate": 0.00019276934772398114,
      "loss": 0.3385,
      "step": 365
    },
    {
      "epoch": 0.36610296645931667,
      "grad_norm": 0.2602714002132416,
      "learning_rate": 0.00019273016126546323,
      "loss": 0.3836,
      "step": 366
    },
    {
      "epoch": 0.3671032477884405,
      "grad_norm": 0.33483123779296875,
      "learning_rate": 0.00019269087291326833,
      "loss": 0.4129,
      "step": 367
    },
    {
      "epoch": 0.36810352911756433,
      "grad_norm": 0.23084044456481934,
      "learning_rate": 0.00019265148271056722,
      "loss": 0.326,
      "step": 368
    },
    {
      "epoch": 0.3691038104466881,
      "grad_norm": 0.26484981179237366,
      "learning_rate": 0.0001926119907006426,
      "loss": 0.3881,
      "step": 369
    },
    {
      "epoch": 0.37010409177581194,
      "grad_norm": 0.24120277166366577,
      "learning_rate": 0.00019257239692688907,
      "loss": 0.3705,
      "step": 370
    },
    {
      "epoch": 0.37110437310493577,
      "grad_norm": 0.23136964440345764,
      "learning_rate": 0.00019253270143281296,
      "loss": 0.3632,
      "step": 371
    },
    {
      "epoch": 0.3721046544340596,
      "grad_norm": 0.24877014756202698,
      "learning_rate": 0.00019249290426203252,
      "loss": 0.3409,
      "step": 372
    },
    {
      "epoch": 0.3731049357631834,
      "grad_norm": 0.23634536564350128,
      "learning_rate": 0.0001924530054582776,
      "loss": 0.3415,
      "step": 373
    },
    {
      "epoch": 0.3741052170923072,
      "grad_norm": 0.6095355153083801,
      "learning_rate": 0.0001924130050653898,
      "loss": 0.3556,
      "step": 374
    },
    {
      "epoch": 0.37510549842143104,
      "grad_norm": 0.2479625940322876,
      "learning_rate": 0.00019237290312732226,
      "loss": 0.345,
      "step": 375
    },
    {
      "epoch": 0.37610577975055487,
      "grad_norm": 0.22326619923114777,
      "learning_rate": 0.00019233269968813984,
      "loss": 0.3442,
      "step": 376
    },
    {
      "epoch": 0.37710606107967864,
      "grad_norm": 0.2376389503479004,
      "learning_rate": 0.00019229239479201876,
      "loss": 0.3251,
      "step": 377
    },
    {
      "epoch": 0.3781063424088025,
      "grad_norm": 0.2835697829723358,
      "learning_rate": 0.0001922519884832469,
      "loss": 0.3834,
      "step": 378
    },
    {
      "epoch": 0.3791066237379263,
      "grad_norm": 0.27211618423461914,
      "learning_rate": 0.0001922114808062234,
      "loss": 0.3642,
      "step": 379
    },
    {
      "epoch": 0.3801069050670501,
      "grad_norm": 0.3268214166164398,
      "learning_rate": 0.00019217087180545893,
      "loss": 0.3909,
      "step": 380
    },
    {
      "epoch": 0.3811071863961739,
      "grad_norm": 0.2260449230670929,
      "learning_rate": 0.0001921301615255754,
      "loss": 0.3796,
      "step": 381
    },
    {
      "epoch": 0.38210746772529774,
      "grad_norm": 0.21646197140216827,
      "learning_rate": 0.0001920893500113061,
      "loss": 0.3455,
      "step": 382
    },
    {
      "epoch": 0.3831077490544216,
      "grad_norm": 0.27988046407699585,
      "learning_rate": 0.00019204843730749547,
      "loss": 0.3443,
      "step": 383
    },
    {
      "epoch": 0.38410803038354535,
      "grad_norm": 0.2434566169977188,
      "learning_rate": 0.00019200742345909915,
      "loss": 0.36,
      "step": 384
    },
    {
      "epoch": 0.3851083117126692,
      "grad_norm": 0.21464578807353973,
      "learning_rate": 0.00019196630851118398,
      "loss": 0.3434,
      "step": 385
    },
    {
      "epoch": 0.386108593041793,
      "grad_norm": 0.2649652063846588,
      "learning_rate": 0.0001919250925089278,
      "loss": 0.3438,
      "step": 386
    },
    {
      "epoch": 0.38710887437091684,
      "grad_norm": 0.21409133076667786,
      "learning_rate": 0.00019188377549761963,
      "loss": 0.3675,
      "step": 387
    },
    {
      "epoch": 0.3881091557000406,
      "grad_norm": 0.27577823400497437,
      "learning_rate": 0.00019184235752265928,
      "loss": 0.3486,
      "step": 388
    },
    {
      "epoch": 0.38910943702916445,
      "grad_norm": 0.3003029227256775,
      "learning_rate": 0.00019180083862955772,
      "loss": 0.3968,
      "step": 389
    },
    {
      "epoch": 0.3901097183582883,
      "grad_norm": 0.2916281521320343,
      "learning_rate": 0.00019175921886393666,
      "loss": 0.3835,
      "step": 390
    },
    {
      "epoch": 0.39110999968741206,
      "grad_norm": 0.27574488520622253,
      "learning_rate": 0.00019171749827152869,
      "loss": 0.3958,
      "step": 391
    },
    {
      "epoch": 0.3921102810165359,
      "grad_norm": 0.25350049138069153,
      "learning_rate": 0.0001916756768981772,
      "loss": 0.3579,
      "step": 392
    },
    {
      "epoch": 0.3931105623456597,
      "grad_norm": 0.2273736596107483,
      "learning_rate": 0.00019163375478983632,
      "loss": 0.3348,
      "step": 393
    },
    {
      "epoch": 0.39411084367478355,
      "grad_norm": 0.31927919387817383,
      "learning_rate": 0.00019159173199257085,
      "loss": 0.3714,
      "step": 394
    },
    {
      "epoch": 0.3951111250039073,
      "grad_norm": 0.22905245423316956,
      "learning_rate": 0.00019154960855255628,
      "loss": 0.3707,
      "step": 395
    },
    {
      "epoch": 0.39611140633303116,
      "grad_norm": 0.22161217033863068,
      "learning_rate": 0.0001915073845160786,
      "loss": 0.3591,
      "step": 396
    },
    {
      "epoch": 0.397111687662155,
      "grad_norm": 0.24688982963562012,
      "learning_rate": 0.00019146505992953446,
      "loss": 0.3113,
      "step": 397
    },
    {
      "epoch": 0.3981119689912788,
      "grad_norm": 0.3700356185436249,
      "learning_rate": 0.00019142263483943085,
      "loss": 0.3675,
      "step": 398
    },
    {
      "epoch": 0.3991122503204026,
      "grad_norm": 0.2571461796760559,
      "learning_rate": 0.00019138010929238534,
      "loss": 0.3741,
      "step": 399
    },
    {
      "epoch": 0.4001125316495264,
      "grad_norm": 0.25479117035865784,
      "learning_rate": 0.00019133748333512575,
      "loss": 0.3374,
      "step": 400
    },
    {
      "epoch": 0.40111281297865026,
      "grad_norm": 0.300425261259079,
      "learning_rate": 0.00019129475701449035,
      "loss": 0.3541,
      "step": 401
    },
    {
      "epoch": 0.4021130943077741,
      "grad_norm": 0.2358659952878952,
      "learning_rate": 0.0001912519303774276,
      "loss": 0.3808,
      "step": 402
    },
    {
      "epoch": 0.40311337563689786,
      "grad_norm": 0.22720776498317719,
      "learning_rate": 0.0001912090034709963,
      "loss": 0.3359,
      "step": 403
    },
    {
      "epoch": 0.4041136569660217,
      "grad_norm": 0.2396584004163742,
      "learning_rate": 0.00019116597634236525,
      "loss": 0.3397,
      "step": 404
    },
    {
      "epoch": 0.4051139382951455,
      "grad_norm": 0.3370227813720703,
      "learning_rate": 0.0001911228490388136,
      "loss": 0.347,
      "step": 405
    },
    {
      "epoch": 0.4061142196242693,
      "grad_norm": 0.22302855551242828,
      "learning_rate": 0.00019107962160773035,
      "loss": 0.3055,
      "step": 406
    },
    {
      "epoch": 0.40711450095339313,
      "grad_norm": 0.2422298789024353,
      "learning_rate": 0.0001910362940966147,
      "loss": 0.3662,
      "step": 407
    },
    {
      "epoch": 0.40811478228251696,
      "grad_norm": 0.24443741142749786,
      "learning_rate": 0.00019099286655307568,
      "loss": 0.3929,
      "step": 408
    },
    {
      "epoch": 0.4091150636116408,
      "grad_norm": 0.2055700719356537,
      "learning_rate": 0.0001909493390248324,
      "loss": 0.3078,
      "step": 409
    },
    {
      "epoch": 0.41011534494076457,
      "grad_norm": 0.2374294251203537,
      "learning_rate": 0.00019090571155971366,
      "loss": 0.353,
      "step": 410
    },
    {
      "epoch": 0.4111156262698884,
      "grad_norm": 0.29268407821655273,
      "learning_rate": 0.00019086198420565823,
      "loss": 0.3204,
      "step": 411
    },
    {
      "epoch": 0.41211590759901223,
      "grad_norm": 0.26289287209510803,
      "learning_rate": 0.00019081815701071445,
      "loss": 0.292,
      "step": 412
    },
    {
      "epoch": 0.41311618892813606,
      "grad_norm": 0.2546895742416382,
      "learning_rate": 0.0001907742300230406,
      "loss": 0.3856,
      "step": 413
    },
    {
      "epoch": 0.41411647025725984,
      "grad_norm": 0.23252642154693604,
      "learning_rate": 0.00019073020329090444,
      "loss": 0.3482,
      "step": 414
    },
    {
      "epoch": 0.41511675158638367,
      "grad_norm": 0.2277359515428543,
      "learning_rate": 0.0001906860768626834,
      "loss": 0.3185,
      "step": 415
    },
    {
      "epoch": 0.4161170329155075,
      "grad_norm": 0.2553969919681549,
      "learning_rate": 0.00019064185078686443,
      "loss": 0.3476,
      "step": 416
    },
    {
      "epoch": 0.4171173142446313,
      "grad_norm": 0.26211032271385193,
      "learning_rate": 0.000190597525112044,
      "loss": 0.3506,
      "step": 417
    },
    {
      "epoch": 0.4181175955737551,
      "grad_norm": 0.32664135098457336,
      "learning_rate": 0.000190553099886928,
      "loss": 0.3519,
      "step": 418
    },
    {
      "epoch": 0.41911787690287894,
      "grad_norm": 0.2600671947002411,
      "learning_rate": 0.00019050857516033173,
      "loss": 0.3422,
      "step": 419
    },
    {
      "epoch": 0.42011815823200277,
      "grad_norm": 0.24694529175758362,
      "learning_rate": 0.00019046395098117983,
      "loss": 0.3411,
      "step": 420
    },
    {
      "epoch": 0.42111843956112655,
      "grad_norm": 0.25047555565834045,
      "learning_rate": 0.00019041922739850616,
      "loss": 0.3654,
      "step": 421
    },
    {
      "epoch": 0.4221187208902504,
      "grad_norm": 0.2995339035987854,
      "learning_rate": 0.00019037440446145385,
      "loss": 0.4087,
      "step": 422
    },
    {
      "epoch": 0.4231190022193742,
      "grad_norm": 0.25556957721710205,
      "learning_rate": 0.00019032948221927524,
      "loss": 0.3692,
      "step": 423
    },
    {
      "epoch": 0.42411928354849804,
      "grad_norm": 0.26491332054138184,
      "learning_rate": 0.00019028446072133175,
      "loss": 0.3866,
      "step": 424
    },
    {
      "epoch": 0.4251195648776218,
      "grad_norm": 0.23855310678482056,
      "learning_rate": 0.00019023934001709383,
      "loss": 0.329,
      "step": 425
    },
    {
      "epoch": 0.42611984620674564,
      "grad_norm": 0.22823669016361237,
      "learning_rate": 0.00019019412015614098,
      "loss": 0.3594,
      "step": 426
    },
    {
      "epoch": 0.4271201275358695,
      "grad_norm": 0.3833972215652466,
      "learning_rate": 0.00019014880118816164,
      "loss": 0.3786,
      "step": 427
    },
    {
      "epoch": 0.4281204088649933,
      "grad_norm": 0.2663598954677582,
      "learning_rate": 0.0001901033831629532,
      "loss": 0.3619,
      "step": 428
    },
    {
      "epoch": 0.4291206901941171,
      "grad_norm": 0.28205564618110657,
      "learning_rate": 0.00019005786613042185,
      "loss": 0.3879,
      "step": 429
    },
    {
      "epoch": 0.4301209715232409,
      "grad_norm": 0.2598055899143219,
      "learning_rate": 0.00019001225014058255,
      "loss": 0.3448,
      "step": 430
    },
    {
      "epoch": 0.43112125285236474,
      "grad_norm": 0.279117614030838,
      "learning_rate": 0.00018996653524355902,
      "loss": 0.3787,
      "step": 431
    },
    {
      "epoch": 0.4321215341814885,
      "grad_norm": 0.22764593362808228,
      "learning_rate": 0.00018992072148958368,
      "loss": 0.3862,
      "step": 432
    },
    {
      "epoch": 0.43312181551061235,
      "grad_norm": 0.38709068298339844,
      "learning_rate": 0.00018987480892899758,
      "loss": 0.3459,
      "step": 433
    },
    {
      "epoch": 0.4341220968397362,
      "grad_norm": 0.2217923253774643,
      "learning_rate": 0.00018982879761225027,
      "loss": 0.3517,
      "step": 434
    },
    {
      "epoch": 0.43512237816886,
      "grad_norm": 0.23652343451976776,
      "learning_rate": 0.00018978268758989991,
      "loss": 0.3644,
      "step": 435
    },
    {
      "epoch": 0.4361226594979838,
      "grad_norm": 0.5657463669776917,
      "learning_rate": 0.00018973647891261307,
      "loss": 0.3325,
      "step": 436
    },
    {
      "epoch": 0.4371229408271076,
      "grad_norm": 0.32180288434028625,
      "learning_rate": 0.00018969017163116472,
      "loss": 0.3864,
      "step": 437
    },
    {
      "epoch": 0.43812322215623145,
      "grad_norm": 0.3013526201248169,
      "learning_rate": 0.0001896437657964382,
      "loss": 0.4002,
      "step": 438
    },
    {
      "epoch": 0.4391235034853553,
      "grad_norm": 0.26990970969200134,
      "learning_rate": 0.00018959726145942508,
      "loss": 0.3331,
      "step": 439
    },
    {
      "epoch": 0.44012378481447906,
      "grad_norm": 0.19621394574642181,
      "learning_rate": 0.00018955065867122528,
      "loss": 0.2931,
      "step": 440
    },
    {
      "epoch": 0.4411240661436029,
      "grad_norm": 0.2245367169380188,
      "learning_rate": 0.00018950395748304678,
      "loss": 0.333,
      "step": 441
    },
    {
      "epoch": 0.4421243474727267,
      "grad_norm": 0.2619412839412689,
      "learning_rate": 0.0001894571579462058,
      "loss": 0.3442,
      "step": 442
    },
    {
      "epoch": 0.4431246288018505,
      "grad_norm": 0.26194262504577637,
      "learning_rate": 0.00018941026011212654,
      "loss": 0.3369,
      "step": 443
    },
    {
      "epoch": 0.4441249101309743,
      "grad_norm": 0.2473272681236267,
      "learning_rate": 0.00018936326403234125,
      "loss": 0.3724,
      "step": 444
    },
    {
      "epoch": 0.44512519146009816,
      "grad_norm": 0.32609519362449646,
      "learning_rate": 0.00018931616975849006,
      "loss": 0.3933,
      "step": 445
    },
    {
      "epoch": 0.446125472789222,
      "grad_norm": 0.3470418155193329,
      "learning_rate": 0.00018926897734232115,
      "loss": 0.4049,
      "step": 446
    },
    {
      "epoch": 0.44712575411834576,
      "grad_norm": 0.2971716821193695,
      "learning_rate": 0.0001892216868356904,
      "loss": 0.3469,
      "step": 447
    },
    {
      "epoch": 0.4481260354474696,
      "grad_norm": 0.24874372780323029,
      "learning_rate": 0.0001891742982905615,
      "loss": 0.3531,
      "step": 448
    },
    {
      "epoch": 0.4491263167765934,
      "grad_norm": 0.2729557752609253,
      "learning_rate": 0.00018912681175900598,
      "loss": 0.3627,
      "step": 449
    },
    {
      "epoch": 0.45012659810571726,
      "grad_norm": 0.2681429982185364,
      "learning_rate": 0.00018907922729320285,
      "loss": 0.3675,
      "step": 450
    },
    {
      "epoch": 0.45112687943484103,
      "grad_norm": 0.2941266894340515,
      "learning_rate": 0.00018903154494543889,
      "loss": 0.4168,
      "step": 451
    },
    {
      "epoch": 0.45212716076396486,
      "grad_norm": 0.24707025289535522,
      "learning_rate": 0.00018898376476810834,
      "loss": 0.3437,
      "step": 452
    },
    {
      "epoch": 0.4531274420930887,
      "grad_norm": 0.3138103187084198,
      "learning_rate": 0.00018893588681371303,
      "loss": 0.3738,
      "step": 453
    },
    {
      "epoch": 0.45412772342221247,
      "grad_norm": 0.32942667603492737,
      "learning_rate": 0.00018888791113486213,
      "loss": 0.4186,
      "step": 454
    },
    {
      "epoch": 0.4551280047513363,
      "grad_norm": 0.26734960079193115,
      "learning_rate": 0.00018883983778427227,
      "loss": 0.3519,
      "step": 455
    },
    {
      "epoch": 0.45612828608046013,
      "grad_norm": 0.23201091587543488,
      "learning_rate": 0.0001887916668147673,
      "loss": 0.3613,
      "step": 456
    },
    {
      "epoch": 0.45712856740958396,
      "grad_norm": 0.2691316306591034,
      "learning_rate": 0.00018874339827927846,
      "loss": 0.3924,
      "step": 457
    },
    {
      "epoch": 0.45812884873870774,
      "grad_norm": 0.27234920859336853,
      "learning_rate": 0.00018869503223084414,
      "loss": 0.3704,
      "step": 458
    },
    {
      "epoch": 0.45912913006783157,
      "grad_norm": 0.24237160384655,
      "learning_rate": 0.00018864656872260985,
      "loss": 0.3009,
      "step": 459
    },
    {
      "epoch": 0.4601294113969554,
      "grad_norm": 0.2326439768075943,
      "learning_rate": 0.00018859800780782828,
      "loss": 0.3243,
      "step": 460
    },
    {
      "epoch": 0.46112969272607923,
      "grad_norm": 0.250873863697052,
      "learning_rate": 0.000188549349539859,
      "loss": 0.3778,
      "step": 461
    },
    {
      "epoch": 0.462129974055203,
      "grad_norm": 0.22086897492408752,
      "learning_rate": 0.00018850059397216876,
      "loss": 0.3322,
      "step": 462
    },
    {
      "epoch": 0.46313025538432684,
      "grad_norm": 0.20893535017967224,
      "learning_rate": 0.00018845174115833099,
      "loss": 0.3278,
      "step": 463
    },
    {
      "epoch": 0.46413053671345067,
      "grad_norm": 0.20041684806346893,
      "learning_rate": 0.0001884027911520262,
      "loss": 0.3055,
      "step": 464
    },
    {
      "epoch": 0.4651308180425745,
      "grad_norm": 0.30201271176338196,
      "learning_rate": 0.00018835374400704154,
      "loss": 0.3257,
      "step": 465
    },
    {
      "epoch": 0.4661310993716983,
      "grad_norm": 0.23991478979587555,
      "learning_rate": 0.00018830459977727096,
      "loss": 0.3396,
      "step": 466
    },
    {
      "epoch": 0.4671313807008221,
      "grad_norm": 0.22463761270046234,
      "learning_rate": 0.0001882553585167151,
      "loss": 0.2841,
      "step": 467
    },
    {
      "epoch": 0.46813166202994594,
      "grad_norm": 0.24309033155441284,
      "learning_rate": 0.00018820602027948114,
      "loss": 0.332,
      "step": 468
    },
    {
      "epoch": 0.4691319433590697,
      "grad_norm": 0.24418720602989197,
      "learning_rate": 0.00018815658511978298,
      "loss": 0.3274,
      "step": 469
    },
    {
      "epoch": 0.47013222468819355,
      "grad_norm": 0.2473025619983673,
      "learning_rate": 0.00018810705309194083,
      "loss": 0.3641,
      "step": 470
    },
    {
      "epoch": 0.4711325060173174,
      "grad_norm": 0.22359922528266907,
      "learning_rate": 0.00018805742425038145,
      "loss": 0.3158,
      "step": 471
    },
    {
      "epoch": 0.4721327873464412,
      "grad_norm": 0.25090882182121277,
      "learning_rate": 0.00018800769864963802,
      "loss": 0.3622,
      "step": 472
    },
    {
      "epoch": 0.473133068675565,
      "grad_norm": 0.22978092730045319,
      "learning_rate": 0.00018795787634434994,
      "loss": 0.3546,
      "step": 473
    },
    {
      "epoch": 0.4741333500046888,
      "grad_norm": 0.2290024757385254,
      "learning_rate": 0.0001879079573892629,
      "loss": 0.3277,
      "step": 474
    },
    {
      "epoch": 0.47513363133381264,
      "grad_norm": 0.2646024525165558,
      "learning_rate": 0.00018785794183922883,
      "loss": 0.3446,
      "step": 475
    },
    {
      "epoch": 0.4761339126629365,
      "grad_norm": 0.3361208140850067,
      "learning_rate": 0.00018780782974920572,
      "loss": 0.3321,
      "step": 476
    },
    {
      "epoch": 0.47713419399206025,
      "grad_norm": 0.23654255270957947,
      "learning_rate": 0.00018775762117425777,
      "loss": 0.3294,
      "step": 477
    },
    {
      "epoch": 0.4781344753211841,
      "grad_norm": 0.266557902097702,
      "learning_rate": 0.0001877073161695551,
      "loss": 0.363,
      "step": 478
    },
    {
      "epoch": 0.4791347566503079,
      "grad_norm": 0.32463106513023376,
      "learning_rate": 0.00018765691479037376,
      "loss": 0.3751,
      "step": 479
    },
    {
      "epoch": 0.4801350379794317,
      "grad_norm": 0.24274414777755737,
      "learning_rate": 0.00018760641709209583,
      "loss": 0.3638,
      "step": 480
    },
    {
      "epoch": 0.4811353193085555,
      "grad_norm": 0.2479117512702942,
      "learning_rate": 0.0001875558231302091,
      "loss": 0.3269,
      "step": 481
    },
    {
      "epoch": 0.48213560063767935,
      "grad_norm": 0.24928362667560577,
      "learning_rate": 0.00018750513296030718,
      "loss": 0.3452,
      "step": 482
    },
    {
      "epoch": 0.4831358819668032,
      "grad_norm": 0.2932121753692627,
      "learning_rate": 0.00018745434663808942,
      "loss": 0.4229,
      "step": 483
    },
    {
      "epoch": 0.48413616329592696,
      "grad_norm": 0.3067155182361603,
      "learning_rate": 0.0001874034642193608,
      "loss": 0.3521,
      "step": 484
    },
    {
      "epoch": 0.4851364446250508,
      "grad_norm": 0.29176706075668335,
      "learning_rate": 0.0001873524857600319,
      "loss": 0.3618,
      "step": 485
    },
    {
      "epoch": 0.4861367259541746,
      "grad_norm": 0.22693835198879242,
      "learning_rate": 0.00018730141131611882,
      "loss": 0.3389,
      "step": 486
    },
    {
      "epoch": 0.48713700728329845,
      "grad_norm": 0.2182817906141281,
      "learning_rate": 0.00018725024094374315,
      "loss": 0.3478,
      "step": 487
    },
    {
      "epoch": 0.4881372886124222,
      "grad_norm": 0.2273852825164795,
      "learning_rate": 0.00018719897469913184,
      "loss": 0.3236,
      "step": 488
    },
    {
      "epoch": 0.48913756994154606,
      "grad_norm": 0.24996232986450195,
      "learning_rate": 0.00018714761263861728,
      "loss": 0.3005,
      "step": 489
    },
    {
      "epoch": 0.4901378512706699,
      "grad_norm": 0.47928324341773987,
      "learning_rate": 0.000187096154818637,
      "loss": 0.3412,
      "step": 490
    },
    {
      "epoch": 0.4911381325997937,
      "grad_norm": 0.22586360573768616,
      "learning_rate": 0.00018704460129573391,
      "loss": 0.3113,
      "step": 491
    },
    {
      "epoch": 0.4921384139289175,
      "grad_norm": 0.23548360168933868,
      "learning_rate": 0.00018699295212655596,
      "loss": 0.3246,
      "step": 492
    },
    {
      "epoch": 0.4931386952580413,
      "grad_norm": 0.26637163758277893,
      "learning_rate": 0.00018694120736785632,
      "loss": 0.3434,
      "step": 493
    },
    {
      "epoch": 0.49413897658716516,
      "grad_norm": 0.2589665949344635,
      "learning_rate": 0.00018688936707649304,
      "loss": 0.3459,
      "step": 494
    },
    {
      "epoch": 0.49513925791628893,
      "grad_norm": 0.24287012219429016,
      "learning_rate": 0.00018683743130942928,
      "loss": 0.3509,
      "step": 495
    },
    {
      "epoch": 0.49613953924541276,
      "grad_norm": 0.2158239781856537,
      "learning_rate": 0.00018678540012373302,
      "loss": 0.3393,
      "step": 496
    },
    {
      "epoch": 0.4971398205745366,
      "grad_norm": 0.2196134328842163,
      "learning_rate": 0.00018673327357657715,
      "loss": 0.3316,
      "step": 497
    },
    {
      "epoch": 0.4981401019036604,
      "grad_norm": 0.29125913977622986,
      "learning_rate": 0.0001866810517252393,
      "loss": 0.3905,
      "step": 498
    },
    {
      "epoch": 0.4991403832327842,
      "grad_norm": 0.31574273109436035,
      "learning_rate": 0.00018662873462710184,
      "loss": 0.356,
      "step": 499
    },
    {
      "epoch": 0.5001406645619081,
      "grad_norm": 0.2551417946815491,
      "learning_rate": 0.0001865763223396518,
      "loss": 0.354,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.272745519752806e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}