{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.06824198450828513,
  "eval_steps": 348,
  "global_step": 348,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00019609765663300324,
      "grad_norm": 20.33372688293457,
      "learning_rate": 2e-05,
      "loss": 3.0843,
      "step": 1
    },
    {
      "epoch": 0.00019609765663300324,
      "eval_loss": 1.1017773151397705,
      "eval_runtime": 79.9135,
      "eval_samples_per_second": 26.879,
      "eval_steps_per_second": 13.44,
      "step": 1
    },
    {
      "epoch": 0.0003921953132660065,
      "grad_norm": 19.32895278930664,
      "learning_rate": 4e-05,
      "loss": 3.2221,
      "step": 2
    },
    {
      "epoch": 0.0005882929698990097,
      "grad_norm": 18.6882266998291,
      "learning_rate": 6e-05,
      "loss": 3.8951,
      "step": 3
    },
    {
      "epoch": 0.000784390626532013,
      "grad_norm": 43.008060455322266,
      "learning_rate": 8e-05,
      "loss": 5.167,
      "step": 4
    },
    {
      "epoch": 0.0009804882831650162,
      "grad_norm": 21.642993927001953,
      "learning_rate": 0.0001,
      "loss": 3.1304,
      "step": 5
    },
    {
      "epoch": 0.0011765859397980193,
      "grad_norm": 29.79266929626465,
      "learning_rate": 0.00012,
      "loss": 4.5153,
      "step": 6
    },
    {
      "epoch": 0.0013726835964310226,
      "grad_norm": 25.503681182861328,
      "learning_rate": 0.00014,
      "loss": 3.8083,
      "step": 7
    },
    {
      "epoch": 0.001568781253064026,
      "grad_norm": 32.35524368286133,
      "learning_rate": 0.00016,
      "loss": 4.253,
      "step": 8
    },
    {
      "epoch": 0.0017648789096970292,
      "grad_norm": 21.053390502929688,
      "learning_rate": 0.00018,
      "loss": 3.3757,
      "step": 9
    },
    {
      "epoch": 0.0019609765663300325,
      "grad_norm": 25.7067928314209,
      "learning_rate": 0.0002,
      "loss": 3.2484,
      "step": 10
    },
    {
      "epoch": 0.0021570742229630358,
      "grad_norm": 22.57227897644043,
      "learning_rate": 0.00019999974049780868,
      "loss": 2.8378,
      "step": 11
    },
    {
      "epoch": 0.0023531718795960386,
      "grad_norm": 19.06597900390625,
      "learning_rate": 0.00019999896199258152,
      "loss": 3.231,
      "step": 12
    },
    {
      "epoch": 0.002549269536229042,
      "grad_norm": 17.590620040893555,
      "learning_rate": 0.000199997664488359,
      "loss": 2.2391,
      "step": 13
    },
    {
      "epoch": 0.002745367192862045,
      "grad_norm": 8.627043724060059,
      "learning_rate": 0.00019999584799187522,
      "loss": 1.7095,
      "step": 14
    },
    {
      "epoch": 0.0029414648494950485,
      "grad_norm": 21.60858917236328,
      "learning_rate": 0.0001999935125125579,
      "loss": 3.9299,
      "step": 15
    },
    {
      "epoch": 0.003137562506128052,
      "grad_norm": 8.075380325317383,
      "learning_rate": 0.00019999065806252829,
      "loss": 1.7939,
      "step": 16
    },
    {
      "epoch": 0.003333660162761055,
      "grad_norm": 11.393594741821289,
      "learning_rate": 0.00019998728465660105,
      "loss": 1.601,
      "step": 17
    },
    {
      "epoch": 0.0035297578193940584,
      "grad_norm": 8.256339073181152,
      "learning_rate": 0.00019998339231228434,
      "loss": 3.1556,
      "step": 18
    },
    {
      "epoch": 0.0037258554760270617,
      "grad_norm": 20.03615951538086,
      "learning_rate": 0.0001999789810497796,
      "loss": 2.0883,
      "step": 19
    },
    {
      "epoch": 0.003921953132660065,
      "grad_norm": 10.166353225708008,
      "learning_rate": 0.0001999740508919815,
      "loss": 3.5616,
      "step": 20
    },
    {
      "epoch": 0.004118050789293068,
      "grad_norm": 15.80553913116455,
      "learning_rate": 0.0001999686018644777,
      "loss": 3.0344,
      "step": 21
    },
    {
      "epoch": 0.0043141484459260715,
      "grad_norm": 7.451974391937256,
      "learning_rate": 0.00019996263399554897,
      "loss": 2.1049,
      "step": 22
    },
    {
      "epoch": 0.004510246102559075,
      "grad_norm": 5.434274673461914,
      "learning_rate": 0.00019995614731616875,
      "loss": 2.3178,
      "step": 23
    },
    {
      "epoch": 0.004706343759192077,
      "grad_norm": 10.594315528869629,
      "learning_rate": 0.00019994914186000328,
      "loss": 1.7096,
      "step": 24
    },
    {
      "epoch": 0.0049024414158250805,
      "grad_norm": 5.348718166351318,
      "learning_rate": 0.0001999416176634111,
      "loss": 2.695,
      "step": 25
    },
    {
      "epoch": 0.005098539072458084,
      "grad_norm": 17.776073455810547,
      "learning_rate": 0.00019993357476544312,
      "loss": 1.7411,
      "step": 26
    },
    {
      "epoch": 0.005294636729091087,
      "grad_norm": 10.051606178283691,
      "learning_rate": 0.0001999250132078424,
      "loss": 2.6161,
      "step": 27
    },
    {
      "epoch": 0.00549073438572409,
      "grad_norm": 26.03020668029785,
      "learning_rate": 0.00019991593303504376,
      "loss": 3.3977,
      "step": 28
    },
    {
      "epoch": 0.005686832042357094,
      "grad_norm": 10.213540077209473,
      "learning_rate": 0.00019990633429417363,
      "loss": 1.2442,
      "step": 29
    },
    {
      "epoch": 0.005882929698990097,
      "grad_norm": 11.69288444519043,
      "learning_rate": 0.00019989621703505,
      "loss": 1.4702,
      "step": 30
    },
    {
      "epoch": 0.0060790273556231,
      "grad_norm": 4.343452453613281,
      "learning_rate": 0.00019988558131018186,
      "loss": 1.0779,
      "step": 31
    },
    {
      "epoch": 0.006275125012256104,
      "grad_norm": 9.106976509094238,
      "learning_rate": 0.00019987442717476906,
      "loss": 2.5887,
      "step": 32
    },
    {
      "epoch": 0.006471222668889107,
      "grad_norm": 17.658370971679688,
      "learning_rate": 0.00019986275468670205,
      "loss": 2.2258,
      "step": 33
    },
    {
      "epoch": 0.00666732032552211,
      "grad_norm": 6.7451090812683105,
      "learning_rate": 0.00019985056390656162,
      "loss": 1.7206,
      "step": 34
    },
    {
      "epoch": 0.0068634179821551134,
      "grad_norm": 28.07065200805664,
      "learning_rate": 0.00019983785489761837,
      "loss": 2.7356,
      "step": 35
    },
    {
      "epoch": 0.007059515638788117,
      "grad_norm": 11.387879371643066,
      "learning_rate": 0.00019982462772583266,
      "loss": 1.973,
      "step": 36
    },
    {
      "epoch": 0.00725561329542112,
      "grad_norm": 9.64372444152832,
      "learning_rate": 0.00019981088245985408,
      "loss": 2.7339,
      "step": 37
    },
    {
      "epoch": 0.007451710952054123,
      "grad_norm": 9.302544593811035,
      "learning_rate": 0.00019979661917102115,
      "loss": 1.7498,
      "step": 38
    },
    {
      "epoch": 0.007647808608687127,
      "grad_norm": 15.064400672912598,
      "learning_rate": 0.000199781837933361,
      "loss": 3.0109,
      "step": 39
    },
    {
      "epoch": 0.00784390626532013,
      "grad_norm": 7.281099319458008,
      "learning_rate": 0.00019976653882358884,
      "loss": 1.3118,
      "step": 40
    },
    {
      "epoch": 0.008040003921953132,
      "grad_norm": 6.4474873542785645,
      "learning_rate": 0.0001997507219211078,
      "loss": 1.408,
      "step": 41
    },
    {
      "epoch": 0.008236101578586136,
      "grad_norm": 13.101079940795898,
      "learning_rate": 0.00019973438730800822,
      "loss": 2.3367,
      "step": 42
    },
    {
      "epoch": 0.008432199235219139,
      "grad_norm": 5.951049327850342,
      "learning_rate": 0.00019971753506906753,
      "loss": 0.9101,
      "step": 43
    },
    {
      "epoch": 0.008628296891852143,
      "grad_norm": 11.212276458740234,
      "learning_rate": 0.00019970016529174947,
      "loss": 2.7058,
      "step": 44
    },
    {
      "epoch": 0.008824394548485145,
      "grad_norm": 8.68136978149414,
      "learning_rate": 0.0001996822780662041,
      "loss": 2.0276,
      "step": 45
    },
    {
      "epoch": 0.00902049220511815,
      "grad_norm": 17.70038414001465,
      "learning_rate": 0.00019966387348526683,
      "loss": 2.7989,
      "step": 46
    },
    {
      "epoch": 0.009216589861751152,
      "grad_norm": 10.247598648071289,
      "learning_rate": 0.00019964495164445824,
      "loss": 1.9618,
      "step": 47
    },
    {
      "epoch": 0.009412687518384154,
      "grad_norm": 10.378255844116211,
      "learning_rate": 0.0001996255126419835,
      "loss": 1.8003,
      "step": 48
    },
    {
      "epoch": 0.009608785175017159,
      "grad_norm": 31.620820999145508,
      "learning_rate": 0.0001996055565787319,
      "loss": 2.8785,
      "step": 49
    },
    {
      "epoch": 0.009804882831650161,
      "grad_norm": 9.976147651672363,
      "learning_rate": 0.0001995850835582763,
      "loss": 2.5605,
      "step": 50
    },
    {
      "epoch": 0.010000980488283165,
      "grad_norm": 11.751899719238281,
      "learning_rate": 0.00019956409368687258,
      "loss": 2.7556,
      "step": 51
    },
    {
      "epoch": 0.010197078144916168,
      "grad_norm": 15.828932762145996,
      "learning_rate": 0.000199542587073459,
      "loss": 2.7773,
      "step": 52
    },
    {
      "epoch": 0.010393175801549172,
      "grad_norm": 10.772979736328125,
      "learning_rate": 0.00019952056382965597,
      "loss": 1.9553,
      "step": 53
    },
    {
      "epoch": 0.010589273458182174,
      "grad_norm": 10.821427345275879,
      "learning_rate": 0.00019949802406976495,
      "loss": 1.8528,
      "step": 54
    },
    {
      "epoch": 0.010785371114815178,
      "grad_norm": 7.228662490844727,
      "learning_rate": 0.00019947496791076837,
      "loss": 1.1844,
      "step": 55
    },
    {
      "epoch": 0.01098146877144818,
      "grad_norm": 7.164773941040039,
      "learning_rate": 0.00019945139547232872,
      "loss": 1.0291,
      "step": 56
    },
    {
      "epoch": 0.011177566428081185,
      "grad_norm": 13.927733421325684,
      "learning_rate": 0.0001994273068767879,
      "loss": 1.5417,
      "step": 57
    },
    {
      "epoch": 0.011373664084714187,
      "grad_norm": 10.366493225097656,
      "learning_rate": 0.00019940270224916688,
      "loss": 1.5122,
      "step": 58
    },
    {
      "epoch": 0.011569761741347192,
      "grad_norm": 11.2214994430542,
      "learning_rate": 0.00019937758171716468,
      "loss": 1.6003,
      "step": 59
    },
    {
      "epoch": 0.011765859397980194,
      "grad_norm": 14.360090255737305,
      "learning_rate": 0.000199351945411158,
      "loss": 1.5651,
      "step": 60
    },
    {
      "epoch": 0.011961957054613198,
      "grad_norm": 17.97150993347168,
      "learning_rate": 0.00019932579346420038,
      "loss": 1.6064,
      "step": 61
    },
    {
      "epoch": 0.0121580547112462,
      "grad_norm": 10.190518379211426,
      "learning_rate": 0.00019929912601202151,
      "loss": 1.9151,
      "step": 62
    },
    {
      "epoch": 0.012354152367879203,
      "grad_norm": 13.573248863220215,
      "learning_rate": 0.00019927194319302677,
      "loss": 4.0602,
      "step": 63
    },
    {
      "epoch": 0.012550250024512207,
      "grad_norm": 16.919841766357422,
      "learning_rate": 0.00019924424514829606,
      "loss": 2.8292,
      "step": 64
    },
    {
      "epoch": 0.01274634768114521,
      "grad_norm": 58.470252990722656,
      "learning_rate": 0.00019921603202158354,
      "loss": 1.9637,
      "step": 65
    },
    {
      "epoch": 0.012942445337778214,
      "grad_norm": 18.334800720214844,
      "learning_rate": 0.00019918730395931649,
      "loss": 2.5609,
      "step": 66
    },
    {
      "epoch": 0.013138542994411216,
      "grad_norm": 12.280759811401367,
      "learning_rate": 0.00019915806111059486,
      "loss": 1.2495,
      "step": 67
    },
    {
      "epoch": 0.01333464065104422,
      "grad_norm": 8.015874862670898,
      "learning_rate": 0.0001991283036271903,
      "loss": 1.505,
      "step": 68
    },
    {
      "epoch": 0.013530738307677223,
      "grad_norm": 7.713284969329834,
      "learning_rate": 0.0001990980316635455,
      "loss": 2.3898,
      "step": 69
    },
    {
      "epoch": 0.013726835964310227,
      "grad_norm": 18.01800537109375,
      "learning_rate": 0.00019906724537677316,
      "loss": 3.0263,
      "step": 70
    },
    {
      "epoch": 0.01392293362094323,
      "grad_norm": 21.270421981811523,
      "learning_rate": 0.00019903594492665558,
      "loss": 3.2547,
      "step": 71
    },
    {
      "epoch": 0.014119031277576233,
      "grad_norm": 21.60205841064453,
      "learning_rate": 0.0001990041304756434,
      "loss": 2.577,
      "step": 72
    },
    {
      "epoch": 0.014315128934209236,
      "grad_norm": 10.01419734954834,
      "learning_rate": 0.00019897180218885507,
      "loss": 1.9092,
      "step": 73
    },
    {
      "epoch": 0.01451122659084224,
      "grad_norm": 14.10943603515625,
      "learning_rate": 0.00019893896023407578,
      "loss": 2.2377,
      "step": 74
    },
    {
      "epoch": 0.014707324247475242,
      "grad_norm": 11.310667037963867,
      "learning_rate": 0.0001989056047817567,
      "loss": 1.6645,
      "step": 75
    },
    {
      "epoch": 0.014903421904108247,
      "grad_norm": 6.586666107177734,
      "learning_rate": 0.0001988717360050141,
      "loss": 2.2651,
      "step": 76
    },
    {
      "epoch": 0.015099519560741249,
      "grad_norm": 4.402716159820557,
      "learning_rate": 0.00019883735407962846,
      "loss": 1.3483,
      "step": 77
    },
    {
      "epoch": 0.015295617217374253,
      "grad_norm": 9.384387016296387,
      "learning_rate": 0.00019880245918404342,
      "loss": 2.6391,
      "step": 78
    },
    {
      "epoch": 0.015491714874007256,
      "grad_norm": 6.753894329071045,
      "learning_rate": 0.000198767051499365,
      "loss": 2.9391,
      "step": 79
    },
    {
      "epoch": 0.01568781253064026,
      "grad_norm": 6.399787902832031,
      "learning_rate": 0.00019873113120936074,
      "loss": 3.7452,
      "step": 80
    },
    {
      "epoch": 0.01588391018727326,
      "grad_norm": 8.880107879638672,
      "learning_rate": 0.00019869469850045842,
      "loss": 1.2771,
      "step": 81
    },
    {
      "epoch": 0.016080007843906265,
      "grad_norm": 12.630661964416504,
      "learning_rate": 0.00019865775356174545,
      "loss": 2.2072,
      "step": 82
    },
    {
      "epoch": 0.01627610550053927,
      "grad_norm": 7.974503993988037,
      "learning_rate": 0.00019862029658496762,
      "loss": 1.9795,
      "step": 83
    },
    {
      "epoch": 0.016472203157172273,
      "grad_norm": 50.43594741821289,
      "learning_rate": 0.00019858232776452837,
      "loss": 1.5331,
      "step": 84
    },
    {
      "epoch": 0.016668300813805274,
      "grad_norm": 7.273484230041504,
      "learning_rate": 0.00019854384729748746,
      "loss": 2.4005,
      "step": 85
    },
    {
      "epoch": 0.016864398470438278,
      "grad_norm": 5.826492786407471,
      "learning_rate": 0.00019850485538356027,
      "loss": 2.1915,
      "step": 86
    },
    {
      "epoch": 0.017060496127071282,
      "grad_norm": 9.881019592285156,
      "learning_rate": 0.0001984653522251165,
      "loss": 2.3309,
      "step": 87
    },
    {
      "epoch": 0.017256593783704286,
      "grad_norm": 9.147713661193848,
      "learning_rate": 0.00019842533802717923,
      "loss": 1.1404,
      "step": 88
    },
    {
      "epoch": 0.017452691440337287,
      "grad_norm": 13.98263931274414,
      "learning_rate": 0.00019838481299742398,
      "loss": 1.2166,
      "step": 89
    },
    {
      "epoch": 0.01764878909697029,
      "grad_norm": 8.206791877746582,
      "learning_rate": 0.0001983437773461774,
      "loss": 2.6039,
      "step": 90
    },
    {
      "epoch": 0.017844886753603295,
      "grad_norm": 10.445443153381348,
      "learning_rate": 0.00019830223128641637,
      "loss": 2.3554,
      "step": 91
    },
    {
      "epoch": 0.0180409844102363,
      "grad_norm": 11.756292343139648,
      "learning_rate": 0.00019826017503376666,
      "loss": 1.7371,
      "step": 92
    },
    {
      "epoch": 0.0182370820668693,
      "grad_norm": 7.509032249450684,
      "learning_rate": 0.00019821760880650214,
      "loss": 1.389,
      "step": 93
    },
    {
      "epoch": 0.018433179723502304,
      "grad_norm": 8.619280815124512,
      "learning_rate": 0.00019817453282554333,
      "loss": 1.6818,
      "step": 94
    },
    {
      "epoch": 0.01862927738013531,
      "grad_norm": 9.11640739440918,
      "learning_rate": 0.00019813094731445654,
      "loss": 1.631,
      "step": 95
    },
    {
      "epoch": 0.01882537503676831,
      "grad_norm": 14.109521865844727,
      "learning_rate": 0.00019808685249945245,
      "loss": 2.0497,
      "step": 96
    },
    {
      "epoch": 0.019021472693401313,
      "grad_norm": 10.804281234741211,
      "learning_rate": 0.00019804224860938506,
      "loss": 2.2364,
      "step": 97
    },
    {
      "epoch": 0.019217570350034317,
      "grad_norm": 7.363731384277344,
      "learning_rate": 0.0001979971358757505,
      "loss": 1.0967,
      "step": 98
    },
    {
      "epoch": 0.01941366800666732,
      "grad_norm": 15.269912719726562,
      "learning_rate": 0.0001979515145326859,
      "loss": 2.8752,
      "step": 99
    },
    {
      "epoch": 0.019609765663300322,
      "grad_norm": 5.457535266876221,
      "learning_rate": 0.000197905384816968,
      "loss": 1.7098,
      "step": 100
    },
    {
      "epoch": 0.019805863319933326,
      "grad_norm": 4.689967632293701,
      "learning_rate": 0.00019785874696801202,
      "loss": 2.2133,
      "step": 101
    },
    {
      "epoch": 0.02000196097656633,
      "grad_norm": 10.993409156799316,
      "learning_rate": 0.00019781160122787046,
      "loss": 2.314,
      "step": 102
    },
    {
      "epoch": 0.020198058633199335,
      "grad_norm": 8.199251174926758,
      "learning_rate": 0.00019776394784123177,
      "loss": 2.5164,
      "step": 103
    },
    {
      "epoch": 0.020394156289832335,
      "grad_norm": 15.144885063171387,
      "learning_rate": 0.00019771578705541916,
      "loss": 2.0058,
      "step": 104
    },
    {
      "epoch": 0.02059025394646534,
      "grad_norm": 5.252450466156006,
      "learning_rate": 0.00019766711912038915,
      "loss": 1.7012,
      "step": 105
    },
    {
      "epoch": 0.020786351603098344,
      "grad_norm": 8.265049934387207,
      "learning_rate": 0.0001976179442887305,
      "loss": 1.8646,
      "step": 106
    },
    {
      "epoch": 0.020982449259731348,
      "grad_norm": 8.365408897399902,
      "learning_rate": 0.00019756826281566272,
      "loss": 1.9615,
      "step": 107
    },
    {
      "epoch": 0.02117854691636435,
      "grad_norm": 7.514213562011719,
      "learning_rate": 0.00019751807495903484,
      "loss": 1.4897,
      "step": 108
    },
    {
      "epoch": 0.021374644572997353,
      "grad_norm": 15.234655380249023,
      "learning_rate": 0.00019746738097932407,
      "loss": 2.0467,
      "step": 109
    },
    {
      "epoch": 0.021570742229630357,
      "grad_norm": 6.856448650360107,
      "learning_rate": 0.0001974161811396343,
      "loss": 1.4492,
      "step": 110
    },
    {
      "epoch": 0.021766839886263357,
      "grad_norm": 7.893224716186523,
      "learning_rate": 0.00019736447570569503,
      "loss": 1.919,
      "step": 111
    },
    {
      "epoch": 0.02196293754289636,
      "grad_norm": 8.966511726379395,
      "learning_rate": 0.0001973122649458597,
      "loss": 2.4484,
      "step": 112
    },
    {
      "epoch": 0.022159035199529366,
      "grad_norm": 7.631579875946045,
      "learning_rate": 0.00019725954913110442,
      "loss": 1.4992,
      "step": 113
    },
    {
      "epoch": 0.02235513285616237,
      "grad_norm": 7.418518543243408,
      "learning_rate": 0.0001972063285350266,
      "loss": 0.8401,
      "step": 114
    },
    {
      "epoch": 0.02255123051279537,
      "grad_norm": 7.739930629730225,
      "learning_rate": 0.00019715260343384347,
      "loss": 2.0713,
      "step": 115
    },
    {
      "epoch": 0.022747328169428375,
      "grad_norm": 6.441893100738525,
      "learning_rate": 0.00019709837410639063,
      "loss": 1.4438,
      "step": 116
    },
    {
      "epoch": 0.02294342582606138,
      "grad_norm": 6.008083820343018,
      "learning_rate": 0.0001970436408341207,
      "loss": 1.3503,
      "step": 117
    },
    {
      "epoch": 0.023139523482694383,
      "grad_norm": 7.100820541381836,
      "learning_rate": 0.00019698840390110176,
      "loss": 1.4726,
      "step": 118
    },
    {
      "epoch": 0.023335621139327384,
      "grad_norm": 10.213756561279297,
      "learning_rate": 0.0001969326635940159,
      "loss": 0.8107,
      "step": 119
    },
    {
      "epoch": 0.023531718795960388,
      "grad_norm": 5.251387119293213,
      "learning_rate": 0.00019687642020215775,
      "loss": 1.5542,
      "step": 120
    },
    {
      "epoch": 0.023727816452593392,
      "grad_norm": 6.100740432739258,
      "learning_rate": 0.00019681967401743297,
      "loss": 1.2512,
      "step": 121
    },
    {
      "epoch": 0.023923914109226396,
      "grad_norm": 7.356696128845215,
      "learning_rate": 0.00019676242533435678,
      "loss": 2.4725,
      "step": 122
    },
    {
      "epoch": 0.024120011765859397,
      "grad_norm": 11.542431831359863,
      "learning_rate": 0.00019670467445005233,
      "loss": 3.0307,
      "step": 123
    },
    {
      "epoch": 0.0243161094224924,
      "grad_norm": 12.166086196899414,
      "learning_rate": 0.00019664642166424928,
      "loss": 1.2784,
      "step": 124
    },
    {
      "epoch": 0.024512207079125405,
      "grad_norm": 5.222433090209961,
      "learning_rate": 0.00019658766727928206,
      "loss": 1.1759,
      "step": 125
    },
    {
      "epoch": 0.024708304735758406,
      "grad_norm": 4.77174711227417,
      "learning_rate": 0.00019652841160008858,
      "loss": 1.1041,
      "step": 126
    },
    {
      "epoch": 0.02490440239239141,
      "grad_norm": 4.879274368286133,
      "learning_rate": 0.0001964686549342084,
      "loss": 2.6326,
      "step": 127
    },
    {
      "epoch": 0.025100500049024414,
      "grad_norm": 14.171689987182617,
      "learning_rate": 0.00019640839759178116,
      "loss": 3.4144,
      "step": 128
    },
    {
      "epoch": 0.02529659770565742,
      "grad_norm": 7.598373889923096,
      "learning_rate": 0.00019634763988554522,
      "loss": 2.0596,
      "step": 129
    },
    {
      "epoch": 0.02549269536229042,
      "grad_norm": 6.88770866394043,
      "learning_rate": 0.00019628638213083565,
      "loss": 1.4691,
      "step": 130
    },
    {
      "epoch": 0.025688793018923423,
      "grad_norm": 7.128096580505371,
      "learning_rate": 0.00019622462464558295,
      "loss": 1.3307,
      "step": 131
    },
    {
      "epoch": 0.025884890675556427,
      "grad_norm": 6.430881500244141,
      "learning_rate": 0.00019616236775031113,
      "loss": 0.9491,
      "step": 132
    },
    {
      "epoch": 0.02608098833218943,
      "grad_norm": 9.912070274353027,
      "learning_rate": 0.00019609961176813624,
      "loss": 2.5006,
      "step": 133
    },
    {
      "epoch": 0.026277085988822432,
      "grad_norm": 8.550467491149902,
      "learning_rate": 0.0001960363570247645,
      "loss": 2.4952,
      "step": 134
    },
    {
      "epoch": 0.026473183645455436,
      "grad_norm": 4.201476573944092,
      "learning_rate": 0.0001959726038484909,
      "loss": 0.9033,
      "step": 135
    },
    {
      "epoch": 0.02666928130208844,
      "grad_norm": 5.774847984313965,
      "learning_rate": 0.00019590835257019714,
      "loss": 2.1291,
      "step": 136
    },
    {
      "epoch": 0.026865378958721445,
      "grad_norm": 8.179195404052734,
      "learning_rate": 0.00019584360352335023,
      "loss": 2.7527,
      "step": 137
    },
    {
      "epoch": 0.027061476615354445,
      "grad_norm": 15.658841133117676,
      "learning_rate": 0.0001957783570440005,
      "loss": 1.8304,
      "step": 138
    },
    {
      "epoch": 0.02725757427198745,
      "grad_norm": 5.7399163246154785,
      "learning_rate": 0.0001957126134707801,
      "loss": 1.7071,
      "step": 139
    },
    {
      "epoch": 0.027453671928620454,
      "grad_norm": 5.0817389488220215,
      "learning_rate": 0.00019564637314490108,
      "loss": 1.8933,
      "step": 140
    },
    {
      "epoch": 0.027649769585253458,
      "grad_norm": 5.634946346282959,
      "learning_rate": 0.0001955796364101535,
      "loss": 1.7343,
      "step": 141
    },
    {
      "epoch": 0.02784586724188646,
      "grad_norm": 6.406938552856445,
      "learning_rate": 0.00019551240361290407,
      "loss": 2.3013,
      "step": 142
    },
    {
      "epoch": 0.028041964898519463,
      "grad_norm": 8.239458084106445,
      "learning_rate": 0.00019544467510209388,
      "loss": 1.2177,
      "step": 143
    },
    {
      "epoch": 0.028238062555152467,
      "grad_norm": 11.887965202331543,
      "learning_rate": 0.0001953764512292369,
      "loss": 2.4312,
      "step": 144
    },
    {
      "epoch": 0.028434160211785468,
      "grad_norm": 7.482359409332275,
      "learning_rate": 0.00019530773234841803,
      "loss": 1.1083,
      "step": 145
    },
    {
      "epoch": 0.028630257868418472,
      "grad_norm": 8.86729621887207,
      "learning_rate": 0.00019523851881629126,
      "loss": 1.6451,
      "step": 146
    },
    {
      "epoch": 0.028826355525051476,
      "grad_norm": 7.395509719848633,
      "learning_rate": 0.0001951688109920778,
      "loss": 1.31,
      "step": 147
    },
    {
      "epoch": 0.02902245318168448,
      "grad_norm": 4.955163955688477,
      "learning_rate": 0.00019509860923756442,
      "loss": 2.5206,
      "step": 148
    },
    {
      "epoch": 0.02921855083831748,
      "grad_norm": 5.034746170043945,
      "learning_rate": 0.00019502791391710125,
      "loss": 0.9336,
      "step": 149
    },
    {
      "epoch": 0.029414648494950485,
      "grad_norm": 12.375234603881836,
      "learning_rate": 0.00019495672539760007,
      "loss": 2.1276,
      "step": 150
    },
    {
      "epoch": 0.02961074615158349,
      "grad_norm": 5.832932949066162,
      "learning_rate": 0.00019488504404853248,
      "loss": 1.3252,
      "step": 151
    },
    {
      "epoch": 0.029806843808216493,
      "grad_norm": 5.934417724609375,
      "learning_rate": 0.00019481287024192775,
      "loss": 1.5907,
      "step": 152
    },
    {
      "epoch": 0.030002941464849494,
      "grad_norm": 9.238896369934082,
      "learning_rate": 0.00019474020435237117,
      "loss": 1.1184,
      "step": 153
    },
    {
      "epoch": 0.030199039121482498,
      "grad_norm": 9.787931442260742,
      "learning_rate": 0.00019466704675700185,
      "loss": 1.4931,
      "step": 154
    },
    {
      "epoch": 0.030395136778115502,
      "grad_norm": 7.260796070098877,
      "learning_rate": 0.00019459339783551094,
      "loss": 0.8924,
      "step": 155
    },
    {
      "epoch": 0.030591234434748506,
      "grad_norm": 8.712836265563965,
      "learning_rate": 0.00019451925797013954,
      "loss": 1.586,
      "step": 156
    },
    {
      "epoch": 0.030787332091381507,
      "grad_norm": 11.15104866027832,
      "learning_rate": 0.00019444462754567682,
      "loss": 1.5007,
      "step": 157
    },
    {
      "epoch": 0.03098342974801451,
      "grad_norm": 7.158255100250244,
      "learning_rate": 0.00019436950694945798,
      "loss": 2.4118,
      "step": 158
    },
    {
      "epoch": 0.031179527404647515,
      "grad_norm": 11.58385944366455,
      "learning_rate": 0.00019429389657136213,
      "loss": 2.1638,
      "step": 159
    },
    {
      "epoch": 0.03137562506128052,
      "grad_norm": 7.469117641448975,
      "learning_rate": 0.00019421779680381054,
      "loss": 3.0682,
      "step": 160
    },
    {
      "epoch": 0.031571722717913524,
      "grad_norm": 10.78966999053955,
      "learning_rate": 0.00019414120804176426,
      "loss": 1.1822,
      "step": 161
    },
    {
      "epoch": 0.03176782037454652,
      "grad_norm": 9.68694019317627,
      "learning_rate": 0.00019406413068272238,
      "loss": 2.5351,
      "step": 162
    },
    {
      "epoch": 0.031963918031179525,
      "grad_norm": 11.67428970336914,
      "learning_rate": 0.00019398656512671972,
      "loss": 1.9244,
      "step": 163
    },
    {
      "epoch": 0.03216001568781253,
      "grad_norm": 12.72513198852539,
      "learning_rate": 0.00019390851177632497,
      "loss": 3.2138,
      "step": 164
    },
    {
      "epoch": 0.03235611334444553,
      "grad_norm": 8.345921516418457,
      "learning_rate": 0.00019382997103663838,
      "loss": 2.6435,
      "step": 165
    },
    {
      "epoch": 0.03255221100107854,
      "grad_norm": 7.740304470062256,
      "learning_rate": 0.0001937509433152899,
      "loss": 0.8189,
      "step": 166
    },
    {
      "epoch": 0.03274830865771154,
      "grad_norm": 9.329862594604492,
      "learning_rate": 0.0001936714290224368,
      "loss": 1.4106,
      "step": 167
    },
    {
      "epoch": 0.032944406314344546,
      "grad_norm": 7.179844379425049,
      "learning_rate": 0.00019359142857076176,
      "loss": 1.8125,
      "step": 168
    },
    {
      "epoch": 0.03314050397097755,
      "grad_norm": 7.835447311401367,
      "learning_rate": 0.00019351094237547066,
      "loss": 1.6617,
      "step": 169
    },
    {
      "epoch": 0.03333660162761055,
      "grad_norm": 6.018518924713135,
      "learning_rate": 0.0001934299708542904,
      "loss": 2.4333,
      "step": 170
    },
    {
      "epoch": 0.03353269928424355,
      "grad_norm": 8.176468849182129,
      "learning_rate": 0.00019334851442746664,
      "loss": 2.5915,
      "step": 171
    },
    {
      "epoch": 0.033728796940876556,
      "grad_norm": 8.241739273071289,
      "learning_rate": 0.00019326657351776186,
      "loss": 1.666,
      "step": 172
    },
    {
      "epoch": 0.03392489459750956,
      "grad_norm": 8.064835548400879,
      "learning_rate": 0.000193184148550453,
      "loss": 1.477,
      "step": 173
    },
    {
      "epoch": 0.034120992254142564,
      "grad_norm": 5.790217399597168,
      "learning_rate": 0.00019310123995332917,
      "loss": 0.7703,
      "step": 174
    },
    {
      "epoch": 0.03431708991077557,
      "grad_norm": 9.38430118560791,
      "learning_rate": 0.00019301784815668974,
      "loss": 1.5785,
      "step": 175
    },
    {
      "epoch": 0.03451318756740857,
      "grad_norm": 8.252826690673828,
      "learning_rate": 0.00019293397359334167,
      "loss": 2.1462,
      "step": 176
    },
    {
      "epoch": 0.03470928522404157,
      "grad_norm": 12.65652847290039,
      "learning_rate": 0.00019284961669859766,
      "loss": 1.3009,
      "step": 177
    },
    {
      "epoch": 0.034905382880674574,
      "grad_norm": 6.8490753173828125,
      "learning_rate": 0.00019276477791027374,
      "loss": 2.4905,
      "step": 178
    },
    {
      "epoch": 0.03510148053730758,
      "grad_norm": 4.2581048011779785,
      "learning_rate": 0.0001926794576686869,
      "loss": 0.9042,
      "step": 179
    },
    {
      "epoch": 0.03529757819394058,
      "grad_norm": 6.415445327758789,
      "learning_rate": 0.0001925936564166529,
      "loss": 2.238,
      "step": 180
    },
    {
      "epoch": 0.035493675850573586,
      "grad_norm": 13.620756149291992,
      "learning_rate": 0.00019250737459948405,
      "loss": 1.5966,
      "step": 181
    },
    {
      "epoch": 0.03568977350720659,
      "grad_norm": 10.609662055969238,
      "learning_rate": 0.00019242061266498675,
      "loss": 1.081,
      "step": 182
    },
    {
      "epoch": 0.035885871163839594,
      "grad_norm": 8.404073715209961,
      "learning_rate": 0.00019233337106345925,
      "loss": 1.849,
      "step": 183
    },
    {
      "epoch": 0.0360819688204726,
      "grad_norm": 5.560455322265625,
      "learning_rate": 0.00019224565024768926,
      "loss": 1.4533,
      "step": 184
    },
    {
      "epoch": 0.036278066477105596,
      "grad_norm": 7.896220684051514,
      "learning_rate": 0.00019215745067295169,
      "loss": 2.482,
      "step": 185
    },
    {
      "epoch": 0.0364741641337386,
      "grad_norm": 9.554024696350098,
      "learning_rate": 0.00019206877279700612,
      "loss": 1.9367,
      "step": 186
    },
    {
      "epoch": 0.036670261790371604,
      "grad_norm": 3.333113193511963,
      "learning_rate": 0.00019197961708009473,
      "loss": 1.1477,
      "step": 187
    },
    {
      "epoch": 0.03686635944700461,
      "grad_norm": 9.468240737915039,
      "learning_rate": 0.00019188998398493953,
      "loss": 1.0849,
      "step": 188
    },
    {
      "epoch": 0.03706245710363761,
      "grad_norm": 10.807921409606934,
      "learning_rate": 0.00019179987397674022,
      "loss": 2.0192,
      "step": 189
    },
    {
      "epoch": 0.03725855476027062,
      "grad_norm": 7.14724588394165,
      "learning_rate": 0.0001917092875231717,
      "loss": 2.1502,
      "step": 190
    },
    {
      "epoch": 0.03745465241690362,
      "grad_norm": 12.262707710266113,
      "learning_rate": 0.00019161822509438162,
      "loss": 2.423,
      "step": 191
    },
    {
      "epoch": 0.03765075007353662,
      "grad_norm": 35.0489387512207,
      "learning_rate": 0.000191526687162988,
      "loss": 2.5959,
      "step": 192
    },
    {
      "epoch": 0.03784684773016962,
      "grad_norm": 6.615735054016113,
      "learning_rate": 0.0001914346742040767,
      "loss": 1.7733,
      "step": 193
    },
    {
      "epoch": 0.038042945386802626,
      "grad_norm": 4.537426471710205,
      "learning_rate": 0.00019134218669519896,
      "loss": 1.0028,
      "step": 194
    },
    {
      "epoch": 0.03823904304343563,
      "grad_norm": 5.247801303863525,
      "learning_rate": 0.00019124922511636912,
      "loss": 0.8412,
      "step": 195
    },
    {
      "epoch": 0.038435140700068635,
      "grad_norm": 6.2183918952941895,
      "learning_rate": 0.00019115578995006173,
      "loss": 1.7212,
      "step": 196
    },
    {
      "epoch": 0.03863123835670164,
      "grad_norm": 9.330825805664062,
      "learning_rate": 0.00019106188168120948,
      "loss": 1.5341,
      "step": 197
    },
    {
      "epoch": 0.03882733601333464,
      "grad_norm": 9.86260986328125,
      "learning_rate": 0.00019096750079720037,
      "loss": 2.765,
      "step": 198
    },
    {
      "epoch": 0.03902343366996765,
      "grad_norm": 10.341052055358887,
      "learning_rate": 0.00019087264778787534,
      "loss": 1.9024,
      "step": 199
    },
    {
      "epoch": 0.039219531326600644,
      "grad_norm": 9.549159049987793,
      "learning_rate": 0.00019077732314552566,
      "loss": 1.2644,
      "step": 200
    },
    {
      "epoch": 0.03941562898323365,
      "grad_norm": 5.25094747543335,
      "learning_rate": 0.00019068152736489036,
      "loss": 1.334,
      "step": 201
    },
    {
      "epoch": 0.03961172663986665,
      "grad_norm": 7.197662830352783,
      "learning_rate": 0.00019058526094315378,
      "loss": 1.9093,
      "step": 202
    },
    {
      "epoch": 0.03980782429649966,
      "grad_norm": 8.476766586303711,
      "learning_rate": 0.0001904885243799429,
      "loss": 1.477,
      "step": 203
    },
    {
      "epoch": 0.04000392195313266,
      "grad_norm": 8.232537269592285,
      "learning_rate": 0.00019039131817732462,
      "loss": 1.4013,
      "step": 204
    },
    {
      "epoch": 0.040200019609765665,
      "grad_norm": 15.687997817993164,
      "learning_rate": 0.0001902936428398035,
      "loss": 1.6772,
      "step": 205
    },
    {
      "epoch": 0.04039611726639867,
      "grad_norm": 7.573246479034424,
      "learning_rate": 0.00019019549887431877,
      "loss": 1.5007,
      "step": 206
    },
    {
      "epoch": 0.040592214923031666,
      "grad_norm": 11.531679153442383,
      "learning_rate": 0.0001900968867902419,
      "loss": 2.6798,
      "step": 207
    },
    {
      "epoch": 0.04078831257966467,
      "grad_norm": 6.225399494171143,
      "learning_rate": 0.00018999780709937398,
      "loss": 1.3078,
      "step": 208
    },
    {
      "epoch": 0.040984410236297675,
      "grad_norm": 10.358306884765625,
      "learning_rate": 0.0001898982603159429,
      "loss": 1.7353,
      "step": 209
    },
    {
      "epoch": 0.04118050789293068,
      "grad_norm": 8.146821975708008,
      "learning_rate": 0.00018979824695660087,
      "loss": 1.415,
      "step": 210
    },
    {
      "epoch": 0.04137660554956368,
      "grad_norm": 4.390834808349609,
      "learning_rate": 0.00018969776754042156,
      "loss": 1.7612,
      "step": 211
    },
    {
      "epoch": 0.04157270320619669,
      "grad_norm": 7.958174228668213,
      "learning_rate": 0.0001895968225888976,
      "loss": 2.6614,
      "step": 212
    },
    {
      "epoch": 0.04176880086282969,
      "grad_norm": 9.981225967407227,
      "learning_rate": 0.00018949541262593762,
      "loss": 2.0158,
      "step": 213
    },
    {
      "epoch": 0.041964898519462696,
      "grad_norm": 4.456605911254883,
      "learning_rate": 0.00018939353817786387,
      "loss": 1.0621,
      "step": 214
    },
    {
      "epoch": 0.04216099617609569,
      "grad_norm": 7.546274662017822,
      "learning_rate": 0.00018929119977340917,
      "loss": 1.7333,
      "step": 215
    },
    {
      "epoch": 0.0423570938327287,
      "grad_norm": 11.629569053649902,
      "learning_rate": 0.0001891883979437143,
      "loss": 1.4268,
      "step": 216
    },
    {
      "epoch": 0.0425531914893617,
      "grad_norm": 17.710948944091797,
      "learning_rate": 0.00018908513322232528,
      "loss": 2.8701,
      "step": 217
    },
    {
      "epoch": 0.042749289145994705,
      "grad_norm": 6.267049789428711,
      "learning_rate": 0.00018898140614519054,
      "loss": 1.6313,
      "step": 218
    },
    {
      "epoch": 0.04294538680262771,
      "grad_norm": 4.971591949462891,
      "learning_rate": 0.00018887721725065814,
      "loss": 2.0962,
      "step": 219
    },
    {
      "epoch": 0.043141484459260714,
      "grad_norm": 5.603585243225098,
      "learning_rate": 0.00018877256707947306,
      "loss": 0.6683,
      "step": 220
    },
    {
      "epoch": 0.04333758211589372,
      "grad_norm": 6.029137134552002,
      "learning_rate": 0.00018866745617477423,
      "loss": 1.5375,
      "step": 221
    },
    {
      "epoch": 0.043533679772526715,
      "grad_norm": 7.4105143547058105,
      "learning_rate": 0.00018856188508209183,
      "loss": 1.9524,
      "step": 222
    },
    {
      "epoch": 0.04372977742915972,
      "grad_norm": 8.321500778198242,
      "learning_rate": 0.00018845585434934452,
      "loss": 2.1109,
      "step": 223
    },
    {
      "epoch": 0.04392587508579272,
      "grad_norm": 9.238992691040039,
      "learning_rate": 0.00018834936452683638,
      "loss": 1.4247,
      "step": 224
    },
    {
      "epoch": 0.04412197274242573,
      "grad_norm": 5.125700950622559,
      "learning_rate": 0.00018824241616725434,
      "loss": 1.1266,
      "step": 225
    },
    {
      "epoch": 0.04431807039905873,
      "grad_norm": 7.538069725036621,
      "learning_rate": 0.000188135009825665,
      "loss": 2.1554,
      "step": 226
    },
    {
      "epoch": 0.044514168055691736,
      "grad_norm": 8.309137344360352,
      "learning_rate": 0.00018802714605951199,
      "loss": 1.1435,
      "step": 227
    },
    {
      "epoch": 0.04471026571232474,
      "grad_norm": 22.02942657470703,
      "learning_rate": 0.00018791882542861302,
      "loss": 1.8154,
      "step": 228
    },
    {
      "epoch": 0.044906363368957744,
      "grad_norm": 7.017299652099609,
      "learning_rate": 0.0001878100484951569,
      "loss": 1.4998,
      "step": 229
    },
    {
      "epoch": 0.04510246102559074,
      "grad_norm": 18.39406394958496,
      "learning_rate": 0.00018770081582370068,
      "loss": 2.1662,
      "step": 230
    },
    {
      "epoch": 0.045298558682223745,
      "grad_norm": 9.11802864074707,
      "learning_rate": 0.0001875911279811667,
      "loss": 0.7446,
      "step": 231
    },
    {
      "epoch": 0.04549465633885675,
      "grad_norm": 7.193735122680664,
      "learning_rate": 0.00018748098553683968,
      "loss": 1.9472,
      "step": 232
    },
    {
      "epoch": 0.045690753995489754,
      "grad_norm": 23.407245635986328,
      "learning_rate": 0.0001873703890623637,
      "loss": 2.1782,
      "step": 233
    },
    {
      "epoch": 0.04588685165212276,
      "grad_norm": 6.547053813934326,
      "learning_rate": 0.00018725933913173938,
      "loss": 1.9687,
      "step": 234
    },
    {
      "epoch": 0.04608294930875576,
      "grad_norm": 10.576699256896973,
      "learning_rate": 0.00018714783632132068,
      "loss": 1.8832,
      "step": 235
    },
    {
      "epoch": 0.046279046965388766,
      "grad_norm": 5.852027416229248,
      "learning_rate": 0.00018703588120981207,
      "loss": 1.8932,
      "step": 236
    },
    {
      "epoch": 0.04647514462202176,
      "grad_norm": 7.023755073547363,
      "learning_rate": 0.00018692347437826548,
      "loss": 3.7953,
      "step": 237
    },
    {
      "epoch": 0.04667124227865477,
      "grad_norm": 13.61612606048584,
      "learning_rate": 0.00018681061641007737,
      "loss": 1.9077,
      "step": 238
    },
    {
      "epoch": 0.04686733993528777,
      "grad_norm": 5.3344526290893555,
      "learning_rate": 0.0001866973078909854,
      "loss": 1.4342,
      "step": 239
    },
    {
      "epoch": 0.047063437591920776,
      "grad_norm": 38.80408477783203,
      "learning_rate": 0.00018658354940906586,
      "loss": 2.3665,
      "step": 240
    },
    {
      "epoch": 0.04725953524855378,
      "grad_norm": 9.670344352722168,
      "learning_rate": 0.00018646934155473022,
      "loss": 0.9006,
      "step": 241
    },
    {
      "epoch": 0.047455632905186784,
      "grad_norm": 5.1102495193481445,
      "learning_rate": 0.00018635468492072228,
      "loss": 1.2289,
      "step": 242
    },
    {
      "epoch": 0.04765173056181979,
      "grad_norm": 9.1209077835083,
      "learning_rate": 0.00018623958010211493,
      "loss": 1.6009,
      "step": 243
    },
    {
      "epoch": 0.04784782821845279,
      "grad_norm": 16.793027877807617,
      "learning_rate": 0.0001861240276963073,
      "loss": 0.94,
      "step": 244
    },
    {
      "epoch": 0.04804392587508579,
      "grad_norm": 6.90054988861084,
      "learning_rate": 0.00018600802830302134,
      "loss": 1.559,
      "step": 245
    },
    {
      "epoch": 0.048240023531718794,
      "grad_norm": 13.111268043518066,
      "learning_rate": 0.0001858915825242991,
      "loss": 2.1186,
      "step": 246
    },
    {
      "epoch": 0.0484361211883518,
      "grad_norm": 6.356579780578613,
      "learning_rate": 0.00018577469096449925,
      "loss": 1.6653,
      "step": 247
    },
    {
      "epoch": 0.0486322188449848,
      "grad_norm": 9.505541801452637,
      "learning_rate": 0.00018565735423029404,
      "loss": 0.9774,
      "step": 248
    },
    {
      "epoch": 0.048828316501617806,
      "grad_norm": 8.927581787109375,
      "learning_rate": 0.00018553957293066632,
      "loss": 2.6455,
      "step": 249
    },
    {
      "epoch": 0.04902441415825081,
      "grad_norm": 7.568793773651123,
      "learning_rate": 0.00018542134767690616,
      "loss": 1.1464,
      "step": 250
    },
    {
      "epoch": 0.049220511814883815,
      "grad_norm": 7.632232189178467,
      "learning_rate": 0.00018530267908260784,
      "loss": 1.2671,
      "step": 251
    },
    {
      "epoch": 0.04941660947151681,
      "grad_norm": 4.4279561042785645,
      "learning_rate": 0.00018518356776366657,
      "loss": 2.0384,
      "step": 252
    },
    {
      "epoch": 0.049612707128149816,
      "grad_norm": 10.818602561950684,
      "learning_rate": 0.00018506401433827528,
      "loss": 1.0559,
      "step": 253
    },
    {
      "epoch": 0.04980880478478282,
      "grad_norm": 5.57148551940918,
      "learning_rate": 0.00018494401942692153,
      "loss": 0.9603,
      "step": 254
    },
    {
      "epoch": 0.050004902441415824,
      "grad_norm": 11.1985502243042,
      "learning_rate": 0.00018482358365238413,
      "loss": 2.4928,
      "step": 255
    },
    {
      "epoch": 0.05020100009804883,
      "grad_norm": 4.890799522399902,
      "learning_rate": 0.00018470270763973004,
      "loss": 1.4034,
      "step": 256
    },
    {
      "epoch": 0.05039709775468183,
      "grad_norm": 6.2078680992126465,
      "learning_rate": 0.00018458139201631108,
      "loss": 1.782,
      "step": 257
    },
    {
      "epoch": 0.05059319541131484,
      "grad_norm": 24.89278221130371,
      "learning_rate": 0.00018445963741176065,
      "loss": 3.7879,
      "step": 258
    },
    {
      "epoch": 0.05078929306794784,
      "grad_norm": 5.363570213317871,
      "learning_rate": 0.00018433744445799045,
      "loss": 1.4292,
      "step": 259
    },
    {
      "epoch": 0.05098539072458084,
      "grad_norm": 7.669764041900635,
      "learning_rate": 0.0001842148137891873,
      "loss": 2.0483,
      "step": 260
    },
    {
      "epoch": 0.05118148838121384,
      "grad_norm": 5.229150295257568,
      "learning_rate": 0.00018409174604180976,
      "loss": 3.2863,
      "step": 261
    },
    {
      "epoch": 0.05137758603784685,
      "grad_norm": 5.850373268127441,
      "learning_rate": 0.0001839682418545848,
      "loss": 1.8197,
      "step": 262
    },
    {
      "epoch": 0.05157368369447985,
      "grad_norm": 7.138283729553223,
      "learning_rate": 0.00018384430186850454,
      "loss": 2.7101,
      "step": 263
    },
    {
      "epoch": 0.051769781351112855,
      "grad_norm": 10.918169975280762,
      "learning_rate": 0.000183719926726823,
      "loss": 1.8243,
      "step": 264
    },
    {
      "epoch": 0.05196587900774586,
      "grad_norm": 9.205517768859863,
      "learning_rate": 0.00018359511707505258,
      "loss": 1.4992,
      "step": 265
    },
    {
      "epoch": 0.05216197666437886,
      "grad_norm": 8.567139625549316,
      "learning_rate": 0.00018346987356096086,
      "loss": 1.051,
      "step": 266
    },
    {
      "epoch": 0.05235807432101187,
      "grad_norm": 10.313075065612793,
      "learning_rate": 0.00018334419683456717,
      "loss": 2.6062,
      "step": 267
    },
    {
      "epoch": 0.052554171977644865,
      "grad_norm": 7.515801906585693,
      "learning_rate": 0.0001832180875481392,
      "loss": 1.266,
      "step": 268
    },
    {
      "epoch": 0.05275026963427787,
      "grad_norm": 5.345809459686279,
      "learning_rate": 0.00018309154635618965,
      "loss": 1.2526,
      "step": 269
    },
    {
      "epoch": 0.05294636729091087,
      "grad_norm": 13.568882942199707,
      "learning_rate": 0.00018296457391547296,
      "loss": 2.5183,
      "step": 270
    },
    {
      "epoch": 0.05314246494754388,
      "grad_norm": 10.022235870361328,
      "learning_rate": 0.00018283717088498155,
      "loss": 2.2774,
      "step": 271
    },
    {
      "epoch": 0.05333856260417688,
      "grad_norm": 6.537176132202148,
      "learning_rate": 0.0001827093379259428,
      "loss": 1.4989,
      "step": 272
    },
    {
      "epoch": 0.053534660260809885,
      "grad_norm": 17.213987350463867,
      "learning_rate": 0.00018258107570181533,
      "loss": 2.4885,
      "step": 273
    },
    {
      "epoch": 0.05373075791744289,
      "grad_norm": 6.48647403717041,
      "learning_rate": 0.00018245238487828573,
      "loss": 1.2309,
      "step": 274
    },
    {
      "epoch": 0.05392685557407589,
      "grad_norm": 5.479822158813477,
      "learning_rate": 0.000182323266123265,
      "loss": 1.8959,
      "step": 275
    },
    {
      "epoch": 0.05412295323070889,
      "grad_norm": 7.716124534606934,
      "learning_rate": 0.00018219372010688515,
      "loss": 1.8321,
      "step": 276
    },
    {
      "epoch": 0.054319050887341895,
      "grad_norm": 9.968965530395508,
      "learning_rate": 0.00018206374750149567,
      "loss": 4.1652,
      "step": 277
    },
    {
      "epoch": 0.0545151485439749,
      "grad_norm": 6.009235382080078,
      "learning_rate": 0.00018193334898166007,
      "loss": 0.8178,
      "step": 278
    },
    {
      "epoch": 0.0547112462006079,
      "grad_norm": 8.031886100769043,
      "learning_rate": 0.00018180252522415242,
      "loss": 1.783,
      "step": 279
    },
    {
      "epoch": 0.05490734385724091,
      "grad_norm": 5.5589680671691895,
      "learning_rate": 0.00018167127690795368,
      "loss": 1.3049,
      "step": 280
    },
    {
      "epoch": 0.05510344151387391,
      "grad_norm": 5.04995059967041,
      "learning_rate": 0.0001815396047142485,
      "loss": 0.8962,
      "step": 281
    },
    {
      "epoch": 0.055299539170506916,
      "grad_norm": 5.3526692390441895,
      "learning_rate": 0.0001814075093264212,
      "loss": 1.201,
      "step": 282
    },
    {
      "epoch": 0.05549563682713991,
      "grad_norm": 11.980429649353027,
      "learning_rate": 0.00018127499143005268,
      "loss": 0.6955,
      "step": 283
    },
    {
      "epoch": 0.05569173448377292,
      "grad_norm": 38.28229904174805,
      "learning_rate": 0.00018114205171291663,
      "loss": 1.7335,
      "step": 284
    },
    {
      "epoch": 0.05588783214040592,
      "grad_norm": 6.15138053894043,
      "learning_rate": 0.000181008690864976,
      "loss": 1.2766,
      "step": 285
    },
    {
      "epoch": 0.056083929797038926,
      "grad_norm": 7.846836566925049,
      "learning_rate": 0.00018087490957837944,
      "loss": 1.155,
      "step": 286
    },
    {
      "epoch": 0.05628002745367193,
      "grad_norm": 7.675628185272217,
      "learning_rate": 0.00018074070854745772,
      "loss": 1.6129,
      "step": 287
    },
    {
      "epoch": 0.056476125110304934,
      "grad_norm": 12.245649337768555,
      "learning_rate": 0.00018060608846872005,
      "loss": 1.7585,
      "step": 288
    },
    {
      "epoch": 0.05667222276693794,
      "grad_norm": 10.520101547241211,
      "learning_rate": 0.00018047105004085053,
      "loss": 1.9265,
      "step": 289
    },
    {
      "epoch": 0.056868320423570935,
      "grad_norm": 7.400151252746582,
      "learning_rate": 0.00018033559396470454,
      "loss": 1.4189,
      "step": 290
    },
    {
      "epoch": 0.05706441808020394,
      "grad_norm": 12.058060646057129,
      "learning_rate": 0.00018019972094330503,
      "loss": 2.3312,
      "step": 291
    },
    {
      "epoch": 0.057260515736836944,
      "grad_norm": 5.313794136047363,
      "learning_rate": 0.00018006343168183893,
      "loss": 2.0051,
      "step": 292
    },
    {
      "epoch": 0.05745661339346995,
      "grad_norm": 11.182997703552246,
      "learning_rate": 0.0001799267268876535,
      "loss": 1.4779,
      "step": 293
    },
    {
      "epoch": 0.05765271105010295,
      "grad_norm": 16.24866485595703,
      "learning_rate": 0.0001797896072702526,
      "loss": 2.4689,
      "step": 294
    },
    {
      "epoch": 0.057848808706735956,
      "grad_norm": 7.471411228179932,
      "learning_rate": 0.00017965207354129307,
      "loss": 3.0599,
      "step": 295
    },
    {
      "epoch": 0.05804490636336896,
      "grad_norm": 7.715878486633301,
      "learning_rate": 0.00017951412641458098,
      "loss": 0.8256,
      "step": 296
    },
    {
      "epoch": 0.058241004020001964,
      "grad_norm": 22.084482192993164,
      "learning_rate": 0.000179375766606068,
      "loss": 2.457,
      "step": 297
    },
    {
      "epoch": 0.05843710167663496,
      "grad_norm": 8.041847229003906,
      "learning_rate": 0.00017923699483384753,
      "loss": 1.5642,
      "step": 298
    },
    {
      "epoch": 0.058633199333267966,
      "grad_norm": 12.814888000488281,
      "learning_rate": 0.00017909781181815117,
      "loss": 1.5129,
      "step": 299
    },
    {
      "epoch": 0.05882929698990097,
      "grad_norm": 9.216371536254883,
      "learning_rate": 0.0001789582182813449,
      "loss": 2.0632,
      "step": 300
    },
    {
      "epoch": 0.059025394646533974,
      "grad_norm": 12.80371379852295,
      "learning_rate": 0.00017881821494792528,
      "loss": 2.8705,
      "step": 301
    },
    {
      "epoch": 0.05922149230316698,
      "grad_norm": 7.234943389892578,
      "learning_rate": 0.00017867780254451576,
      "loss": 2.6664,
      "step": 302
    },
    {
      "epoch": 0.05941758995979998,
      "grad_norm": 11.168726921081543,
      "learning_rate": 0.00017853698179986282,
      "loss": 1.347,
      "step": 303
    },
    {
      "epoch": 0.059613687616432987,
      "grad_norm": 19.369266510009766,
      "learning_rate": 0.00017839575344483238,
      "loss": 2.68,
      "step": 304
    },
    {
      "epoch": 0.059809785273065984,
      "grad_norm": 7.1730570793151855,
      "learning_rate": 0.0001782541182124057,
      "loss": 2.3908,
      "step": 305
    },
    {
      "epoch": 0.06000588292969899,
      "grad_norm": 7.243929862976074,
      "learning_rate": 0.0001781120768376759,
      "loss": 1.0056,
      "step": 306
    },
    {
      "epoch": 0.06020198058633199,
      "grad_norm": 7.748988628387451,
      "learning_rate": 0.00017796963005784394,
      "loss": 2.1776,
      "step": 307
    },
    {
      "epoch": 0.060398078242964996,
      "grad_norm": 13.446945190429688,
      "learning_rate": 0.0001778267786122148,
      "loss": 2.3275,
      "step": 308
    },
    {
      "epoch": 0.060594175899598,
      "grad_norm": 10.720627784729004,
      "learning_rate": 0.0001776835232421938,
      "loss": 1.046,
      "step": 309
    },
    {
      "epoch": 0.060790273556231005,
      "grad_norm": 11.274985313415527,
      "learning_rate": 0.00017753986469128257,
      "loss": 2.4269,
      "step": 310
    },
    {
      "epoch": 0.06098637121286401,
      "grad_norm": 8.671335220336914,
      "learning_rate": 0.00017739580370507532,
      "loss": 2.1488,
      "step": 311
    },
    {
      "epoch": 0.06118246886949701,
      "grad_norm": 8.375978469848633,
      "learning_rate": 0.0001772513410312548,
      "loss": 1.8458,
      "step": 312
    },
    {
      "epoch": 0.06137856652613001,
      "grad_norm": 11.178112983703613,
      "learning_rate": 0.00017710647741958868,
      "loss": 2.7169,
      "step": 313
    },
    {
      "epoch": 0.061574664182763014,
      "grad_norm": 8.29799747467041,
      "learning_rate": 0.00017696121362192544,
      "loss": 1.455,
      "step": 314
    },
    {
      "epoch": 0.06177076183939602,
      "grad_norm": 6.712766647338867,
      "learning_rate": 0.00017681555039219054,
      "loss": 1.2604,
      "step": 315
    },
    {
      "epoch": 0.06196685949602902,
      "grad_norm": 7.891608238220215,
      "learning_rate": 0.00017666948848638257,
      "loss": 2.1795,
      "step": 316
    },
    {
      "epoch": 0.06216295715266203,
      "grad_norm": 5.039219379425049,
      "learning_rate": 0.00017652302866256916,
      "loss": 0.9069,
      "step": 317
    },
    {
      "epoch": 0.06235905480929503,
      "grad_norm": 9.421103477478027,
      "learning_rate": 0.00017637617168088325,
      "loss": 2.4256,
      "step": 318
    },
    {
      "epoch": 0.06255515246592804,
      "grad_norm": 4.435902118682861,
      "learning_rate": 0.000176228918303519,
      "loss": 1.9269,
      "step": 319
    },
    {
      "epoch": 0.06275125012256104,
      "grad_norm": 10.938987731933594,
      "learning_rate": 0.00017608126929472795,
      "loss": 1.4649,
      "step": 320
    },
    {
      "epoch": 0.06294734777919404,
      "grad_norm": 6.332970142364502,
      "learning_rate": 0.00017593322542081485,
      "loss": 2.0089,
      "step": 321
    },
    {
      "epoch": 0.06314344543582705,
      "grad_norm": 6.731532573699951,
      "learning_rate": 0.00017578478745013392,
      "loss": 2.4046,
      "step": 322
    },
    {
      "epoch": 0.06333954309246005,
      "grad_norm": 8.772012710571289,
      "learning_rate": 0.00017563595615308474,
      "loss": 1.4935,
      "step": 323
    },
    {
      "epoch": 0.06353564074909304,
      "grad_norm": 5.693745136260986,
      "learning_rate": 0.00017548673230210823,
      "loss": 1.848,
      "step": 324
    },
    {
      "epoch": 0.06373173840572605,
      "grad_norm": 15.056157112121582,
      "learning_rate": 0.0001753371166716828,
      "loss": 1.4598,
      "step": 325
    },
    {
      "epoch": 0.06392783606235905,
      "grad_norm": 9.370506286621094,
      "learning_rate": 0.00017518711003832002,
      "loss": 1.4809,
      "step": 326
    },
    {
      "epoch": 0.06412393371899205,
      "grad_norm": 19.398839950561523,
      "learning_rate": 0.000175036713180561,
      "loss": 1.0093,
      "step": 327
    },
    {
      "epoch": 0.06432003137562506,
      "grad_norm": 4.393742084503174,
      "learning_rate": 0.00017488592687897193,
      "loss": 0.817,
      "step": 328
    },
    {
      "epoch": 0.06451612903225806,
      "grad_norm": 6.7713799476623535,
      "learning_rate": 0.00017473475191614037,
      "loss": 2.1701,
      "step": 329
    },
    {
      "epoch": 0.06471222668889107,
      "grad_norm": 5.920267581939697,
      "learning_rate": 0.00017458318907667098,
      "loss": 3.3491,
      "step": 330
    },
    {
      "epoch": 0.06490832434552407,
      "grad_norm": 15.095996856689453,
      "learning_rate": 0.0001744312391471816,
      "loss": 1.7637,
      "step": 331
    },
    {
      "epoch": 0.06510442200215708,
      "grad_norm": 9.470211029052734,
      "learning_rate": 0.00017427890291629893,
      "loss": 2.7744,
      "step": 332
    },
    {
      "epoch": 0.06530051965879008,
      "grad_norm": 9.082067489624023,
      "learning_rate": 0.00017412618117465477,
      "loss": 3.1791,
      "step": 333
    },
    {
      "epoch": 0.06549661731542308,
      "grad_norm": 5.174635410308838,
      "learning_rate": 0.0001739730747148816,
      "loss": 1.2189,
      "step": 334
    },
    {
      "epoch": 0.06569271497205609,
      "grad_norm": 5.053405284881592,
      "learning_rate": 0.00017381958433160865,
      "loss": 1.7119,
      "step": 335
    },
    {
      "epoch": 0.06588881262868909,
      "grad_norm": 5.771046161651611,
      "learning_rate": 0.0001736657108214578,
      "loss": 1.4188,
      "step": 336
    },
    {
      "epoch": 0.0660849102853221,
      "grad_norm": 8.400517463684082,
      "learning_rate": 0.00017351145498303925,
      "loss": 2.3167,
      "step": 337
    },
    {
      "epoch": 0.0662810079419551,
      "grad_norm": 4.6646728515625,
      "learning_rate": 0.0001733568176169476,
      "loss": 1.2102,
      "step": 338
    },
    {
      "epoch": 0.06647710559858809,
      "grad_norm": 8.288646697998047,
      "learning_rate": 0.0001732017995257575,
      "loss": 2.4803,
      "step": 339
    },
    {
      "epoch": 0.0666732032552211,
      "grad_norm": 10.970074653625488,
      "learning_rate": 0.00017304640151401967,
      "loss": 2.5839,
      "step": 340
    },
    {
      "epoch": 0.0668693009118541,
      "grad_norm": 6.0125732421875,
      "learning_rate": 0.00017289062438825665,
      "loss": 1.5807,
      "step": 341
    },
    {
      "epoch": 0.0670653985684871,
      "grad_norm": 5.844028472900391,
      "learning_rate": 0.0001727344689569585,
      "loss": 3.34,
      "step": 342
    },
    {
      "epoch": 0.06726149622512011,
      "grad_norm": 7.1026387214660645,
      "learning_rate": 0.00017257793603057871,
      "loss": 1.4347,
      "step": 343
    },
    {
      "epoch": 0.06745759388175311,
      "grad_norm": 9.198262214660645,
      "learning_rate": 0.00017242102642153016,
      "loss": 1.834,
      "step": 344
    },
    {
      "epoch": 0.06765369153838612,
      "grad_norm": 5.76854133605957,
      "learning_rate": 0.00017226374094418044,
      "loss": 0.9294,
      "step": 345
    },
    {
      "epoch": 0.06784978919501912,
      "grad_norm": 10.319186210632324,
      "learning_rate": 0.0001721060804148482,
      "loss": 2.0088,
      "step": 346
    },
    {
      "epoch": 0.06804588685165212,
      "grad_norm": 22.298240661621094,
      "learning_rate": 0.00017194804565179842,
      "loss": 2.6901,
      "step": 347
    },
    {
      "epoch": 0.06824198450828513,
      "grad_norm": 11.38401985168457,
      "learning_rate": 0.00017178963747523847,
      "loss": 2.6342,
      "step": 348
    },
    {
      "epoch": 0.06824198450828513,
      "eval_loss": 0.4400941729545593,
      "eval_runtime": 78.7276,
      "eval_samples_per_second": 27.284,
      "eval_steps_per_second": 13.642,
      "step": 348
    }
  ],
  "logging_steps": 1,
  "max_steps": 1389,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 348,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.492671941640192e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}