{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9958417169684776,
  "eval_steps": 500,
  "global_step": 232,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004292421193829644,
      "grad_norm": 3.342827936598552,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 0.679,
      "step": 1
    },
    {
      "epoch": 0.008584842387659289,
      "grad_norm": 2.896501931914046,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 0.6709,
      "step": 2
    },
    {
      "epoch": 0.012877263581488933,
      "grad_norm": 3.3718581451101177,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 0.6619,
      "step": 3
    },
    {
      "epoch": 0.017169684775318578,
      "grad_norm": 2.529609856453942,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.7753,
      "step": 4
    },
    {
      "epoch": 0.021462105969148222,
      "grad_norm": 2.4044812444703103,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 0.7833,
      "step": 5
    },
    {
      "epoch": 0.025754527162977867,
      "grad_norm": 2.152377713157304,
      "learning_rate": 8.571428571428571e-06,
      "loss": 0.7411,
      "step": 6
    },
    {
      "epoch": 0.03004694835680751,
      "grad_norm": 7.163921239673759,
      "learning_rate": 1e-05,
      "loss": 0.6751,
      "step": 7
    },
    {
      "epoch": 0.034339369550637155,
      "grad_norm": 1.9811371567636475,
      "learning_rate": 9.999512620046523e-06,
      "loss": 0.6019,
      "step": 8
    },
    {
      "epoch": 0.0386317907444668,
      "grad_norm": 2.2991427332066663,
      "learning_rate": 9.998050575201772e-06,
      "loss": 0.6195,
      "step": 9
    },
    {
      "epoch": 0.042924211938296444,
      "grad_norm": 2.633277122254043,
      "learning_rate": 9.995614150494293e-06,
      "loss": 0.6773,
      "step": 10
    },
    {
      "epoch": 0.04721663313212609,
      "grad_norm": 0.8137858553380385,
      "learning_rate": 9.992203820909906e-06,
      "loss": 0.44,
      "step": 11
    },
    {
      "epoch": 0.05150905432595573,
      "grad_norm": 2.5789476047477495,
      "learning_rate": 9.987820251299121e-06,
      "loss": 0.6856,
      "step": 12
    },
    {
      "epoch": 0.05580147551978538,
      "grad_norm": 2.325022123665024,
      "learning_rate": 9.982464296247523e-06,
      "loss": 0.6573,
      "step": 13
    },
    {
      "epoch": 0.06009389671361502,
      "grad_norm": 2.1016133484431814,
      "learning_rate": 9.976136999909156e-06,
      "loss": 0.5795,
      "step": 14
    },
    {
      "epoch": 0.06438631790744467,
      "grad_norm": 2.203575792232459,
      "learning_rate": 9.968839595802982e-06,
      "loss": 0.5512,
      "step": 15
    },
    {
      "epoch": 0.06867873910127431,
      "grad_norm": 1.7649770138174061,
      "learning_rate": 9.960573506572391e-06,
      "loss": 0.5991,
      "step": 16
    },
    {
      "epoch": 0.07297116029510396,
      "grad_norm": 2.125017710581645,
      "learning_rate": 9.951340343707852e-06,
      "loss": 0.6961,
      "step": 17
    },
    {
      "epoch": 0.0772635814889336,
      "grad_norm": 1.9648109979705357,
      "learning_rate": 9.941141907232766e-06,
      "loss": 0.6274,
      "step": 18
    },
    {
      "epoch": 0.08155600268276325,
      "grad_norm": 2.1430042253940997,
      "learning_rate": 9.929980185352525e-06,
      "loss": 0.5933,
      "step": 19
    },
    {
      "epoch": 0.08584842387659289,
      "grad_norm": 2.187065629596637,
      "learning_rate": 9.91785735406693e-06,
      "loss": 0.6537,
      "step": 20
    },
    {
      "epoch": 0.09014084507042254,
      "grad_norm": 2.0411377847693366,
      "learning_rate": 9.904775776745959e-06,
      "loss": 0.5277,
      "step": 21
    },
    {
      "epoch": 0.09443326626425218,
      "grad_norm": 1.9222476925913479,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.6009,
      "step": 22
    },
    {
      "epoch": 0.09872568745808183,
      "grad_norm": 2.202338617043183,
      "learning_rate": 9.875746771527817e-06,
      "loss": 0.6808,
      "step": 23
    },
    {
      "epoch": 0.10301810865191147,
      "grad_norm": 1.8704298128204246,
      "learning_rate": 9.859805002892733e-06,
      "loss": 0.5834,
      "step": 24
    },
    {
      "epoch": 0.10731052984574112,
      "grad_norm": 2.26606517803316,
      "learning_rate": 9.842915805643156e-06,
      "loss": 0.6416,
      "step": 25
    },
    {
      "epoch": 0.11160295103957076,
      "grad_norm": 1.957724168370316,
      "learning_rate": 9.825082472361558e-06,
      "loss": 0.5565,
      "step": 26
    },
    {
      "epoch": 0.1158953722334004,
      "grad_norm": 2.7304041309488367,
      "learning_rate": 9.806308479691595e-06,
      "loss": 0.6941,
      "step": 27
    },
    {
      "epoch": 0.12018779342723004,
      "grad_norm": 2.252924472572446,
      "learning_rate": 9.786597487660336e-06,
      "loss": 0.6806,
      "step": 28
    },
    {
      "epoch": 0.1244802146210597,
      "grad_norm": 2.208805819097276,
      "learning_rate": 9.765953338964736e-06,
      "loss": 0.6123,
      "step": 29
    },
    {
      "epoch": 0.12877263581488935,
      "grad_norm": 2.0330562196938238,
      "learning_rate": 9.744380058222483e-06,
      "loss": 0.5966,
      "step": 30
    },
    {
      "epoch": 0.13306505700871898,
      "grad_norm": 2.4286898255231533,
      "learning_rate": 9.721881851187406e-06,
      "loss": 0.6148,
      "step": 31
    },
    {
      "epoch": 0.13735747820254862,
      "grad_norm": 1.911339918779555,
      "learning_rate": 9.698463103929542e-06,
      "loss": 0.5662,
      "step": 32
    },
    {
      "epoch": 0.14164989939637826,
      "grad_norm": 2.0543803346952583,
      "learning_rate": 9.674128381980073e-06,
      "loss": 0.6602,
      "step": 33
    },
    {
      "epoch": 0.14594232059020792,
      "grad_norm": 2.086982956085146,
      "learning_rate": 9.648882429441258e-06,
      "loss": 0.5432,
      "step": 34
    },
    {
      "epoch": 0.15023474178403756,
      "grad_norm": 2.0901966575723474,
      "learning_rate": 9.622730168061568e-06,
      "loss": 0.5274,
      "step": 35
    },
    {
      "epoch": 0.1545271629778672,
      "grad_norm": 4.381472469442544,
      "learning_rate": 9.595676696276173e-06,
      "loss": 0.583,
      "step": 36
    },
    {
      "epoch": 0.15881958417169684,
      "grad_norm": 2.3265871885088503,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.5623,
      "step": 37
    },
    {
      "epoch": 0.1631120053655265,
      "grad_norm": 2.0021378156582963,
      "learning_rate": 9.538887392664544e-06,
      "loss": 0.6318,
      "step": 38
    },
    {
      "epoch": 0.16740442655935614,
      "grad_norm": 1.0984958441170714,
      "learning_rate": 9.50916263202557e-06,
      "loss": 0.4548,
      "step": 39
    },
    {
      "epoch": 0.17169684775318578,
      "grad_norm": 1.8318936065536409,
      "learning_rate": 9.478558801197065e-06,
      "loss": 0.5985,
      "step": 40
    },
    {
      "epoch": 0.17598926894701541,
      "grad_norm": 2.7994275799684076,
      "learning_rate": 9.44708186645649e-06,
      "loss": 0.5926,
      "step": 41
    },
    {
      "epoch": 0.18028169014084508,
      "grad_norm": 2.072220455177539,
      "learning_rate": 9.414737964294636e-06,
      "loss": 0.5318,
      "step": 42
    },
    {
      "epoch": 0.18457411133467472,
      "grad_norm": 2.2812531791006645,
      "learning_rate": 9.381533400219319e-06,
      "loss": 0.5948,
      "step": 43
    },
    {
      "epoch": 0.18886653252850436,
      "grad_norm": 1.8766396474219533,
      "learning_rate": 9.347474647526095e-06,
      "loss": 0.5514,
      "step": 44
    },
    {
      "epoch": 0.193158953722334,
      "grad_norm": 1.9908306217286644,
      "learning_rate": 9.312568346036288e-06,
      "loss": 0.5282,
      "step": 45
    },
    {
      "epoch": 0.19745137491616366,
      "grad_norm": 2.2942225048556284,
      "learning_rate": 9.276821300802535e-06,
      "loss": 0.5875,
      "step": 46
    },
    {
      "epoch": 0.2017437961099933,
      "grad_norm": 3.1824151916415087,
      "learning_rate": 9.24024048078213e-06,
      "loss": 0.6144,
      "step": 47
    },
    {
      "epoch": 0.20603621730382293,
      "grad_norm": 0.992721768550404,
      "learning_rate": 9.202833017478421e-06,
      "loss": 0.4847,
      "step": 48
    },
    {
      "epoch": 0.21032863849765257,
      "grad_norm": 1.9310743502843328,
      "learning_rate": 9.164606203550498e-06,
      "loss": 0.5974,
      "step": 49
    },
    {
      "epoch": 0.21462105969148224,
      "grad_norm": 2.1730863843947392,
      "learning_rate": 9.125567491391476e-06,
      "loss": 0.6293,
      "step": 50
    },
    {
      "epoch": 0.21891348088531187,
      "grad_norm": 0.8680338310564839,
      "learning_rate": 9.085724491675642e-06,
      "loss": 0.4757,
      "step": 51
    },
    {
      "epoch": 0.2232059020791415,
      "grad_norm": 2.1915863963781916,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.5925,
      "step": 52
    },
    {
      "epoch": 0.22749832327297115,
      "grad_norm": 2.3492931220336195,
      "learning_rate": 9.003656854743667e-06,
      "loss": 0.6402,
      "step": 53
    },
    {
      "epoch": 0.2317907444668008,
      "grad_norm": 0.8164522682092396,
      "learning_rate": 8.961448216775955e-06,
      "loss": 0.4382,
      "step": 54
    },
    {
      "epoch": 0.23608316566063045,
      "grad_norm": 2.2933091603858418,
      "learning_rate": 8.9184672866292e-06,
      "loss": 0.5367,
      "step": 55
    },
    {
      "epoch": 0.2403755868544601,
      "grad_norm": 2.051540581160434,
      "learning_rate": 8.874722443520898e-06,
      "loss": 0.5736,
      "step": 56
    },
    {
      "epoch": 0.24466800804828973,
      "grad_norm": 1.9508118237454382,
      "learning_rate": 8.83022221559489e-06,
      "loss": 0.6196,
      "step": 57
    },
    {
      "epoch": 0.2489604292421194,
      "grad_norm": 2.4166528722303036,
      "learning_rate": 8.784975278258783e-06,
      "loss": 0.5279,
      "step": 58
    },
    {
      "epoch": 0.25325285043594903,
      "grad_norm": 1.9719959484618856,
      "learning_rate": 8.73899045249266e-06,
      "loss": 0.5248,
      "step": 59
    },
    {
      "epoch": 0.2575452716297787,
      "grad_norm": 2.088094168914141,
      "learning_rate": 8.692276703129421e-06,
      "loss": 0.5436,
      "step": 60
    },
    {
      "epoch": 0.2618376928236083,
      "grad_norm": 2.347294342285649,
      "learning_rate": 8.644843137107058e-06,
      "loss": 0.7,
      "step": 61
    },
    {
      "epoch": 0.26613011401743797,
      "grad_norm": 2.152662276659629,
      "learning_rate": 8.596699001693257e-06,
      "loss": 0.5052,
      "step": 62
    },
    {
      "epoch": 0.2704225352112676,
      "grad_norm": 2.1260026022790552,
      "learning_rate": 8.547853682682605e-06,
      "loss": 0.6121,
      "step": 63
    },
    {
      "epoch": 0.27471495640509724,
      "grad_norm": 1.9148541801577235,
      "learning_rate": 8.498316702566828e-06,
      "loss": 0.5516,
      "step": 64
    },
    {
      "epoch": 0.2790073775989269,
      "grad_norm": 2.134479085245887,
      "learning_rate": 8.44809771867835e-06,
      "loss": 0.6297,
      "step": 65
    },
    {
      "epoch": 0.2832997987927565,
      "grad_norm": 1.776446730314099,
      "learning_rate": 8.397206521307584e-06,
      "loss": 0.5337,
      "step": 66
    },
    {
      "epoch": 0.2875922199865862,
      "grad_norm": 1.955780895559264,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.6187,
      "step": 67
    },
    {
      "epoch": 0.29188464118041585,
      "grad_norm": 2.065677438153802,
      "learning_rate": 8.293447300593402e-06,
      "loss": 0.4712,
      "step": 68
    },
    {
      "epoch": 0.29617706237424546,
      "grad_norm": 1.929637625719691,
      "learning_rate": 8.240599505315656e-06,
      "loss": 0.5638,
      "step": 69
    },
    {
      "epoch": 0.3004694835680751,
      "grad_norm": 2.1234613246255294,
      "learning_rate": 8.18711994874345e-06,
      "loss": 0.5622,
      "step": 70
    },
    {
      "epoch": 0.3047619047619048,
      "grad_norm": 2.4974658951008935,
      "learning_rate": 8.133019056822303e-06,
      "loss": 0.5656,
      "step": 71
    },
    {
      "epoch": 0.3090543259557344,
      "grad_norm": 2.1148121496211028,
      "learning_rate": 8.078307376628292e-06,
      "loss": 0.5706,
      "step": 72
    },
    {
      "epoch": 0.31334674714956406,
      "grad_norm": 1.9637387645317304,
      "learning_rate": 8.022995574311876e-06,
      "loss": 0.6431,
      "step": 73
    },
    {
      "epoch": 0.3176391683433937,
      "grad_norm": 2.2321260500467996,
      "learning_rate": 7.967094433018508e-06,
      "loss": 0.6038,
      "step": 74
    },
    {
      "epoch": 0.32193158953722334,
      "grad_norm": 2.1333068986028136,
      "learning_rate": 7.910614850786448e-06,
      "loss": 0.5666,
      "step": 75
    },
    {
      "epoch": 0.326224010731053,
      "grad_norm": 2.195983244813881,
      "learning_rate": 7.85356783842216e-06,
      "loss": 0.5877,
      "step": 76
    },
    {
      "epoch": 0.3305164319248826,
      "grad_norm": 2.9185529975089644,
      "learning_rate": 7.795964517353734e-06,
      "loss": 0.5221,
      "step": 77
    },
    {
      "epoch": 0.3348088531187123,
      "grad_norm": 2.1356064453519363,
      "learning_rate": 7.737816117462752e-06,
      "loss": 0.5536,
      "step": 78
    },
    {
      "epoch": 0.33910127431254194,
      "grad_norm": 0.9888358843543892,
      "learning_rate": 7.679133974894984e-06,
      "loss": 0.427,
      "step": 79
    },
    {
      "epoch": 0.34339369550637155,
      "grad_norm": 0.9532274184167417,
      "learning_rate": 7.619929529850397e-06,
      "loss": 0.4607,
      "step": 80
    },
    {
      "epoch": 0.3476861167002012,
      "grad_norm": 1.969695168470127,
      "learning_rate": 7.560214324352858e-06,
      "loss": 0.525,
      "step": 81
    },
    {
      "epoch": 0.35197853789403083,
      "grad_norm": 2.2331654792734272,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.6041,
      "step": 82
    },
    {
      "epoch": 0.3562709590878605,
      "grad_norm": 2.4884565474083606,
      "learning_rate": 7.4392982956936644e-06,
      "loss": 0.5886,
      "step": 83
    },
    {
      "epoch": 0.36056338028169016,
      "grad_norm": 1.8314120774170615,
      "learning_rate": 7.378121045351378e-06,
      "loss": 0.5349,
      "step": 84
    },
    {
      "epoch": 0.36485580147551977,
      "grad_norm": 2.1762807073421127,
      "learning_rate": 7.31648017559931e-06,
      "loss": 0.5606,
      "step": 85
    },
    {
      "epoch": 0.36914822266934944,
      "grad_norm": 2.422109142275802,
      "learning_rate": 7.254387703447154e-06,
      "loss": 0.5431,
      "step": 86
    },
    {
      "epoch": 0.3734406438631791,
      "grad_norm": 1.9079552738678454,
      "learning_rate": 7.191855733945388e-06,
      "loss": 0.5553,
      "step": 87
    },
    {
      "epoch": 0.3777330650570087,
      "grad_norm": 1.760125442274873,
      "learning_rate": 7.128896457825364e-06,
      "loss": 0.5808,
      "step": 88
    },
    {
      "epoch": 0.3820254862508384,
      "grad_norm": 0.9979331285164651,
      "learning_rate": 7.06552214912271e-06,
      "loss": 0.4579,
      "step": 89
    },
    {
      "epoch": 0.386317907444668,
      "grad_norm": 1.8496753131089991,
      "learning_rate": 7.0017451627844765e-06,
      "loss": 0.591,
      "step": 90
    },
    {
      "epoch": 0.39061032863849765,
      "grad_norm": 2.3343608471053265,
      "learning_rate": 6.9375779322605154e-06,
      "loss": 0.6091,
      "step": 91
    },
    {
      "epoch": 0.3949027498323273,
      "grad_norm": 1.9668213618430554,
      "learning_rate": 6.873032967079562e-06,
      "loss": 0.6944,
      "step": 92
    },
    {
      "epoch": 0.3991951710261569,
      "grad_norm": 1.8725242772507493,
      "learning_rate": 6.808122850410461e-06,
      "loss": 0.6055,
      "step": 93
    },
    {
      "epoch": 0.4034875922199866,
      "grad_norm": 1.9966003945224369,
      "learning_rate": 6.7428602366090764e-06,
      "loss": 0.5595,
      "step": 94
    },
    {
      "epoch": 0.40778001341381626,
      "grad_norm": 3.053074851635477,
      "learning_rate": 6.677257848751276e-06,
      "loss": 0.5857,
      "step": 95
    },
    {
      "epoch": 0.41207243460764587,
      "grad_norm": 2.010221226340498,
      "learning_rate": 6.611328476152557e-06,
      "loss": 0.5995,
      "step": 96
    },
    {
      "epoch": 0.41636485580147553,
      "grad_norm": 1.742641730594434,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.5389,
      "step": 97
    },
    {
      "epoch": 0.42065727699530514,
      "grad_norm": 0.7858607450939203,
      "learning_rate": 6.4785402502202345e-06,
      "loss": 0.4598,
      "step": 98
    },
    {
      "epoch": 0.4249496981891348,
      "grad_norm": 1.9026874984032676,
      "learning_rate": 6.411707284214384e-06,
      "loss": 0.5824,
      "step": 99
    },
    {
      "epoch": 0.42924211938296447,
      "grad_norm": 1.7972924537621116,
      "learning_rate": 6.344599103076329e-06,
      "loss": 0.5605,
      "step": 100
    },
    {
      "epoch": 0.4335345405767941,
      "grad_norm": 1.6566505050926905,
      "learning_rate": 6.277228789678953e-06,
      "loss": 0.55,
      "step": 101
    },
    {
      "epoch": 0.43782696177062375,
      "grad_norm": 0.7871200662869098,
      "learning_rate": 6.209609477998339e-06,
      "loss": 0.4487,
      "step": 102
    },
    {
      "epoch": 0.4421193829644534,
      "grad_norm": 2.118924742242862,
      "learning_rate": 6.141754350553279e-06,
      "loss": 0.5791,
      "step": 103
    },
    {
      "epoch": 0.446411804158283,
      "grad_norm": 1.9417216221368725,
      "learning_rate": 6.073676635835317e-06,
      "loss": 0.5321,
      "step": 104
    },
    {
      "epoch": 0.4507042253521127,
      "grad_norm": 0.8301362013675532,
      "learning_rate": 6.005389605729824e-06,
      "loss": 0.471,
      "step": 105
    },
    {
      "epoch": 0.4549966465459423,
      "grad_norm": 1.9216842351031653,
      "learning_rate": 5.936906572928625e-06,
      "loss": 0.4981,
      "step": 106
    },
    {
      "epoch": 0.45928906773977196,
      "grad_norm": 2.1309977101502913,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.523,
      "step": 107
    },
    {
      "epoch": 0.4635814889336016,
      "grad_norm": 6.307531137311205,
      "learning_rate": 5.799405938459175e-06,
      "loss": 0.5631,
      "step": 108
    },
    {
      "epoch": 0.46787391012743124,
      "grad_norm": 0.8161016261331794,
      "learning_rate": 5.730415142812059e-06,
      "loss": 0.4739,
      "step": 109
    },
    {
      "epoch": 0.4721663313212609,
      "grad_norm": 1.9009887448608773,
      "learning_rate": 5.661281951285613e-06,
      "loss": 0.6272,
      "step": 110
    },
    {
      "epoch": 0.47645875251509057,
      "grad_norm": 2.2577462172723233,
      "learning_rate": 5.592019841532507e-06,
      "loss": 0.5796,
      "step": 111
    },
    {
      "epoch": 0.4807511737089202,
      "grad_norm": 2.1012817228049814,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.6074,
      "step": 112
    },
    {
      "epoch": 0.48504359490274984,
      "grad_norm": 3.169512060210574,
      "learning_rate": 5.453162900988902e-06,
      "loss": 0.5817,
      "step": 113
    },
    {
      "epoch": 0.48933601609657945,
      "grad_norm": 0.7965146219233805,
      "learning_rate": 5.383595140634093e-06,
      "loss": 0.4628,
      "step": 114
    },
    {
      "epoch": 0.4936284372904091,
      "grad_norm": 2.528449046345642,
      "learning_rate": 5.3139525976465675e-06,
      "loss": 0.6113,
      "step": 115
    },
    {
      "epoch": 0.4979208584842388,
      "grad_norm": 2.169410644155486,
      "learning_rate": 5.244248848978067e-06,
      "loss": 0.6655,
      "step": 116
    },
    {
      "epoch": 0.5022132796780684,
      "grad_norm": 2.114218695353453,
      "learning_rate": 5.174497483512506e-06,
      "loss": 0.5438,
      "step": 117
    },
    {
      "epoch": 0.5065057008718981,
      "grad_norm": 2.322736697640736,
      "learning_rate": 5.1047120994167855e-06,
      "loss": 0.619,
      "step": 118
    },
    {
      "epoch": 0.5107981220657277,
      "grad_norm": 1.9355642603619068,
      "learning_rate": 5.034906301489808e-06,
      "loss": 0.568,
      "step": 119
    },
    {
      "epoch": 0.5150905432595574,
      "grad_norm": 0.8275929454164218,
      "learning_rate": 4.965093698510192e-06,
      "loss": 0.4642,
      "step": 120
    },
    {
      "epoch": 0.5193829644533869,
      "grad_norm": 0.7826606670717038,
      "learning_rate": 4.895287900583216e-06,
      "loss": 0.428,
      "step": 121
    },
    {
      "epoch": 0.5236753856472166,
      "grad_norm": 2.227565392356624,
      "learning_rate": 4.825502516487497e-06,
      "loss": 0.616,
      "step": 122
    },
    {
      "epoch": 0.5279678068410463,
      "grad_norm": 1.9049065560319947,
      "learning_rate": 4.755751151021934e-06,
      "loss": 0.6396,
      "step": 123
    },
    {
      "epoch": 0.5322602280348759,
      "grad_norm": 2.360278324447153,
      "learning_rate": 4.686047402353433e-06,
      "loss": 0.581,
      "step": 124
    },
    {
      "epoch": 0.5365526492287056,
      "grad_norm": 2.1436744795832463,
      "learning_rate": 4.6164048593659076e-06,
      "loss": 0.6027,
      "step": 125
    },
    {
      "epoch": 0.5408450704225352,
      "grad_norm": 2.0687980774005856,
      "learning_rate": 4.546837099011101e-06,
      "loss": 0.5455,
      "step": 126
    },
    {
      "epoch": 0.5451374916163648,
      "grad_norm": 2.094975644434325,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.5659,
      "step": 127
    },
    {
      "epoch": 0.5494299128101945,
      "grad_norm": 2.183849966879906,
      "learning_rate": 4.4079801584674955e-06,
      "loss": 0.578,
      "step": 128
    },
    {
      "epoch": 0.5537223340040242,
      "grad_norm": 1.8848396606946802,
      "learning_rate": 4.3387180487143875e-06,
      "loss": 0.5052,
      "step": 129
    },
    {
      "epoch": 0.5580147551978538,
      "grad_norm": 0.843007824971089,
      "learning_rate": 4.269584857187942e-06,
      "loss": 0.4997,
      "step": 130
    },
    {
      "epoch": 0.5623071763916835,
      "grad_norm": 2.763950937739874,
      "learning_rate": 4.200594061540827e-06,
      "loss": 0.5819,
      "step": 131
    },
    {
      "epoch": 0.566599597585513,
      "grad_norm": 1.7630340741332657,
      "learning_rate": 4.131759111665349e-06,
      "loss": 0.604,
      "step": 132
    },
    {
      "epoch": 0.5708920187793427,
      "grad_norm": 3.2094078977714897,
      "learning_rate": 4.063093427071376e-06,
      "loss": 0.6265,
      "step": 133
    },
    {
      "epoch": 0.5751844399731724,
      "grad_norm": 1.8312046959677455,
      "learning_rate": 3.994610394270178e-06,
      "loss": 0.5885,
      "step": 134
    },
    {
      "epoch": 0.579476861167002,
      "grad_norm": 1.8491871843778356,
      "learning_rate": 3.926323364164684e-06,
      "loss": 0.634,
      "step": 135
    },
    {
      "epoch": 0.5837692823608317,
      "grad_norm": 2.6473671182169167,
      "learning_rate": 3.8582456494467214e-06,
      "loss": 0.6355,
      "step": 136
    },
    {
      "epoch": 0.5880617035546613,
      "grad_norm": 2.999849822112049,
      "learning_rate": 3.790390522001662e-06,
      "loss": 0.529,
      "step": 137
    },
    {
      "epoch": 0.5923541247484909,
      "grad_norm": 1.9445772581815945,
      "learning_rate": 3.7227712103210485e-06,
      "loss": 0.5575,
      "step": 138
    },
    {
      "epoch": 0.5966465459423206,
      "grad_norm": 1.8104560751827103,
      "learning_rate": 3.655400896923672e-06,
      "loss": 0.5254,
      "step": 139
    },
    {
      "epoch": 0.6009389671361502,
      "grad_norm": 2.3671152557639346,
      "learning_rate": 3.5882927157856175e-06,
      "loss": 0.5583,
      "step": 140
    },
    {
      "epoch": 0.6052313883299799,
      "grad_norm": 2.074346620633345,
      "learning_rate": 3.521459749779769e-06,
      "loss": 0.6084,
      "step": 141
    },
    {
      "epoch": 0.6095238095238096,
      "grad_norm": 2.035892537869217,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.5832,
      "step": 142
    },
    {
      "epoch": 0.6138162307176391,
      "grad_norm": 2.1099675037548966,
      "learning_rate": 3.3886715238474454e-06,
      "loss": 0.5579,
      "step": 143
    },
    {
      "epoch": 0.6181086519114688,
      "grad_norm": 2.0468759829486443,
      "learning_rate": 3.322742151248726e-06,
      "loss": 0.5848,
      "step": 144
    },
    {
      "epoch": 0.6224010731052985,
      "grad_norm": 1.9674918076912449,
      "learning_rate": 3.2571397633909252e-06,
      "loss": 0.5398,
      "step": 145
    },
    {
      "epoch": 0.6266934942991281,
      "grad_norm": 1.7459454556549392,
      "learning_rate": 3.1918771495895395e-06,
      "loss": 0.6756,
      "step": 146
    },
    {
      "epoch": 0.6309859154929578,
      "grad_norm": 0.7735085423697842,
      "learning_rate": 3.12696703292044e-06,
      "loss": 0.4194,
      "step": 147
    },
    {
      "epoch": 0.6352783366867873,
      "grad_norm": 1.820298479603609,
      "learning_rate": 3.0624220677394854e-06,
      "loss": 0.5858,
      "step": 148
    },
    {
      "epoch": 0.639570757880617,
      "grad_norm": 3.1630846211682,
      "learning_rate": 2.9982548372155264e-06,
      "loss": 0.5303,
      "step": 149
    },
    {
      "epoch": 0.6438631790744467,
      "grad_norm": 2.413654673008136,
      "learning_rate": 2.934477850877292e-06,
      "loss": 0.5315,
      "step": 150
    },
    {
      "epoch": 0.6481556002682763,
      "grad_norm": 2.508188149902217,
      "learning_rate": 2.871103542174637e-06,
      "loss": 0.5468,
      "step": 151
    },
    {
      "epoch": 0.652448021462106,
      "grad_norm": 1.801696332460669,
      "learning_rate": 2.8081442660546126e-06,
      "loss": 0.5817,
      "step": 152
    },
    {
      "epoch": 0.6567404426559356,
      "grad_norm": 2.26133822829944,
      "learning_rate": 2.7456122965528475e-06,
      "loss": 0.522,
      "step": 153
    },
    {
      "epoch": 0.6610328638497652,
      "grad_norm": 2.4471397871687834,
      "learning_rate": 2.683519824400693e-06,
      "loss": 0.5892,
      "step": 154
    },
    {
      "epoch": 0.6653252850435949,
      "grad_norm": 1.9131453194028882,
      "learning_rate": 2.6218789546486235e-06,
      "loss": 0.5554,
      "step": 155
    },
    {
      "epoch": 0.6696177062374246,
      "grad_norm": 2.4034600764131606,
      "learning_rate": 2.560701704306336e-06,
      "loss": 0.5969,
      "step": 156
    },
    {
      "epoch": 0.6739101274312542,
      "grad_norm": 2.1929545250468423,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.5703,
      "step": 157
    },
    {
      "epoch": 0.6782025486250839,
      "grad_norm": 2.2198350864402348,
      "learning_rate": 2.4397856756471435e-06,
      "loss": 0.5812,
      "step": 158
    },
    {
      "epoch": 0.6824949698189134,
      "grad_norm": 2.0128371852734332,
      "learning_rate": 2.380070470149605e-06,
      "loss": 0.6307,
      "step": 159
    },
    {
      "epoch": 0.6867873910127431,
      "grad_norm": 2.217606056909539,
      "learning_rate": 2.320866025105016e-06,
      "loss": 0.5601,
      "step": 160
    },
    {
      "epoch": 0.6910798122065728,
      "grad_norm": 2.0541472609321065,
      "learning_rate": 2.2621838825372496e-06,
      "loss": 0.6326,
      "step": 161
    },
    {
      "epoch": 0.6953722334004024,
      "grad_norm": 1.7762192766929534,
      "learning_rate": 2.204035482646267e-06,
      "loss": 0.501,
      "step": 162
    },
    {
      "epoch": 0.6996646545942321,
      "grad_norm": 2.4025706210449322,
      "learning_rate": 2.146432161577842e-06,
      "loss": 0.5587,
      "step": 163
    },
    {
      "epoch": 0.7039570757880617,
      "grad_norm": 3.686601475649348,
      "learning_rate": 2.0893851492135536e-06,
      "loss": 0.5732,
      "step": 164
    },
    {
      "epoch": 0.7082494969818913,
      "grad_norm": 1.8617954191026609,
      "learning_rate": 2.0329055669814936e-06,
      "loss": 0.5018,
      "step": 165
    },
    {
      "epoch": 0.712541918175721,
      "grad_norm": 3.1015155924797804,
      "learning_rate": 1.977004425688126e-06,
      "loss": 0.6083,
      "step": 166
    },
    {
      "epoch": 0.7168343393695507,
      "grad_norm": 11.133598367985865,
      "learning_rate": 1.9216926233717087e-06,
      "loss": 0.5543,
      "step": 167
    },
    {
      "epoch": 0.7211267605633803,
      "grad_norm": 1.8192465630367278,
      "learning_rate": 1.8669809431776991e-06,
      "loss": 0.5863,
      "step": 168
    },
    {
      "epoch": 0.7254191817572099,
      "grad_norm": 2.0788969216539757,
      "learning_rate": 1.8128800512565514e-06,
      "loss": 0.5291,
      "step": 169
    },
    {
      "epoch": 0.7297116029510395,
      "grad_norm": 1.8368930044145042,
      "learning_rate": 1.7594004946843458e-06,
      "loss": 0.5378,
      "step": 170
    },
    {
      "epoch": 0.7340040241448692,
      "grad_norm": 2.0502782524895315,
      "learning_rate": 1.7065526994065973e-06,
      "loss": 0.5478,
      "step": 171
    },
    {
      "epoch": 0.7382964453386989,
      "grad_norm": 1.7763584764334766,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.5448,
      "step": 172
    },
    {
      "epoch": 0.7425888665325285,
      "grad_norm": 1.7918310960966537,
      "learning_rate": 1.6027934786924187e-06,
      "loss": 0.6076,
      "step": 173
    },
    {
      "epoch": 0.7468812877263582,
      "grad_norm": 2.3563140946836523,
      "learning_rate": 1.551902281321651e-06,
      "loss": 0.565,
      "step": 174
    },
    {
      "epoch": 0.7511737089201878,
      "grad_norm": 1.8987524052104983,
      "learning_rate": 1.5016832974331725e-06,
      "loss": 0.5367,
      "step": 175
    },
    {
      "epoch": 0.7554661301140174,
      "grad_norm": 1.9189932438009185,
      "learning_rate": 1.4521463173173966e-06,
      "loss": 0.5435,
      "step": 176
    },
    {
      "epoch": 0.7597585513078471,
      "grad_norm": 2.42896938145852,
      "learning_rate": 1.4033009983067454e-06,
      "loss": 0.5141,
      "step": 177
    },
    {
      "epoch": 0.7640509725016768,
      "grad_norm": 2.5249548327367766,
      "learning_rate": 1.3551568628929434e-06,
      "loss": 0.674,
      "step": 178
    },
    {
      "epoch": 0.7683433936955064,
      "grad_norm": 2.806361559097513,
      "learning_rate": 1.3077232968705805e-06,
      "loss": 0.5517,
      "step": 179
    },
    {
      "epoch": 0.772635814889336,
      "grad_norm": 2.2153752757917555,
      "learning_rate": 1.2610095475073415e-06,
      "loss": 0.6408,
      "step": 180
    },
    {
      "epoch": 0.7769282360831656,
      "grad_norm": 2.5628201933255124,
      "learning_rate": 1.2150247217412186e-06,
      "loss": 0.5957,
      "step": 181
    },
    {
      "epoch": 0.7812206572769953,
      "grad_norm": 2.079321530887709,
      "learning_rate": 1.1697777844051105e-06,
      "loss": 0.6155,
      "step": 182
    },
    {
      "epoch": 0.785513078470825,
      "grad_norm": 3.2999932447363416,
      "learning_rate": 1.1252775564791023e-06,
      "loss": 0.5515,
      "step": 183
    },
    {
      "epoch": 0.7898054996646546,
      "grad_norm": 0.7905984900012323,
      "learning_rate": 1.0815327133708015e-06,
      "loss": 0.4313,
      "step": 184
    },
    {
      "epoch": 0.7940979208584842,
      "grad_norm": 1.9368152867312256,
      "learning_rate": 1.0385517832240472e-06,
      "loss": 0.6071,
      "step": 185
    },
    {
      "epoch": 0.7983903420523139,
      "grad_norm": 2.366580720258878,
      "learning_rate": 9.963431452563331e-07,
      "loss": 0.5578,
      "step": 186
    },
    {
      "epoch": 0.8026827632461435,
      "grad_norm": 2.3822261396548576,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.499,
      "step": 187
    },
    {
      "epoch": 0.8069751844399732,
      "grad_norm": 1.7417985644400509,
      "learning_rate": 9.142755083243577e-07,
      "loss": 0.5696,
      "step": 188
    },
    {
      "epoch": 0.8112676056338028,
      "grad_norm": 3.9103317258184376,
      "learning_rate": 8.744325086085248e-07,
      "loss": 0.5079,
      "step": 189
    },
    {
      "epoch": 0.8155600268276325,
      "grad_norm": 3.0678763221012586,
      "learning_rate": 8.353937964495029e-07,
      "loss": 0.5418,
      "step": 190
    },
    {
      "epoch": 0.8198524480214621,
      "grad_norm": 2.042653243276808,
      "learning_rate": 7.971669825215789e-07,
      "loss": 0.5748,
      "step": 191
    },
    {
      "epoch": 0.8241448692152917,
      "grad_norm": 2.39821396059648,
      "learning_rate": 7.597595192178702e-07,
      "loss": 0.5428,
      "step": 192
    },
    {
      "epoch": 0.8284372904091214,
      "grad_norm": 2.762719441820346,
      "learning_rate": 7.23178699197467e-07,
      "loss": 0.5673,
      "step": 193
    },
    {
      "epoch": 0.8327297116029511,
      "grad_norm": 1.9044906325399893,
      "learning_rate": 6.874316539637127e-07,
      "loss": 0.5932,
      "step": 194
    },
    {
      "epoch": 0.8370221327967807,
      "grad_norm": 1.856435008918489,
      "learning_rate": 6.52525352473905e-07,
      "loss": 0.5787,
      "step": 195
    },
    {
      "epoch": 0.8413145539906103,
      "grad_norm": 2.3652417447261995,
      "learning_rate": 6.184665997806832e-07,
      "loss": 0.5167,
      "step": 196
    },
    {
      "epoch": 0.84560697518444,
      "grad_norm": 2.723770526598716,
      "learning_rate": 5.852620357053651e-07,
      "loss": 0.571,
      "step": 197
    },
    {
      "epoch": 0.8498993963782696,
      "grad_norm": 2.4817466195220437,
      "learning_rate": 5.529181335435124e-07,
      "loss": 0.58,
      "step": 198
    },
    {
      "epoch": 0.8541918175720993,
      "grad_norm": 2.0455723374907397,
      "learning_rate": 5.214411988029355e-07,
      "loss": 0.5313,
      "step": 199
    },
    {
      "epoch": 0.8584842387659289,
      "grad_norm": 1.9318379229380933,
      "learning_rate": 4.908373679744316e-07,
      "loss": 0.5439,
      "step": 200
    },
    {
      "epoch": 0.8627766599597585,
      "grad_norm": 0.7959253630064103,
      "learning_rate": 4.6111260733545714e-07,
      "loss": 0.4454,
      "step": 201
    },
    {
      "epoch": 0.8670690811535882,
      "grad_norm": 1.9913907969524394,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.5008,
      "step": 202
    },
    {
      "epoch": 0.8713615023474178,
      "grad_norm": 2.1458695326057082,
      "learning_rate": 4.043233037238281e-07,
      "loss": 0.5459,
      "step": 203
    },
    {
      "epoch": 0.8756539235412475,
      "grad_norm": 1.9619666249914482,
      "learning_rate": 3.772698319384349e-07,
      "loss": 0.4999,
      "step": 204
    },
    {
      "epoch": 0.8799463447350772,
      "grad_norm": 4.3401484961367744,
      "learning_rate": 3.511175705587433e-07,
      "loss": 0.5758,
      "step": 205
    },
    {
      "epoch": 0.8842387659289068,
      "grad_norm": 2.5430076829964174,
      "learning_rate": 3.258716180199278e-07,
      "loss": 0.5761,
      "step": 206
    },
    {
      "epoch": 0.8885311871227364,
      "grad_norm": 0.767231711493243,
      "learning_rate": 3.015368960704584e-07,
      "loss": 0.4614,
      "step": 207
    },
    {
      "epoch": 0.892823608316566,
      "grad_norm": 2.193097249503329,
      "learning_rate": 2.7811814881259503e-07,
      "loss": 0.6247,
      "step": 208
    },
    {
      "epoch": 0.8971160295103957,
      "grad_norm": 1.9331493033717462,
      "learning_rate": 2.556199417775174e-07,
      "loss": 0.56,
      "step": 209
    },
    {
      "epoch": 0.9014084507042254,
      "grad_norm": 2.057966480432949,
      "learning_rate": 2.3404666103526542e-07,
      "loss": 0.5618,
      "step": 210
    },
    {
      "epoch": 0.905700871898055,
      "grad_norm": 2.351596162489054,
      "learning_rate": 2.134025123396638e-07,
      "loss": 0.4834,
      "step": 211
    },
    {
      "epoch": 0.9099932930918846,
      "grad_norm": 1.9120169610557864,
      "learning_rate": 1.9369152030840553e-07,
      "loss": 0.5541,
      "step": 212
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 2.2049446195768763,
      "learning_rate": 1.7491752763844294e-07,
      "loss": 0.5889,
      "step": 213
    },
    {
      "epoch": 0.9185781354795439,
      "grad_norm": 2.0434712402893145,
      "learning_rate": 1.5708419435684463e-07,
      "loss": 0.5638,
      "step": 214
    },
    {
      "epoch": 0.9228705566733736,
      "grad_norm": 2.1757829918674916,
      "learning_rate": 1.4019499710726913e-07,
      "loss": 0.642,
      "step": 215
    },
    {
      "epoch": 0.9271629778672033,
      "grad_norm": 1.9838844707633234,
      "learning_rate": 1.2425322847218368e-07,
      "loss": 0.6747,
      "step": 216
    },
    {
      "epoch": 0.9314553990610329,
      "grad_norm": 2.3027124358123836,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.5337,
      "step": 217
    },
    {
      "epoch": 0.9357478202548625,
      "grad_norm": 1.8834569228162623,
      "learning_rate": 9.522422325404234e-08,
      "loss": 0.4917,
      "step": 218
    },
    {
      "epoch": 0.9400402414486921,
      "grad_norm": 1.8818328820882426,
      "learning_rate": 8.214264593307097e-08,
      "loss": 0.5806,
      "step": 219
    },
    {
      "epoch": 0.9443326626425218,
      "grad_norm": 2.121602640833575,
      "learning_rate": 7.001981464747565e-08,
      "loss": 0.6032,
      "step": 220
    },
    {
      "epoch": 0.9486250838363515,
      "grad_norm": 1.7145222003788387,
      "learning_rate": 5.8858092767236084e-08,
      "loss": 0.5455,
      "step": 221
    },
    {
      "epoch": 0.9529175050301811,
      "grad_norm": 1.9516038348847442,
      "learning_rate": 4.865965629214819e-08,
      "loss": 0.4984,
      "step": 222
    },
    {
      "epoch": 0.9572099262240107,
      "grad_norm": 2.156767335001454,
      "learning_rate": 3.9426493427611177e-08,
      "loss": 0.5284,
      "step": 223
    },
    {
      "epoch": 0.9615023474178404,
      "grad_norm": 1.8089215543018735,
      "learning_rate": 3.1160404197018155e-08,
      "loss": 0.5852,
      "step": 224
    },
    {
      "epoch": 0.96579476861167,
      "grad_norm": 3.2819358158442062,
      "learning_rate": 2.386300009084408e-08,
      "loss": 0.535,
      "step": 225
    },
    {
      "epoch": 0.9700871898054997,
      "grad_norm": 1.7304415336384433,
      "learning_rate": 1.753570375247815e-08,
      "loss": 0.5152,
      "step": 226
    },
    {
      "epoch": 0.9743796109993293,
      "grad_norm": 2.7057456233183785,
      "learning_rate": 1.2179748700879013e-08,
      "loss": 0.5592,
      "step": 227
    },
    {
      "epoch": 0.9786720321931589,
      "grad_norm": 2.0005265863891872,
      "learning_rate": 7.796179090094891e-09,
      "loss": 0.548,
      "step": 228
    },
    {
      "epoch": 0.9829644533869886,
      "grad_norm": 3.129587457460373,
      "learning_rate": 4.385849505708084e-09,
      "loss": 0.5154,
      "step": 229
    },
    {
      "epoch": 0.9872568745808182,
      "grad_norm": 2.082407038336782,
      "learning_rate": 1.9494247982282386e-09,
      "loss": 0.5263,
      "step": 230
    },
    {
      "epoch": 0.9915492957746479,
      "grad_norm": 2.035928140228953,
      "learning_rate": 4.87379953478806e-10,
      "loss": 0.5001,
      "step": 231
    },
    {
      "epoch": 0.9958417169684776,
      "grad_norm": 1.9493320834510575,
      "learning_rate": 0.0,
      "loss": 0.5539,
      "step": 232
    },
    {
      "epoch": 0.9958417169684776,
      "step": 232,
      "total_flos": 57740051464192.0,
      "train_loss": 0.5700135146235598,
      "train_runtime": 22473.305,
      "train_samples_per_second": 1.327,
      "train_steps_per_second": 0.01
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 232,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5000.0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 57740051464192.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}