{
  "best_metric": 0.5032929182052612,
  "best_model_checkpoint": "miner_id_24/checkpoint-500",
  "epoch": 0.6809669731018045,
  "eval_steps": 100,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001361933946203609,
      "grad_norm": 6.723554611206055,
      "learning_rate": 5e-06,
      "loss": 0.8744,
      "step": 1
    },
    {
      "epoch": 0.001361933946203609,
      "eval_loss": 1.29380464553833,
      "eval_runtime": 151.1717,
      "eval_samples_per_second": 8.183,
      "eval_steps_per_second": 2.051,
      "step": 1
    },
    {
      "epoch": 0.002723867892407218,
      "grad_norm": 4.329668045043945,
      "learning_rate": 1e-05,
      "loss": 1.1685,
      "step": 2
    },
    {
      "epoch": 0.0040858018386108275,
      "grad_norm": 3.8606443405151367,
      "learning_rate": 1.5e-05,
      "loss": 0.9956,
      "step": 3
    },
    {
      "epoch": 0.005447735784814436,
      "grad_norm": 2.3737833499908447,
      "learning_rate": 2e-05,
      "loss": 0.8493,
      "step": 4
    },
    {
      "epoch": 0.0068096697310180455,
      "grad_norm": 2.2130537033081055,
      "learning_rate": 2.5e-05,
      "loss": 1.0386,
      "step": 5
    },
    {
      "epoch": 0.008171603677221655,
      "grad_norm": 1.7720967531204224,
      "learning_rate": 3e-05,
      "loss": 0.9583,
      "step": 6
    },
    {
      "epoch": 0.009533537623425264,
      "grad_norm": 2.0626909732818604,
      "learning_rate": 3.5e-05,
      "loss": 0.9117,
      "step": 7
    },
    {
      "epoch": 0.010895471569628872,
      "grad_norm": 1.2568459510803223,
      "learning_rate": 4e-05,
      "loss": 0.7207,
      "step": 8
    },
    {
      "epoch": 0.012257405515832482,
      "grad_norm": 1.2235935926437378,
      "learning_rate": 4.5e-05,
      "loss": 0.7072,
      "step": 9
    },
    {
      "epoch": 0.013619339462036091,
      "grad_norm": 1.4364486932754517,
      "learning_rate": 5e-05,
      "loss": 0.7842,
      "step": 10
    },
    {
      "epoch": 0.0149812734082397,
      "grad_norm": 1.6137148141860962,
      "learning_rate": 5.500000000000001e-05,
      "loss": 0.7727,
      "step": 11
    },
    {
      "epoch": 0.01634320735444331,
      "grad_norm": 1.0391216278076172,
      "learning_rate": 6e-05,
      "loss": 0.6199,
      "step": 12
    },
    {
      "epoch": 0.017705141300646918,
      "grad_norm": 1.2817028760910034,
      "learning_rate": 6.500000000000001e-05,
      "loss": 0.8018,
      "step": 13
    },
    {
      "epoch": 0.01906707524685053,
      "grad_norm": 1.4122101068496704,
      "learning_rate": 7e-05,
      "loss": 0.7091,
      "step": 14
    },
    {
      "epoch": 0.020429009193054137,
      "grad_norm": 1.2346439361572266,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.6746,
      "step": 15
    },
    {
      "epoch": 0.021790943139257744,
      "grad_norm": 1.101813554763794,
      "learning_rate": 8e-05,
      "loss": 0.7234,
      "step": 16
    },
    {
      "epoch": 0.023152877085461356,
      "grad_norm": 1.3349500894546509,
      "learning_rate": 8.5e-05,
      "loss": 0.7135,
      "step": 17
    },
    {
      "epoch": 0.024514811031664963,
      "grad_norm": 1.274173617362976,
      "learning_rate": 9e-05,
      "loss": 0.6312,
      "step": 18
    },
    {
      "epoch": 0.025876744977868574,
      "grad_norm": 1.210764765739441,
      "learning_rate": 9.5e-05,
      "loss": 0.6636,
      "step": 19
    },
    {
      "epoch": 0.027238678924072182,
      "grad_norm": 1.1875956058502197,
      "learning_rate": 0.0001,
      "loss": 0.6469,
      "step": 20
    },
    {
      "epoch": 0.028600612870275793,
      "grad_norm": 1.3556970357894897,
      "learning_rate": 9.999892908320647e-05,
      "loss": 0.6853,
      "step": 21
    },
    {
      "epoch": 0.0299625468164794,
      "grad_norm": 1.341584324836731,
      "learning_rate": 9.999571637870036e-05,
      "loss": 0.6772,
      "step": 22
    },
    {
      "epoch": 0.03132448076268301,
      "grad_norm": 1.3170078992843628,
      "learning_rate": 9.999036202410325e-05,
      "loss": 0.6173,
      "step": 23
    },
    {
      "epoch": 0.03268641470888662,
      "grad_norm": 1.4024509191513062,
      "learning_rate": 9.998286624877786e-05,
      "loss": 0.679,
      "step": 24
    },
    {
      "epoch": 0.03404834865509023,
      "grad_norm": 1.5314940214157104,
      "learning_rate": 9.997322937381829e-05,
      "loss": 0.6211,
      "step": 25
    },
    {
      "epoch": 0.035410282601293835,
      "grad_norm": 1.3796993494033813,
      "learning_rate": 9.996145181203615e-05,
      "loss": 0.663,
      "step": 26
    },
    {
      "epoch": 0.03677221654749745,
      "grad_norm": 1.162498116493225,
      "learning_rate": 9.994753406794301e-05,
      "loss": 0.5981,
      "step": 27
    },
    {
      "epoch": 0.03813415049370106,
      "grad_norm": 1.214526653289795,
      "learning_rate": 9.99314767377287e-05,
      "loss": 0.7738,
      "step": 28
    },
    {
      "epoch": 0.03949608443990466,
      "grad_norm": 1.251236915588379,
      "learning_rate": 9.991328050923581e-05,
      "loss": 0.6466,
      "step": 29
    },
    {
      "epoch": 0.04085801838610827,
      "grad_norm": 1.2019500732421875,
      "learning_rate": 9.989294616193017e-05,
      "loss": 0.6285,
      "step": 30
    },
    {
      "epoch": 0.042219952332311884,
      "grad_norm": 1.506635308265686,
      "learning_rate": 9.98704745668676e-05,
      "loss": 0.7005,
      "step": 31
    },
    {
      "epoch": 0.04358188627851549,
      "grad_norm": 1.2021490335464478,
      "learning_rate": 9.98458666866564e-05,
      "loss": 0.6355,
      "step": 32
    },
    {
      "epoch": 0.0449438202247191,
      "grad_norm": 1.1416049003601074,
      "learning_rate": 9.981912357541627e-05,
      "loss": 0.6532,
      "step": 33
    },
    {
      "epoch": 0.04630575417092271,
      "grad_norm": 1.165168285369873,
      "learning_rate": 9.97902463787331e-05,
      "loss": 0.6076,
      "step": 34
    },
    {
      "epoch": 0.04766768811712632,
      "grad_norm": 1.1205216646194458,
      "learning_rate": 9.975923633360985e-05,
      "loss": 0.5811,
      "step": 35
    },
    {
      "epoch": 0.049029622063329927,
      "grad_norm": 1.0092084407806396,
      "learning_rate": 9.972609476841367e-05,
      "loss": 0.6334,
      "step": 36
    },
    {
      "epoch": 0.05039155600953354,
      "grad_norm": 1.0040332078933716,
      "learning_rate": 9.969082310281891e-05,
      "loss": 0.5839,
      "step": 37
    },
    {
      "epoch": 0.05175348995573715,
      "grad_norm": 1.1557093858718872,
      "learning_rate": 9.965342284774632e-05,
      "loss": 0.7276,
      "step": 38
    },
    {
      "epoch": 0.05311542390194075,
      "grad_norm": 1.0903106927871704,
      "learning_rate": 9.961389560529836e-05,
      "loss": 0.6504,
      "step": 39
    },
    {
      "epoch": 0.054477357848144364,
      "grad_norm": 1.1416411399841309,
      "learning_rate": 9.957224306869053e-05,
      "loss": 0.5472,
      "step": 40
    },
    {
      "epoch": 0.055839291794347976,
      "grad_norm": 0.9939647912979126,
      "learning_rate": 9.952846702217886e-05,
      "loss": 0.6051,
      "step": 41
    },
    {
      "epoch": 0.05720122574055159,
      "grad_norm": 1.759358286857605,
      "learning_rate": 9.948256934098352e-05,
      "loss": 0.6047,
      "step": 42
    },
    {
      "epoch": 0.05856315968675519,
      "grad_norm": 1.137244701385498,
      "learning_rate": 9.943455199120837e-05,
      "loss": 0.6423,
      "step": 43
    },
    {
      "epoch": 0.0599250936329588,
      "grad_norm": 1.1570804119110107,
      "learning_rate": 9.938441702975689e-05,
      "loss": 0.6577,
      "step": 44
    },
    {
      "epoch": 0.06128702757916241,
      "grad_norm": 1.2725093364715576,
      "learning_rate": 9.933216660424395e-05,
      "loss": 0.7253,
      "step": 45
    },
    {
      "epoch": 0.06264896152536602,
      "grad_norm": 1.4316565990447998,
      "learning_rate": 9.927780295290389e-05,
      "loss": 0.7288,
      "step": 46
    },
    {
      "epoch": 0.06401089547156963,
      "grad_norm": 1.6259329319000244,
      "learning_rate": 9.922132840449459e-05,
      "loss": 0.8127,
      "step": 47
    },
    {
      "epoch": 0.06537282941777324,
      "grad_norm": 1.5872222185134888,
      "learning_rate": 9.916274537819775e-05,
      "loss": 0.744,
      "step": 48
    },
    {
      "epoch": 0.06673476336397685,
      "grad_norm": 2.1103341579437256,
      "learning_rate": 9.91020563835152e-05,
      "loss": 0.7684,
      "step": 49
    },
    {
      "epoch": 0.06809669731018046,
      "grad_norm": 2.860896587371826,
      "learning_rate": 9.903926402016153e-05,
      "loss": 0.8936,
      "step": 50
    },
    {
      "epoch": 0.06945863125638406,
      "grad_norm": 0.8009760975837708,
      "learning_rate": 9.897437097795257e-05,
      "loss": 0.6914,
      "step": 51
    },
    {
      "epoch": 0.07082056520258767,
      "grad_norm": 0.8302339911460876,
      "learning_rate": 9.890738003669029e-05,
      "loss": 0.7927,
      "step": 52
    },
    {
      "epoch": 0.07218249914879128,
      "grad_norm": 0.8460277318954468,
      "learning_rate": 9.883829406604363e-05,
      "loss": 0.6558,
      "step": 53
    },
    {
      "epoch": 0.0735444330949949,
      "grad_norm": 0.6343598961830139,
      "learning_rate": 9.876711602542563e-05,
      "loss": 0.6166,
      "step": 54
    },
    {
      "epoch": 0.0749063670411985,
      "grad_norm": 0.6294136643409729,
      "learning_rate": 9.869384896386668e-05,
      "loss": 0.7247,
      "step": 55
    },
    {
      "epoch": 0.07626830098740212,
      "grad_norm": 0.6533523201942444,
      "learning_rate": 9.861849601988383e-05,
      "loss": 0.6108,
      "step": 56
    },
    {
      "epoch": 0.07763023493360573,
      "grad_norm": 0.7479948997497559,
      "learning_rate": 9.854106042134641e-05,
      "loss": 0.6284,
      "step": 57
    },
    {
      "epoch": 0.07899216887980932,
      "grad_norm": 0.6209579110145569,
      "learning_rate": 9.846154548533773e-05,
      "loss": 0.5844,
      "step": 58
    },
    {
      "epoch": 0.08035410282601294,
      "grad_norm": 0.6447109580039978,
      "learning_rate": 9.837995461801299e-05,
      "loss": 0.5638,
      "step": 59
    },
    {
      "epoch": 0.08171603677221655,
      "grad_norm": 0.6410228610038757,
      "learning_rate": 9.829629131445342e-05,
      "loss": 0.6504,
      "step": 60
    },
    {
      "epoch": 0.08307797071842016,
      "grad_norm": 0.6668742895126343,
      "learning_rate": 9.821055915851647e-05,
      "loss": 0.6025,
      "step": 61
    },
    {
      "epoch": 0.08443990466462377,
      "grad_norm": 1.3477871417999268,
      "learning_rate": 9.812276182268236e-05,
      "loss": 0.5365,
      "step": 62
    },
    {
      "epoch": 0.08580183861082738,
      "grad_norm": 0.6480602622032166,
      "learning_rate": 9.803290306789676e-05,
      "loss": 0.5192,
      "step": 63
    },
    {
      "epoch": 0.08716377255703098,
      "grad_norm": 0.6879251003265381,
      "learning_rate": 9.794098674340965e-05,
      "loss": 0.5351,
      "step": 64
    },
    {
      "epoch": 0.08852570650323459,
      "grad_norm": 0.7362158894538879,
      "learning_rate": 9.784701678661045e-05,
      "loss": 0.5237,
      "step": 65
    },
    {
      "epoch": 0.0898876404494382,
      "grad_norm": 0.7937474250793457,
      "learning_rate": 9.775099722285935e-05,
      "loss": 0.6137,
      "step": 66
    },
    {
      "epoch": 0.09124957439564181,
      "grad_norm": 0.8558733463287354,
      "learning_rate": 9.765293216531486e-05,
      "loss": 0.6135,
      "step": 67
    },
    {
      "epoch": 0.09261150834184542,
      "grad_norm": 0.9930456876754761,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.5677,
      "step": 68
    },
    {
      "epoch": 0.09397344228804903,
      "grad_norm": 0.816164493560791,
      "learning_rate": 9.74506824594107e-05,
      "loss": 0.5234,
      "step": 69
    },
    {
      "epoch": 0.09533537623425264,
      "grad_norm": 0.8402704000473022,
      "learning_rate": 9.73465064747553e-05,
      "loss": 0.4945,
      "step": 70
    },
    {
      "epoch": 0.09669731018045624,
      "grad_norm": 0.7770735025405884,
      "learning_rate": 9.724030232334391e-05,
      "loss": 0.5315,
      "step": 71
    },
    {
      "epoch": 0.09805924412665985,
      "grad_norm": 0.998912513256073,
      "learning_rate": 9.713207455460894e-05,
      "loss": 0.5267,
      "step": 72
    },
    {
      "epoch": 0.09942117807286346,
      "grad_norm": 0.8477897644042969,
      "learning_rate": 9.702182780466775e-05,
      "loss": 0.551,
      "step": 73
    },
    {
      "epoch": 0.10078311201906708,
      "grad_norm": 0.8563628792762756,
      "learning_rate": 9.690956679612421e-05,
      "loss": 0.6059,
      "step": 74
    },
    {
      "epoch": 0.10214504596527069,
      "grad_norm": 0.79013592004776,
      "learning_rate": 9.67952963378663e-05,
      "loss": 0.562,
      "step": 75
    },
    {
      "epoch": 0.1035069799114743,
      "grad_norm": 0.7895680069923401,
      "learning_rate": 9.667902132486009e-05,
      "loss": 0.5271,
      "step": 76
    },
    {
      "epoch": 0.10486891385767791,
      "grad_norm": 0.9388853311538696,
      "learning_rate": 9.656074673794018e-05,
      "loss": 0.6054,
      "step": 77
    },
    {
      "epoch": 0.1062308478038815,
      "grad_norm": 0.9046183228492737,
      "learning_rate": 9.644047764359622e-05,
      "loss": 0.5627,
      "step": 78
    },
    {
      "epoch": 0.10759278175008512,
      "grad_norm": 0.9057890772819519,
      "learning_rate": 9.631821919375591e-05,
      "loss": 0.602,
      "step": 79
    },
    {
      "epoch": 0.10895471569628873,
      "grad_norm": 0.9569646716117859,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.539,
      "step": 80
    },
    {
      "epoch": 0.11031664964249234,
      "grad_norm": 0.9439347386360168,
      "learning_rate": 9.606775526115963e-05,
      "loss": 0.5636,
      "step": 81
    },
    {
      "epoch": 0.11167858358869595,
      "grad_norm": 0.8295052647590637,
      "learning_rate": 9.593956050744492e-05,
      "loss": 0.544,
      "step": 82
    },
    {
      "epoch": 0.11304051753489956,
      "grad_norm": 0.9805134534835815,
      "learning_rate": 9.580939785585681e-05,
      "loss": 0.589,
      "step": 83
    },
    {
      "epoch": 0.11440245148110317,
      "grad_norm": 0.8532008528709412,
      "learning_rate": 9.567727288213005e-05,
      "loss": 0.5225,
      "step": 84
    },
    {
      "epoch": 0.11576438542730677,
      "grad_norm": 1.0181292295455933,
      "learning_rate": 9.554319124605879e-05,
      "loss": 0.6054,
      "step": 85
    },
    {
      "epoch": 0.11712631937351038,
      "grad_norm": 0.882378339767456,
      "learning_rate": 9.540715869125407e-05,
      "loss": 0.5787,
      "step": 86
    },
    {
      "epoch": 0.118488253319714,
      "grad_norm": 0.8436698913574219,
      "learning_rate": 9.526918104489777e-05,
      "loss": 0.5484,
      "step": 87
    },
    {
      "epoch": 0.1198501872659176,
      "grad_norm": 1.0983854532241821,
      "learning_rate": 9.512926421749304e-05,
      "loss": 0.6058,
      "step": 88
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 0.9240631461143494,
      "learning_rate": 9.498741420261108e-05,
      "loss": 0.608,
      "step": 89
    },
    {
      "epoch": 0.12257405515832483,
      "grad_norm": 0.8889957070350647,
      "learning_rate": 9.484363707663442e-05,
      "loss": 0.5661,
      "step": 90
    },
    {
      "epoch": 0.12393598910452842,
      "grad_norm": 0.883852481842041,
      "learning_rate": 9.469793899849661e-05,
      "loss": 0.5305,
      "step": 91
    },
    {
      "epoch": 0.12529792305073204,
      "grad_norm": 0.9092103838920593,
      "learning_rate": 9.45503262094184e-05,
      "loss": 0.5246,
      "step": 92
    },
    {
      "epoch": 0.12665985699693566,
      "grad_norm": 0.990210771560669,
      "learning_rate": 9.440080503264037e-05,
      "loss": 0.5956,
      "step": 93
    },
    {
      "epoch": 0.12802179094313926,
      "grad_norm": 0.981751024723053,
      "learning_rate": 9.42493818731521e-05,
      "loss": 0.5482,
      "step": 94
    },
    {
      "epoch": 0.12938372488934285,
      "grad_norm": 1.1555558443069458,
      "learning_rate": 9.409606321741775e-05,
      "loss": 0.7,
      "step": 95
    },
    {
      "epoch": 0.13074565883554648,
      "grad_norm": 1.1215264797210693,
      "learning_rate": 9.394085563309827e-05,
      "loss": 0.5967,
      "step": 96
    },
    {
      "epoch": 0.13210759278175008,
      "grad_norm": 1.1388212442398071,
      "learning_rate": 9.378376576876999e-05,
      "loss": 0.5699,
      "step": 97
    },
    {
      "epoch": 0.1334695267279537,
      "grad_norm": 1.3690537214279175,
      "learning_rate": 9.362480035363986e-05,
      "loss": 0.6795,
      "step": 98
    },
    {
      "epoch": 0.1348314606741573,
      "grad_norm": 1.632396936416626,
      "learning_rate": 9.34639661972572e-05,
      "loss": 0.7396,
      "step": 99
    },
    {
      "epoch": 0.13619339462036092,
      "grad_norm": 2.967085838317871,
      "learning_rate": 9.330127018922194e-05,
      "loss": 0.9022,
      "step": 100
    },
    {
      "epoch": 0.13619339462036092,
      "eval_loss": 0.6248776316642761,
      "eval_runtime": 153.1736,
      "eval_samples_per_second": 8.076,
      "eval_steps_per_second": 2.024,
      "step": 100
    },
    {
      "epoch": 0.13755532856656452,
      "grad_norm": 0.6180322766304016,
      "learning_rate": 9.31367192988896e-05,
      "loss": 0.5993,
      "step": 101
    },
    {
      "epoch": 0.13891726251276812,
      "grad_norm": 0.6882027387619019,
      "learning_rate": 9.297032057507264e-05,
      "loss": 0.7111,
      "step": 102
    },
    {
      "epoch": 0.14027919645897174,
      "grad_norm": 0.5713951587677002,
      "learning_rate": 9.280208114573859e-05,
      "loss": 0.5262,
      "step": 103
    },
    {
      "epoch": 0.14164113040517534,
      "grad_norm": 0.5399894714355469,
      "learning_rate": 9.263200821770461e-05,
      "loss": 0.5569,
      "step": 104
    },
    {
      "epoch": 0.14300306435137897,
      "grad_norm": 0.6017151474952698,
      "learning_rate": 9.246010907632895e-05,
      "loss": 0.7356,
      "step": 105
    },
    {
      "epoch": 0.14436499829758256,
      "grad_norm": 0.5448001623153687,
      "learning_rate": 9.228639108519868e-05,
      "loss": 0.6401,
      "step": 106
    },
    {
      "epoch": 0.1457269322437862,
      "grad_norm": 0.8159662485122681,
      "learning_rate": 9.211086168581433e-05,
      "loss": 0.6879,
      "step": 107
    },
    {
      "epoch": 0.1470888661899898,
      "grad_norm": 0.6179919838905334,
      "learning_rate": 9.193352839727121e-05,
      "loss": 0.555,
      "step": 108
    },
    {
      "epoch": 0.14845080013619338,
      "grad_norm": 0.5790054798126221,
      "learning_rate": 9.175439881593716e-05,
      "loss": 0.5782,
      "step": 109
    },
    {
      "epoch": 0.149812734082397,
      "grad_norm": 0.5688661932945251,
      "learning_rate": 9.157348061512727e-05,
      "loss": 0.5538,
      "step": 110
    },
    {
      "epoch": 0.1511746680286006,
      "grad_norm": 0.5580296516418457,
      "learning_rate": 9.139078154477512e-05,
      "loss": 0.5394,
      "step": 111
    },
    {
      "epoch": 0.15253660197480423,
      "grad_norm": 0.6444279551506042,
      "learning_rate": 9.120630943110077e-05,
      "loss": 0.6893,
      "step": 112
    },
    {
      "epoch": 0.15389853592100783,
      "grad_norm": 0.6436907649040222,
      "learning_rate": 9.102007217627568e-05,
      "loss": 0.4997,
      "step": 113
    },
    {
      "epoch": 0.15526046986721145,
      "grad_norm": 0.7137179374694824,
      "learning_rate": 9.083207775808396e-05,
      "loss": 0.523,
      "step": 114
    },
    {
      "epoch": 0.15662240381341505,
      "grad_norm": 0.6412702798843384,
      "learning_rate": 9.064233422958077e-05,
      "loss": 0.5796,
      "step": 115
    },
    {
      "epoch": 0.15798433775961865,
      "grad_norm": 0.7804475426673889,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.5524,
      "step": 116
    },
    {
      "epoch": 0.15934627170582227,
      "grad_norm": 0.785469651222229,
      "learning_rate": 9.025763242814291e-05,
      "loss": 0.6496,
      "step": 117
    },
    {
      "epoch": 0.16070820565202587,
      "grad_norm": 0.7303288578987122,
      "learning_rate": 9.006269063455304e-05,
      "loss": 0.5379,
      "step": 118
    },
    {
      "epoch": 0.1620701395982295,
      "grad_norm": 0.7512404322624207,
      "learning_rate": 8.986603268863536e-05,
      "loss": 0.5171,
      "step": 119
    },
    {
      "epoch": 0.1634320735444331,
      "grad_norm": 0.8327203392982483,
      "learning_rate": 8.966766701456177e-05,
      "loss": 0.5642,
      "step": 120
    },
    {
      "epoch": 0.1647940074906367,
      "grad_norm": 0.844458818435669,
      "learning_rate": 8.94676021096575e-05,
      "loss": 0.6107,
      "step": 121
    },
    {
      "epoch": 0.16615594143684032,
      "grad_norm": 0.9365436434745789,
      "learning_rate": 8.926584654403724e-05,
      "loss": 0.4959,
      "step": 122
    },
    {
      "epoch": 0.1675178753830439,
      "grad_norm": 0.7789584398269653,
      "learning_rate": 8.906240896023794e-05,
      "loss": 0.5299,
      "step": 123
    },
    {
      "epoch": 0.16887980932924754,
      "grad_norm": 0.8737173676490784,
      "learning_rate": 8.885729807284856e-05,
      "loss": 0.4669,
      "step": 124
    },
    {
      "epoch": 0.17024174327545114,
      "grad_norm": 1.0036945343017578,
      "learning_rate": 8.865052266813685e-05,
      "loss": 0.5851,
      "step": 125
    },
    {
      "epoch": 0.17160367722165476,
      "grad_norm": 0.9954502582550049,
      "learning_rate": 8.844209160367299e-05,
      "loss": 0.547,
      "step": 126
    },
    {
      "epoch": 0.17296561116785836,
      "grad_norm": 0.9633464813232422,
      "learning_rate": 8.823201380795001e-05,
      "loss": 0.6711,
      "step": 127
    },
    {
      "epoch": 0.17432754511406195,
      "grad_norm": 1.0423957109451294,
      "learning_rate": 8.802029828000156e-05,
      "loss": 0.6222,
      "step": 128
    },
    {
      "epoch": 0.17568947906026558,
      "grad_norm": 1.1178964376449585,
      "learning_rate": 8.780695408901613e-05,
      "loss": 0.5752,
      "step": 129
    },
    {
      "epoch": 0.17705141300646918,
      "grad_norm": 0.7944217324256897,
      "learning_rate": 8.759199037394887e-05,
      "loss": 0.5264,
      "step": 130
    },
    {
      "epoch": 0.1784133469526728,
      "grad_norm": 0.8660529255867004,
      "learning_rate": 8.737541634312985e-05,
      "loss": 0.5387,
      "step": 131
    },
    {
      "epoch": 0.1797752808988764,
      "grad_norm": 1.032966136932373,
      "learning_rate": 8.715724127386972e-05,
      "loss": 0.5572,
      "step": 132
    },
    {
      "epoch": 0.18113721484508002,
      "grad_norm": 0.8581622242927551,
      "learning_rate": 8.693747451206232e-05,
      "loss": 0.5426,
      "step": 133
    },
    {
      "epoch": 0.18249914879128362,
      "grad_norm": 0.9016491770744324,
      "learning_rate": 8.671612547178428e-05,
      "loss": 0.5126,
      "step": 134
    },
    {
      "epoch": 0.18386108273748722,
      "grad_norm": 0.7935693264007568,
      "learning_rate": 8.649320363489179e-05,
      "loss": 0.5126,
      "step": 135
    },
    {
      "epoch": 0.18522301668369084,
      "grad_norm": 0.7998529076576233,
      "learning_rate": 8.626871855061438e-05,
      "loss": 0.4728,
      "step": 136
    },
    {
      "epoch": 0.18658495062989444,
      "grad_norm": 1.1158883571624756,
      "learning_rate": 8.604267983514594e-05,
      "loss": 0.5456,
      "step": 137
    },
    {
      "epoch": 0.18794688457609807,
      "grad_norm": 0.9313867688179016,
      "learning_rate": 8.581509717123273e-05,
      "loss": 0.5756,
      "step": 138
    },
    {
      "epoch": 0.18930881852230166,
      "grad_norm": 0.8501096367835999,
      "learning_rate": 8.558598030775857e-05,
      "loss": 0.5468,
      "step": 139
    },
    {
      "epoch": 0.1906707524685053,
      "grad_norm": 0.8502845764160156,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.5409,
      "step": 140
    },
    {
      "epoch": 0.1920326864147089,
      "grad_norm": 0.8347089886665344,
      "learning_rate": 8.51231833058426e-05,
      "loss": 0.517,
      "step": 141
    },
    {
      "epoch": 0.19339462036091248,
      "grad_norm": 0.9152666330337524,
      "learning_rate": 8.488952299208401e-05,
      "loss": 0.529,
      "step": 142
    },
    {
      "epoch": 0.1947565543071161,
      "grad_norm": 1.0104295015335083,
      "learning_rate": 8.46543681272818e-05,
      "loss": 0.578,
      "step": 143
    },
    {
      "epoch": 0.1961184882533197,
      "grad_norm": 1.0618577003479004,
      "learning_rate": 8.44177287846877e-05,
      "loss": 0.5785,
      "step": 144
    },
    {
      "epoch": 0.19748042219952333,
      "grad_norm": 1.1052486896514893,
      "learning_rate": 8.417961510114356e-05,
      "loss": 0.5952,
      "step": 145
    },
    {
      "epoch": 0.19884235614572693,
      "grad_norm": 1.0076816082000732,
      "learning_rate": 8.39400372766471e-05,
      "loss": 0.5691,
      "step": 146
    },
    {
      "epoch": 0.20020429009193055,
      "grad_norm": 1.0509177446365356,
      "learning_rate": 8.36990055739149e-05,
      "loss": 0.543,
      "step": 147
    },
    {
      "epoch": 0.20156622403813415,
      "grad_norm": 1.0742003917694092,
      "learning_rate": 8.345653031794292e-05,
      "loss": 0.4739,
      "step": 148
    },
    {
      "epoch": 0.20292815798433775,
      "grad_norm": 1.5066368579864502,
      "learning_rate": 8.321262189556409e-05,
      "loss": 0.5987,
      "step": 149
    },
    {
      "epoch": 0.20429009193054137,
      "grad_norm": 3.2690324783325195,
      "learning_rate": 8.296729075500344e-05,
      "loss": 0.9138,
      "step": 150
    },
    {
      "epoch": 0.20565202587674497,
      "grad_norm": 0.5544908046722412,
      "learning_rate": 8.272054740543052e-05,
      "loss": 0.5583,
      "step": 151
    },
    {
      "epoch": 0.2070139598229486,
      "grad_norm": 0.6558340787887573,
      "learning_rate": 8.247240241650918e-05,
      "loss": 0.7693,
      "step": 152
    },
    {
      "epoch": 0.2083758937691522,
      "grad_norm": 0.48983529210090637,
      "learning_rate": 8.222286641794488e-05,
      "loss": 0.5553,
      "step": 153
    },
    {
      "epoch": 0.20973782771535582,
      "grad_norm": 0.5212178230285645,
      "learning_rate": 8.197195009902924e-05,
      "loss": 0.597,
      "step": 154
    },
    {
      "epoch": 0.21109976166155942,
      "grad_norm": 0.515625536441803,
      "learning_rate": 8.171966420818228e-05,
      "loss": 0.6166,
      "step": 155
    },
    {
      "epoch": 0.212461695607763,
      "grad_norm": 0.54818195104599,
      "learning_rate": 8.146601955249188e-05,
      "loss": 0.6416,
      "step": 156
    },
    {
      "epoch": 0.21382362955396664,
      "grad_norm": 0.8764201402664185,
      "learning_rate": 8.121102699725089e-05,
      "loss": 0.5537,
      "step": 157
    },
    {
      "epoch": 0.21518556350017023,
      "grad_norm": 0.49704718589782715,
      "learning_rate": 8.095469746549172e-05,
      "loss": 0.5665,
      "step": 158
    },
    {
      "epoch": 0.21654749744637386,
      "grad_norm": 0.5559098124504089,
      "learning_rate": 8.069704193751832e-05,
      "loss": 0.5524,
      "step": 159
    },
    {
      "epoch": 0.21790943139257746,
      "grad_norm": 0.6416535377502441,
      "learning_rate": 8.043807145043604e-05,
      "loss": 0.4999,
      "step": 160
    },
    {
      "epoch": 0.21927136533878108,
      "grad_norm": 0.58646559715271,
      "learning_rate": 8.017779709767858e-05,
      "loss": 0.5707,
      "step": 161
    },
    {
      "epoch": 0.22063329928498468,
      "grad_norm": 0.6208726763725281,
      "learning_rate": 7.991623002853296e-05,
      "loss": 0.6492,
      "step": 162
    },
    {
      "epoch": 0.22199523323118828,
      "grad_norm": 0.6617854833602905,
      "learning_rate": 7.965338144766186e-05,
      "loss": 0.6366,
      "step": 163
    },
    {
      "epoch": 0.2233571671773919,
      "grad_norm": 0.6566423177719116,
      "learning_rate": 7.938926261462366e-05,
      "loss": 0.6241,
      "step": 164
    },
    {
      "epoch": 0.2247191011235955,
      "grad_norm": 0.5700229406356812,
      "learning_rate": 7.912388484339012e-05,
      "loss": 0.5184,
      "step": 165
    },
    {
      "epoch": 0.22608103506979912,
      "grad_norm": 0.8321027755737305,
      "learning_rate": 7.88572595018617e-05,
      "loss": 0.5799,
      "step": 166
    },
    {
      "epoch": 0.22744296901600272,
      "grad_norm": 0.6786854863166809,
      "learning_rate": 7.858939801138061e-05,
      "loss": 0.5187,
      "step": 167
    },
    {
      "epoch": 0.22880490296220635,
      "grad_norm": 0.7199569344520569,
      "learning_rate": 7.832031184624164e-05,
      "loss": 0.5468,
      "step": 168
    },
    {
      "epoch": 0.23016683690840994,
      "grad_norm": 0.7046637535095215,
      "learning_rate": 7.80500125332005e-05,
      "loss": 0.5081,
      "step": 169
    },
    {
      "epoch": 0.23152877085461354,
      "grad_norm": 0.7412410378456116,
      "learning_rate": 7.777851165098012e-05,
      "loss": 0.6227,
      "step": 170
    },
    {
      "epoch": 0.23289070480081717,
      "grad_norm": 0.9553371667861938,
      "learning_rate": 7.750582082977467e-05,
      "loss": 0.6606,
      "step": 171
    },
    {
      "epoch": 0.23425263874702076,
      "grad_norm": 0.8840705156326294,
      "learning_rate": 7.723195175075136e-05,
      "loss": 0.5144,
      "step": 172
    },
    {
      "epoch": 0.2356145726932244,
      "grad_norm": 0.7185872793197632,
      "learning_rate": 7.695691614555003e-05,
      "loss": 0.4539,
      "step": 173
    },
    {
      "epoch": 0.236976506639428,
      "grad_norm": 0.8578166365623474,
      "learning_rate": 7.668072579578058e-05,
      "loss": 0.5807,
      "step": 174
    },
    {
      "epoch": 0.23833844058563158,
      "grad_norm": 1.1541998386383057,
      "learning_rate": 7.64033925325184e-05,
      "loss": 0.5863,
      "step": 175
    },
    {
      "epoch": 0.2397003745318352,
      "grad_norm": 0.930377185344696,
      "learning_rate": 7.612492823579745e-05,
      "loss": 0.4979,
      "step": 176
    },
    {
      "epoch": 0.2410623084780388,
      "grad_norm": 0.7165592312812805,
      "learning_rate": 7.584534483410137e-05,
      "loss": 0.5226,
      "step": 177
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 0.9593399167060852,
      "learning_rate": 7.55646543038526e-05,
      "loss": 0.5132,
      "step": 178
    },
    {
      "epoch": 0.24378617637044603,
      "grad_norm": 1.0977400541305542,
      "learning_rate": 7.528286866889924e-05,
      "loss": 0.5814,
      "step": 179
    },
    {
      "epoch": 0.24514811031664965,
      "grad_norm": 0.7839404940605164,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.4726,
      "step": 180
    },
    {
      "epoch": 0.24651004426285325,
      "grad_norm": 0.8963006138801575,
      "learning_rate": 7.471606041430723e-05,
      "loss": 0.4825,
      "step": 181
    },
    {
      "epoch": 0.24787197820905685,
      "grad_norm": 0.8852623701095581,
      "learning_rate": 7.443106207484776e-05,
      "loss": 0.6063,
      "step": 182
    },
    {
      "epoch": 0.24923391215526047,
      "grad_norm": 0.7428604960441589,
      "learning_rate": 7.414501719000187e-05,
      "loss": 0.4815,
      "step": 183
    },
    {
      "epoch": 0.25059584610146407,
      "grad_norm": 0.7895641326904297,
      "learning_rate": 7.385793801298042e-05,
      "loss": 0.5377,
      "step": 184
    },
    {
      "epoch": 0.25195778004766767,
      "grad_norm": 0.7758123874664307,
      "learning_rate": 7.35698368412999e-05,
      "loss": 0.4772,
      "step": 185
    },
    {
      "epoch": 0.2533197139938713,
      "grad_norm": 0.7451211214065552,
      "learning_rate": 7.328072601625557e-05,
      "loss": 0.4657,
      "step": 186
    },
    {
      "epoch": 0.2546816479400749,
      "grad_norm": 0.805697500705719,
      "learning_rate": 7.2990617922393e-05,
      "loss": 0.581,
      "step": 187
    },
    {
      "epoch": 0.2560435818862785,
      "grad_norm": 0.7958231568336487,
      "learning_rate": 7.269952498697734e-05,
      "loss": 0.4579,
      "step": 188
    },
    {
      "epoch": 0.2574055158324821,
      "grad_norm": 0.8942751288414001,
      "learning_rate": 7.240745967946113e-05,
      "loss": 0.478,
      "step": 189
    },
    {
      "epoch": 0.2587674497786857,
      "grad_norm": 0.8354119062423706,
      "learning_rate": 7.211443451095007e-05,
      "loss": 0.5682,
      "step": 190
    },
    {
      "epoch": 0.26012938372488936,
      "grad_norm": 0.8643693923950195,
      "learning_rate": 7.18204620336671e-05,
      "loss": 0.5292,
      "step": 191
    },
    {
      "epoch": 0.26149131767109296,
      "grad_norm": 0.9560275673866272,
      "learning_rate": 7.152555484041476e-05,
      "loss": 0.6038,
      "step": 192
    },
    {
      "epoch": 0.26285325161729656,
      "grad_norm": 1.013757348060608,
      "learning_rate": 7.122972556403567e-05,
      "loss": 0.537,
      "step": 193
    },
    {
      "epoch": 0.26421518556350015,
      "grad_norm": 0.9771332740783691,
      "learning_rate": 7.09329868768714e-05,
      "loss": 0.5777,
      "step": 194
    },
    {
      "epoch": 0.26557711950970375,
      "grad_norm": 1.0889893770217896,
      "learning_rate": 7.063535149021973e-05,
      "loss": 0.5884,
      "step": 195
    },
    {
      "epoch": 0.2669390534559074,
      "grad_norm": 1.1477360725402832,
      "learning_rate": 7.033683215379002e-05,
      "loss": 0.5788,
      "step": 196
    },
    {
      "epoch": 0.268300987402111,
      "grad_norm": 1.2506556510925293,
      "learning_rate": 7.003744165515705e-05,
      "loss": 0.5713,
      "step": 197
    },
    {
      "epoch": 0.2696629213483146,
      "grad_norm": 1.0991449356079102,
      "learning_rate": 6.973719281921335e-05,
      "loss": 0.5345,
      "step": 198
    },
    {
      "epoch": 0.2710248552945182,
      "grad_norm": 1.4584896564483643,
      "learning_rate": 6.943609850761979e-05,
      "loss": 0.5686,
      "step": 199
    },
    {
      "epoch": 0.27238678924072185,
      "grad_norm": 2.23221755027771,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.617,
      "step": 200
    },
    {
      "epoch": 0.27238678924072185,
      "eval_loss": 0.5669613480567932,
      "eval_runtime": 153.1251,
      "eval_samples_per_second": 8.078,
      "eval_steps_per_second": 2.024,
      "step": 200
    },
    {
      "epoch": 0.27374872318692545,
      "grad_norm": 0.5449804663658142,
      "learning_rate": 6.883142508466054e-05,
      "loss": 0.6591,
      "step": 201
    },
    {
      "epoch": 0.27511065713312904,
      "grad_norm": 0.5590567588806152,
      "learning_rate": 6.852787187549182e-05,
      "loss": 0.5714,
      "step": 202
    },
    {
      "epoch": 0.27647259107933264,
      "grad_norm": 0.4900493025779724,
      "learning_rate": 6.82235249939575e-05,
      "loss": 0.5715,
      "step": 203
    },
    {
      "epoch": 0.27783452502553624,
      "grad_norm": 0.5053399205207825,
      "learning_rate": 6.7918397477265e-05,
      "loss": 0.5012,
      "step": 204
    },
    {
      "epoch": 0.2791964589717399,
      "grad_norm": 0.5021209120750427,
      "learning_rate": 6.761250239606169e-05,
      "loss": 0.6838,
      "step": 205
    },
    {
      "epoch": 0.2805583929179435,
      "grad_norm": 0.7311335802078247,
      "learning_rate": 6.730585285387465e-05,
      "loss": 0.6743,
      "step": 206
    },
    {
      "epoch": 0.2819203268641471,
      "grad_norm": 0.6293658018112183,
      "learning_rate": 6.699846198654971e-05,
      "loss": 0.6957,
      "step": 207
    },
    {
      "epoch": 0.2832822608103507,
      "grad_norm": 0.6651476621627808,
      "learning_rate": 6.669034296168855e-05,
      "loss": 0.5477,
      "step": 208
    },
    {
      "epoch": 0.2846441947565543,
      "grad_norm": 0.560538113117218,
      "learning_rate": 6.638150897808468e-05,
      "loss": 0.6135,
      "step": 209
    },
    {
      "epoch": 0.28600612870275793,
      "grad_norm": 0.5811340808868408,
      "learning_rate": 6.607197326515808e-05,
      "loss": 0.541,
      "step": 210
    },
    {
      "epoch": 0.28736806264896153,
      "grad_norm": 0.5839016437530518,
      "learning_rate": 6.57617490823885e-05,
      "loss": 0.5165,
      "step": 211
    },
    {
      "epoch": 0.28872999659516513,
      "grad_norm": 0.6369516849517822,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.5536,
      "step": 212
    },
    {
      "epoch": 0.2900919305413687,
      "grad_norm": 0.656177282333374,
      "learning_rate": 6.513928849212873e-05,
      "loss": 0.5031,
      "step": 213
    },
    {
      "epoch": 0.2914538644875724,
      "grad_norm": 0.6746971011161804,
      "learning_rate": 6.482707874877854e-05,
      "loss": 0.5098,
      "step": 214
    },
    {
      "epoch": 0.292815798433776,
      "grad_norm": 0.7272800207138062,
      "learning_rate": 6.451423386272312e-05,
      "loss": 0.5703,
      "step": 215
    },
    {
      "epoch": 0.2941777323799796,
      "grad_norm": 0.641337513923645,
      "learning_rate": 6.420076723519614e-05,
      "loss": 0.5582,
      "step": 216
    },
    {
      "epoch": 0.29553966632618317,
      "grad_norm": 0.669699490070343,
      "learning_rate": 6.388669229406462e-05,
      "loss": 0.4327,
      "step": 217
    },
    {
      "epoch": 0.29690160027238677,
      "grad_norm": 0.8017716407775879,
      "learning_rate": 6.357202249325371e-05,
      "loss": 0.5142,
      "step": 218
    },
    {
      "epoch": 0.2982635342185904,
      "grad_norm": 0.7105624079704285,
      "learning_rate": 6.32567713121704e-05,
      "loss": 0.466,
      "step": 219
    },
    {
      "epoch": 0.299625468164794,
      "grad_norm": 0.7463955879211426,
      "learning_rate": 6.294095225512603e-05,
      "loss": 0.5553,
      "step": 220
    },
    {
      "epoch": 0.3009874021109976,
      "grad_norm": 0.8410119414329529,
      "learning_rate": 6.26245788507579e-05,
      "loss": 0.4517,
      "step": 221
    },
    {
      "epoch": 0.3023493360572012,
      "grad_norm": 0.8605039119720459,
      "learning_rate": 6.230766465144967e-05,
      "loss": 0.5327,
      "step": 222
    },
    {
      "epoch": 0.3037112700034048,
      "grad_norm": 0.812928318977356,
      "learning_rate": 6.199022323275083e-05,
      "loss": 0.525,
      "step": 223
    },
    {
      "epoch": 0.30507320394960846,
      "grad_norm": 0.7954649925231934,
      "learning_rate": 6.167226819279528e-05,
      "loss": 0.5591,
      "step": 224
    },
    {
      "epoch": 0.30643513789581206,
      "grad_norm": 0.7110108733177185,
      "learning_rate": 6.135381315171867e-05,
      "loss": 0.4223,
      "step": 225
    },
    {
      "epoch": 0.30779707184201566,
      "grad_norm": 0.7785506844520569,
      "learning_rate": 6.103487175107507e-05,
      "loss": 0.4665,
      "step": 226
    },
    {
      "epoch": 0.30915900578821925,
      "grad_norm": 0.9806628823280334,
      "learning_rate": 6.071545765325254e-05,
      "loss": 0.5333,
      "step": 227
    },
    {
      "epoch": 0.3105209397344229,
      "grad_norm": 1.0425487756729126,
      "learning_rate": 6.0395584540887963e-05,
      "loss": 0.4521,
      "step": 228
    },
    {
      "epoch": 0.3118828736806265,
      "grad_norm": 0.7474079728126526,
      "learning_rate": 6.007526611628086e-05,
      "loss": 0.4565,
      "step": 229
    },
    {
      "epoch": 0.3132448076268301,
      "grad_norm": 0.7619843482971191,
      "learning_rate": 5.9754516100806423e-05,
      "loss": 0.5255,
      "step": 230
    },
    {
      "epoch": 0.3146067415730337,
      "grad_norm": 0.9061463475227356,
      "learning_rate": 5.9433348234327765e-05,
      "loss": 0.545,
      "step": 231
    },
    {
      "epoch": 0.3159686755192373,
      "grad_norm": 0.8243876099586487,
      "learning_rate": 5.911177627460739e-05,
      "loss": 0.5392,
      "step": 232
    },
    {
      "epoch": 0.31733060946544095,
      "grad_norm": 0.8262293934822083,
      "learning_rate": 5.8789813996717736e-05,
      "loss": 0.5126,
      "step": 233
    },
    {
      "epoch": 0.31869254341164455,
      "grad_norm": 0.6958490014076233,
      "learning_rate": 5.8467475192451226e-05,
      "loss": 0.5295,
      "step": 234
    },
    {
      "epoch": 0.32005447735784814,
      "grad_norm": 0.793052613735199,
      "learning_rate": 5.814477366972945e-05,
      "loss": 0.4696,
      "step": 235
    },
    {
      "epoch": 0.32141641130405174,
      "grad_norm": 0.8174700140953064,
      "learning_rate": 5.782172325201155e-05,
      "loss": 0.5164,
      "step": 236
    },
    {
      "epoch": 0.32277834525025534,
      "grad_norm": 0.782258927822113,
      "learning_rate": 5.749833777770225e-05,
      "loss": 0.5027,
      "step": 237
    },
    {
      "epoch": 0.324140279196459,
      "grad_norm": 0.808543860912323,
      "learning_rate": 5.717463109955896e-05,
      "loss": 0.4562,
      "step": 238
    },
    {
      "epoch": 0.3255022131426626,
      "grad_norm": 0.6752930879592896,
      "learning_rate": 5.685061708409841e-05,
      "loss": 0.4367,
      "step": 239
    },
    {
      "epoch": 0.3268641470888662,
      "grad_norm": 0.8160937428474426,
      "learning_rate": 5.6526309611002594e-05,
      "loss": 0.5194,
      "step": 240
    },
    {
      "epoch": 0.3282260810350698,
      "grad_norm": 0.7545502185821533,
      "learning_rate": 5.6201722572524275e-05,
      "loss": 0.5173,
      "step": 241
    },
    {
      "epoch": 0.3295880149812734,
      "grad_norm": 0.7594610452651978,
      "learning_rate": 5.587686987289189e-05,
      "loss": 0.47,
      "step": 242
    },
    {
      "epoch": 0.33094994892747703,
      "grad_norm": 0.9410940408706665,
      "learning_rate": 5.5551765427713884e-05,
      "loss": 0.5463,
      "step": 243
    },
    {
      "epoch": 0.33231188287368063,
      "grad_norm": 0.855688750743866,
      "learning_rate": 5.522642316338268e-05,
      "loss": 0.5453,
      "step": 244
    },
    {
      "epoch": 0.33367381681988423,
      "grad_norm": 1.0287182331085205,
      "learning_rate": 5.490085701647805e-05,
      "loss": 0.5741,
      "step": 245
    },
    {
      "epoch": 0.3350357507660878,
      "grad_norm": 0.9338936805725098,
      "learning_rate": 5.457508093317013e-05,
      "loss": 0.4976,
      "step": 246
    },
    {
      "epoch": 0.3363976847122915,
      "grad_norm": 1.0984123945236206,
      "learning_rate": 5.4249108868622086e-05,
      "loss": 0.557,
      "step": 247
    },
    {
      "epoch": 0.3377596186584951,
      "grad_norm": 1.164461374282837,
      "learning_rate": 5.392295478639225e-05,
      "loss": 0.5674,
      "step": 248
    },
    {
      "epoch": 0.3391215526046987,
      "grad_norm": 1.5700137615203857,
      "learning_rate": 5.359663265783598e-05,
      "loss": 0.5427,
      "step": 249
    },
    {
      "epoch": 0.34048348655090227,
      "grad_norm": 2.4326469898223877,
      "learning_rate": 5.327015646150716e-05,
      "loss": 0.6406,
      "step": 250
    },
    {
      "epoch": 0.34184542049710587,
      "grad_norm": 0.5486663579940796,
      "learning_rate": 5.294354018255945e-05,
      "loss": 0.5908,
      "step": 251
    },
    {
      "epoch": 0.3432073544433095,
      "grad_norm": 0.5645151138305664,
      "learning_rate": 5.26167978121472e-05,
      "loss": 0.591,
      "step": 252
    },
    {
      "epoch": 0.3445692883895131,
      "grad_norm": 0.48705315589904785,
      "learning_rate": 5.228994334682604e-05,
      "loss": 0.6097,
      "step": 253
    },
    {
      "epoch": 0.3459312223357167,
      "grad_norm": 0.4560736417770386,
      "learning_rate": 5.196299078795344e-05,
      "loss": 0.5431,
      "step": 254
    },
    {
      "epoch": 0.3472931562819203,
      "grad_norm": 0.5636268258094788,
      "learning_rate": 5.1635954141088813e-05,
      "loss": 0.6741,
      "step": 255
    },
    {
      "epoch": 0.3486550902281239,
      "grad_norm": 0.564383327960968,
      "learning_rate": 5.1308847415393666e-05,
      "loss": 0.6625,
      "step": 256
    },
    {
      "epoch": 0.35001702417432756,
      "grad_norm": 0.7027989029884338,
      "learning_rate": 5.0981684623031415e-05,
      "loss": 0.6753,
      "step": 257
    },
    {
      "epoch": 0.35137895812053116,
      "grad_norm": 0.523399293422699,
      "learning_rate": 5.0654479778567223e-05,
      "loss": 0.5031,
      "step": 258
    },
    {
      "epoch": 0.35274089206673476,
      "grad_norm": 0.6486390233039856,
      "learning_rate": 5.0327246898367597e-05,
      "loss": 0.5439,
      "step": 259
    },
    {
      "epoch": 0.35410282601293835,
      "grad_norm": 0.7013056874275208,
      "learning_rate": 5e-05,
      "loss": 0.5079,
      "step": 260
    },
    {
      "epoch": 0.355464759959142,
      "grad_norm": 0.564133882522583,
      "learning_rate": 4.9672753101632415e-05,
      "loss": 0.5627,
      "step": 261
    },
    {
      "epoch": 0.3568266939053456,
      "grad_norm": 0.6743344068527222,
      "learning_rate": 4.934552022143279e-05,
      "loss": 0.6306,
      "step": 262
    },
    {
      "epoch": 0.3581886278515492,
      "grad_norm": 0.6177630424499512,
      "learning_rate": 4.901831537696859e-05,
      "loss": 0.5655,
      "step": 263
    },
    {
      "epoch": 0.3595505617977528,
      "grad_norm": 0.5786964297294617,
      "learning_rate": 4.869115258460635e-05,
      "loss": 0.5098,
      "step": 264
    },
    {
      "epoch": 0.3609124957439564,
      "grad_norm": 0.5963174104690552,
      "learning_rate": 4.83640458589112e-05,
      "loss": 0.459,
      "step": 265
    },
    {
      "epoch": 0.36227442969016005,
      "grad_norm": 0.6427350640296936,
      "learning_rate": 4.8037009212046586e-05,
      "loss": 0.5881,
      "step": 266
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.7243620157241821,
      "learning_rate": 4.7710056653173976e-05,
      "loss": 0.4404,
      "step": 267
    },
    {
      "epoch": 0.36499829758256724,
      "grad_norm": 0.7247153520584106,
      "learning_rate": 4.738320218785281e-05,
      "loss": 0.5293,
      "step": 268
    },
    {
      "epoch": 0.36636023152877084,
      "grad_norm": 0.7374624609947205,
      "learning_rate": 4.7056459817440544e-05,
      "loss": 0.6044,
      "step": 269
    },
    {
      "epoch": 0.36772216547497444,
      "grad_norm": 0.7089643478393555,
      "learning_rate": 4.6729843538492847e-05,
      "loss": 0.4862,
      "step": 270
    },
    {
      "epoch": 0.3690840994211781,
      "grad_norm": 0.8481512665748596,
      "learning_rate": 4.640336734216403e-05,
      "loss": 0.4846,
      "step": 271
    },
    {
      "epoch": 0.3704460333673817,
      "grad_norm": 0.8320870995521545,
      "learning_rate": 4.607704521360776e-05,
      "loss": 0.5487,
      "step": 272
    },
    {
      "epoch": 0.3718079673135853,
      "grad_norm": 0.8000730872154236,
      "learning_rate": 4.575089113137792e-05,
      "loss": 0.4637,
      "step": 273
    },
    {
      "epoch": 0.3731699012597889,
      "grad_norm": 0.8491706252098083,
      "learning_rate": 4.542491906682989e-05,
      "loss": 0.5178,
      "step": 274
    },
    {
      "epoch": 0.37453183520599254,
      "grad_norm": 0.7658562064170837,
      "learning_rate": 4.509914298352197e-05,
      "loss": 0.5065,
      "step": 275
    },
    {
      "epoch": 0.37589376915219613,
      "grad_norm": 0.7216576933860779,
      "learning_rate": 4.477357683661734e-05,
      "loss": 0.4896,
      "step": 276
    },
    {
      "epoch": 0.37725570309839973,
      "grad_norm": 0.650307297706604,
      "learning_rate": 4.444823457228612e-05,
      "loss": 0.5075,
      "step": 277
    },
    {
      "epoch": 0.37861763704460333,
      "grad_norm": 0.9684988260269165,
      "learning_rate": 4.412313012710813e-05,
      "loss": 0.5661,
      "step": 278
    },
    {
      "epoch": 0.3799795709908069,
      "grad_norm": 0.7701312303543091,
      "learning_rate": 4.379827742747575e-05,
      "loss": 0.4176,
      "step": 279
    },
    {
      "epoch": 0.3813415049370106,
      "grad_norm": 0.7678914070129395,
      "learning_rate": 4.347369038899744e-05,
      "loss": 0.4943,
      "step": 280
    },
    {
      "epoch": 0.3827034388832142,
      "grad_norm": 0.8373364210128784,
      "learning_rate": 4.3149382915901606e-05,
      "loss": 0.5418,
      "step": 281
    },
    {
      "epoch": 0.3840653728294178,
      "grad_norm": 0.7830745577812195,
      "learning_rate": 4.282536890044104e-05,
      "loss": 0.5879,
      "step": 282
    },
    {
      "epoch": 0.38542730677562137,
      "grad_norm": 0.7993201613426208,
      "learning_rate": 4.250166222229774e-05,
      "loss": 0.5371,
      "step": 283
    },
    {
      "epoch": 0.38678924072182497,
      "grad_norm": 0.8836315870285034,
      "learning_rate": 4.2178276747988446e-05,
      "loss": 0.508,
      "step": 284
    },
    {
      "epoch": 0.3881511746680286,
      "grad_norm": 0.8252758383750916,
      "learning_rate": 4.185522633027057e-05,
      "loss": 0.5161,
      "step": 285
    },
    {
      "epoch": 0.3895131086142322,
      "grad_norm": 0.8174393773078918,
      "learning_rate": 4.153252480754877e-05,
      "loss": 0.5386,
      "step": 286
    },
    {
      "epoch": 0.3908750425604358,
      "grad_norm": 0.9204825758934021,
      "learning_rate": 4.1210186003282275e-05,
      "loss": 0.5796,
      "step": 287
    },
    {
      "epoch": 0.3922369765066394,
      "grad_norm": 0.8076188564300537,
      "learning_rate": 4.088822372539263e-05,
      "loss": 0.4966,
      "step": 288
    },
    {
      "epoch": 0.393598910452843,
      "grad_norm": 0.7886457443237305,
      "learning_rate": 4.0566651765672246e-05,
      "loss": 0.4809,
      "step": 289
    },
    {
      "epoch": 0.39496084439904666,
      "grad_norm": 0.7853675484657288,
      "learning_rate": 4.0245483899193595e-05,
      "loss": 0.5844,
      "step": 290
    },
    {
      "epoch": 0.39632277834525026,
      "grad_norm": 0.768347442150116,
      "learning_rate": 3.992473388371915e-05,
      "loss": 0.4446,
      "step": 291
    },
    {
      "epoch": 0.39768471229145386,
      "grad_norm": 0.7979101538658142,
      "learning_rate": 3.960441545911204e-05,
      "loss": 0.5028,
      "step": 292
    },
    {
      "epoch": 0.39904664623765745,
      "grad_norm": 0.9402729868888855,
      "learning_rate": 3.928454234674747e-05,
      "loss": 0.5155,
      "step": 293
    },
    {
      "epoch": 0.4004085801838611,
      "grad_norm": 0.8663700222969055,
      "learning_rate": 3.896512824892495e-05,
      "loss": 0.4983,
      "step": 294
    },
    {
      "epoch": 0.4017705141300647,
      "grad_norm": 0.9468212127685547,
      "learning_rate": 3.864618684828134e-05,
      "loss": 0.5147,
      "step": 295
    },
    {
      "epoch": 0.4031324480762683,
      "grad_norm": 0.9500349164009094,
      "learning_rate": 3.832773180720475e-05,
      "loss": 0.5101,
      "step": 296
    },
    {
      "epoch": 0.4044943820224719,
      "grad_norm": 1.0771191120147705,
      "learning_rate": 3.800977676724919e-05,
      "loss": 0.5714,
      "step": 297
    },
    {
      "epoch": 0.4058563159686755,
      "grad_norm": 1.0121628046035767,
      "learning_rate": 3.769233534855035e-05,
      "loss": 0.5329,
      "step": 298
    },
    {
      "epoch": 0.40721824991487915,
      "grad_norm": 1.3342291116714478,
      "learning_rate": 3.73754211492421e-05,
      "loss": 0.5529,
      "step": 299
    },
    {
      "epoch": 0.40858018386108275,
      "grad_norm": 1.7014594078063965,
      "learning_rate": 3.705904774487396e-05,
      "loss": 0.5537,
      "step": 300
    },
    {
      "epoch": 0.40858018386108275,
      "eval_loss": 0.5519266724586487,
      "eval_runtime": 152.7353,
      "eval_samples_per_second": 8.099,
      "eval_steps_per_second": 2.03,
      "step": 300
    },
    {
      "epoch": 0.40994211780728634,
      "grad_norm": 0.7767552733421326,
      "learning_rate": 3.6743228687829595e-05,
      "loss": 0.5897,
      "step": 301
    },
    {
      "epoch": 0.41130405175348994,
      "grad_norm": 0.5848252773284912,
      "learning_rate": 3.642797750674629e-05,
      "loss": 0.7027,
      "step": 302
    },
    {
      "epoch": 0.41266598569969354,
      "grad_norm": 0.4920623004436493,
      "learning_rate": 3.6113307705935396e-05,
      "loss": 0.5718,
      "step": 303
    },
    {
      "epoch": 0.4140279196458972,
      "grad_norm": 0.5429130792617798,
      "learning_rate": 3.579923276480387e-05,
      "loss": 0.497,
      "step": 304
    },
    {
      "epoch": 0.4153898535921008,
      "grad_norm": 0.520773708820343,
      "learning_rate": 3.5485766137276894e-05,
      "loss": 0.5276,
      "step": 305
    },
    {
      "epoch": 0.4167517875383044,
      "grad_norm": 0.4862958788871765,
      "learning_rate": 3.5172921251221455e-05,
      "loss": 0.67,
      "step": 306
    },
    {
      "epoch": 0.418113721484508,
      "grad_norm": 0.6081766486167908,
      "learning_rate": 3.486071150787128e-05,
      "loss": 0.5882,
      "step": 307
    },
    {
      "epoch": 0.41947565543071164,
      "grad_norm": 0.5209944248199463,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.5373,
      "step": 308
    },
    {
      "epoch": 0.42083758937691523,
      "grad_norm": 0.5533971190452576,
      "learning_rate": 3.423825091761153e-05,
      "loss": 0.5028,
      "step": 309
    },
    {
      "epoch": 0.42219952332311883,
      "grad_norm": 0.5679717659950256,
      "learning_rate": 3.392802673484193e-05,
      "loss": 0.4872,
      "step": 310
    },
    {
      "epoch": 0.42356145726932243,
      "grad_norm": 0.5513074398040771,
      "learning_rate": 3.361849102191533e-05,
      "loss": 0.5398,
      "step": 311
    },
    {
      "epoch": 0.424923391215526,
      "grad_norm": 0.5106960535049438,
      "learning_rate": 3.330965703831146e-05,
      "loss": 0.4908,
      "step": 312
    },
    {
      "epoch": 0.4262853251617297,
      "grad_norm": 0.5206043720245361,
      "learning_rate": 3.300153801345028e-05,
      "loss": 0.5134,
      "step": 313
    },
    {
      "epoch": 0.4276472591079333,
      "grad_norm": 0.6002470254898071,
      "learning_rate": 3.2694147146125345e-05,
      "loss": 0.5067,
      "step": 314
    },
    {
      "epoch": 0.4290091930541369,
      "grad_norm": 0.6005515456199646,
      "learning_rate": 3.2387497603938326e-05,
      "loss": 0.5089,
      "step": 315
    },
    {
      "epoch": 0.43037112700034047,
      "grad_norm": 0.5542832612991333,
      "learning_rate": 3.2081602522734986e-05,
      "loss": 0.4994,
      "step": 316
    },
    {
      "epoch": 0.43173306094654407,
      "grad_norm": 0.6923949718475342,
      "learning_rate": 3.177647500604252e-05,
      "loss": 0.4819,
      "step": 317
    },
    {
      "epoch": 0.4330949948927477,
      "grad_norm": 0.6707453727722168,
      "learning_rate": 3.147212812450819e-05,
      "loss": 0.5279,
      "step": 318
    },
    {
      "epoch": 0.4344569288389513,
      "grad_norm": 0.6384219527244568,
      "learning_rate": 3.116857491533947e-05,
      "loss": 0.4877,
      "step": 319
    },
    {
      "epoch": 0.4358188627851549,
      "grad_norm": 0.7258608937263489,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.5075,
      "step": 320
    },
    {
      "epoch": 0.4371807967313585,
      "grad_norm": 0.710105836391449,
      "learning_rate": 3.056390149238022e-05,
      "loss": 0.4318,
      "step": 321
    },
    {
      "epoch": 0.43854273067756216,
      "grad_norm": 1.025390625,
      "learning_rate": 3.0262807180786647e-05,
      "loss": 0.568,
      "step": 322
    },
    {
      "epoch": 0.43990466462376576,
      "grad_norm": 0.7215895652770996,
      "learning_rate": 2.996255834484296e-05,
      "loss": 0.4904,
      "step": 323
    },
    {
      "epoch": 0.44126659856996936,
      "grad_norm": 0.8992422819137573,
      "learning_rate": 2.9663167846209998e-05,
      "loss": 0.4559,
      "step": 324
    },
    {
      "epoch": 0.44262853251617296,
      "grad_norm": 0.7918512225151062,
      "learning_rate": 2.936464850978027e-05,
      "loss": 0.5122,
      "step": 325
    },
    {
      "epoch": 0.44399046646237655,
      "grad_norm": 0.8224153518676758,
      "learning_rate": 2.9067013123128613e-05,
      "loss": 0.5829,
      "step": 326
    },
    {
      "epoch": 0.4453524004085802,
      "grad_norm": 0.8209482431411743,
      "learning_rate": 2.8770274435964355e-05,
      "loss": 0.5027,
      "step": 327
    },
    {
      "epoch": 0.4467143343547838,
      "grad_norm": 0.8795499205589294,
      "learning_rate": 2.8474445159585235e-05,
      "loss": 0.4533,
      "step": 328
    },
    {
      "epoch": 0.4480762683009874,
      "grad_norm": 0.7316797375679016,
      "learning_rate": 2.8179537966332887e-05,
      "loss": 0.406,
      "step": 329
    },
    {
      "epoch": 0.449438202247191,
      "grad_norm": 0.7172899842262268,
      "learning_rate": 2.7885565489049946e-05,
      "loss": 0.4568,
      "step": 330
    },
    {
      "epoch": 0.4508001361933946,
      "grad_norm": 0.9980983734130859,
      "learning_rate": 2.759254032053888e-05,
      "loss": 0.59,
      "step": 331
    },
    {
      "epoch": 0.45216207013959825,
      "grad_norm": 0.7022810578346252,
      "learning_rate": 2.7300475013022663e-05,
      "loss": 0.4542,
      "step": 332
    },
    {
      "epoch": 0.45352400408580185,
      "grad_norm": 0.7838678359985352,
      "learning_rate": 2.700938207760701e-05,
      "loss": 0.508,
      "step": 333
    },
    {
      "epoch": 0.45488593803200544,
      "grad_norm": 0.835180938243866,
      "learning_rate": 2.671927398374443e-05,
      "loss": 0.4383,
      "step": 334
    },
    {
      "epoch": 0.45624787197820904,
      "grad_norm": 0.6343052983283997,
      "learning_rate": 2.6430163158700115e-05,
      "loss": 0.4295,
      "step": 335
    },
    {
      "epoch": 0.4576098059244127,
      "grad_norm": 0.6634835600852966,
      "learning_rate": 2.6142061987019577e-05,
      "loss": 0.4956,
      "step": 336
    },
    {
      "epoch": 0.4589717398706163,
      "grad_norm": 0.7041996717453003,
      "learning_rate": 2.5854982809998153e-05,
      "loss": 0.4548,
      "step": 337
    },
    {
      "epoch": 0.4603336738168199,
      "grad_norm": 0.8104836940765381,
      "learning_rate": 2.556893792515227e-05,
      "loss": 0.4807,
      "step": 338
    },
    {
      "epoch": 0.4616956077630235,
      "grad_norm": 0.7566388845443726,
      "learning_rate": 2.5283939585692783e-05,
      "loss": 0.4644,
      "step": 339
    },
    {
      "epoch": 0.4630575417092271,
      "grad_norm": 0.7137734293937683,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.3881,
      "step": 340
    },
    {
      "epoch": 0.46441947565543074,
      "grad_norm": 0.785618782043457,
      "learning_rate": 2.471713133110078e-05,
      "loss": 0.4656,
      "step": 341
    },
    {
      "epoch": 0.46578140960163433,
      "grad_norm": 0.891506016254425,
      "learning_rate": 2.4435345696147403e-05,
      "loss": 0.5149,
      "step": 342
    },
    {
      "epoch": 0.46714334354783793,
      "grad_norm": 1.0425530672073364,
      "learning_rate": 2.4154655165898627e-05,
      "loss": 0.4701,
      "step": 343
    },
    {
      "epoch": 0.4685052774940415,
      "grad_norm": 0.8959038853645325,
      "learning_rate": 2.3875071764202563e-05,
      "loss": 0.5629,
      "step": 344
    },
    {
      "epoch": 0.4698672114402451,
      "grad_norm": 0.8744730353355408,
      "learning_rate": 2.3596607467481603e-05,
      "loss": 0.4825,
      "step": 345
    },
    {
      "epoch": 0.4712291453864488,
      "grad_norm": 0.9143437147140503,
      "learning_rate": 2.3319274204219428e-05,
      "loss": 0.506,
      "step": 346
    },
    {
      "epoch": 0.4725910793326524,
      "grad_norm": 0.9251874685287476,
      "learning_rate": 2.3043083854449988e-05,
      "loss": 0.4871,
      "step": 347
    },
    {
      "epoch": 0.473953013278856,
      "grad_norm": 1.0879664421081543,
      "learning_rate": 2.2768048249248648e-05,
      "loss": 0.5618,
      "step": 348
    },
    {
      "epoch": 0.47531494722505957,
      "grad_norm": 1.1853846311569214,
      "learning_rate": 2.2494179170225333e-05,
      "loss": 0.5121,
      "step": 349
    },
    {
      "epoch": 0.47667688117126317,
      "grad_norm": 1.999249815940857,
      "learning_rate": 2.2221488349019903e-05,
      "loss": 0.7677,
      "step": 350
    },
    {
      "epoch": 0.4780388151174668,
      "grad_norm": 0.4879932403564453,
      "learning_rate": 2.194998746679952e-05,
      "loss": 0.557,
      "step": 351
    },
    {
      "epoch": 0.4794007490636704,
      "grad_norm": 0.5313215851783752,
      "learning_rate": 2.167968815375837e-05,
      "loss": 0.6362,
      "step": 352
    },
    {
      "epoch": 0.480762683009874,
      "grad_norm": 0.4870143532752991,
      "learning_rate": 2.1410601988619394e-05,
      "loss": 0.5461,
      "step": 353
    },
    {
      "epoch": 0.4821246169560776,
      "grad_norm": 0.40748482942581177,
      "learning_rate": 2.1142740498138324e-05,
      "loss": 0.5394,
      "step": 354
    },
    {
      "epoch": 0.48348655090228126,
      "grad_norm": 0.5128802061080933,
      "learning_rate": 2.08761151566099e-05,
      "loss": 0.5926,
      "step": 355
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 0.45034193992614746,
      "learning_rate": 2.061073738537635e-05,
      "loss": 0.5749,
      "step": 356
    },
    {
      "epoch": 0.48621041879468846,
      "grad_norm": 0.7784144878387451,
      "learning_rate": 2.034661855233815e-05,
      "loss": 0.4702,
      "step": 357
    },
    {
      "epoch": 0.48757235274089206,
      "grad_norm": 0.48969149589538574,
      "learning_rate": 2.008376997146705e-05,
      "loss": 0.5136,
      "step": 358
    },
    {
      "epoch": 0.48893428668709565,
      "grad_norm": 0.475238174200058,
      "learning_rate": 1.982220290232143e-05,
      "loss": 0.5247,
      "step": 359
    },
    {
      "epoch": 0.4902962206332993,
      "grad_norm": 0.4993893802165985,
      "learning_rate": 1.9561928549563968e-05,
      "loss": 0.4939,
      "step": 360
    },
    {
      "epoch": 0.4916581545795029,
      "grad_norm": 0.5764222741127014,
      "learning_rate": 1.9302958062481673e-05,
      "loss": 0.6013,
      "step": 361
    },
    {
      "epoch": 0.4930200885257065,
      "grad_norm": 0.4824737310409546,
      "learning_rate": 1.9045302534508297e-05,
      "loss": 0.4703,
      "step": 362
    },
    {
      "epoch": 0.4943820224719101,
      "grad_norm": 0.6622003316879272,
      "learning_rate": 1.8788973002749112e-05,
      "loss": 0.5234,
      "step": 363
    },
    {
      "epoch": 0.4957439564181137,
      "grad_norm": 0.6316631436347961,
      "learning_rate": 1.8533980447508137e-05,
      "loss": 0.4794,
      "step": 364
    },
    {
      "epoch": 0.49710589036431735,
      "grad_norm": 0.5674468874931335,
      "learning_rate": 1.8280335791817733e-05,
      "loss": 0.5004,
      "step": 365
    },
    {
      "epoch": 0.49846782431052095,
      "grad_norm": 0.5479894876480103,
      "learning_rate": 1.8028049900970767e-05,
      "loss": 0.4401,
      "step": 366
    },
    {
      "epoch": 0.49982975825672454,
      "grad_norm": 0.6312588453292847,
      "learning_rate": 1.777713358205514e-05,
      "loss": 0.5078,
      "step": 367
    },
    {
      "epoch": 0.5011916922029281,
      "grad_norm": 0.6721518635749817,
      "learning_rate": 1.7527597583490822e-05,
      "loss": 0.4853,
      "step": 368
    },
    {
      "epoch": 0.5025536261491318,
      "grad_norm": 0.7594829201698303,
      "learning_rate": 1.7279452594569483e-05,
      "loss": 0.5361,
      "step": 369
    },
    {
      "epoch": 0.5039155600953353,
      "grad_norm": 0.6352505683898926,
      "learning_rate": 1.703270924499656e-05,
      "loss": 0.4291,
      "step": 370
    },
    {
      "epoch": 0.505277494041539,
      "grad_norm": 0.6971171498298645,
      "learning_rate": 1.678737810443593e-05,
      "loss": 0.613,
      "step": 371
    },
    {
      "epoch": 0.5066394279877426,
      "grad_norm": 0.8621500730514526,
      "learning_rate": 1.6543469682057106e-05,
      "loss": 0.5023,
      "step": 372
    },
    {
      "epoch": 0.5080013619339462,
      "grad_norm": 0.7477943897247314,
      "learning_rate": 1.6300994426085103e-05,
      "loss": 0.4928,
      "step": 373
    },
    {
      "epoch": 0.5093632958801498,
      "grad_norm": 0.6949486136436462,
      "learning_rate": 1.605996272335291e-05,
      "loss": 0.4106,
      "step": 374
    },
    {
      "epoch": 0.5107252298263534,
      "grad_norm": 0.7850691080093384,
      "learning_rate": 1.5820384898856434e-05,
      "loss": 0.4416,
      "step": 375
    },
    {
      "epoch": 0.512087163772557,
      "grad_norm": 1.0420745611190796,
      "learning_rate": 1.5582271215312294e-05,
      "loss": 0.5823,
      "step": 376
    },
    {
      "epoch": 0.5134490977187607,
      "grad_norm": 0.8832230567932129,
      "learning_rate": 1.5345631872718214e-05,
      "loss": 0.511,
      "step": 377
    },
    {
      "epoch": 0.5148110316649642,
      "grad_norm": 0.8311038613319397,
      "learning_rate": 1.5110477007916001e-05,
      "loss": 0.4482,
      "step": 378
    },
    {
      "epoch": 0.5161729656111679,
      "grad_norm": 0.7399735450744629,
      "learning_rate": 1.4876816694157419e-05,
      "loss": 0.4744,
      "step": 379
    },
    {
      "epoch": 0.5175348995573714,
      "grad_norm": 0.8263036608695984,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.5225,
      "step": 380
    },
    {
      "epoch": 0.5188968335035751,
      "grad_norm": 0.7773760557174683,
      "learning_rate": 1.4414019692241437e-05,
      "loss": 0.5006,
      "step": 381
    },
    {
      "epoch": 0.5202587674497787,
      "grad_norm": 0.7417269349098206,
      "learning_rate": 1.4184902828767287e-05,
      "loss": 0.4494,
      "step": 382
    },
    {
      "epoch": 0.5216207013959823,
      "grad_norm": 0.7255116105079651,
      "learning_rate": 1.3957320164854059e-05,
      "loss": 0.4674,
      "step": 383
    },
    {
      "epoch": 0.5229826353421859,
      "grad_norm": 0.8162254691123962,
      "learning_rate": 1.373128144938563e-05,
      "loss": 0.4882,
      "step": 384
    },
    {
      "epoch": 0.5243445692883895,
      "grad_norm": 0.8375689387321472,
      "learning_rate": 1.3506796365108232e-05,
      "loss": 0.5161,
      "step": 385
    },
    {
      "epoch": 0.5257065032345931,
      "grad_norm": 0.806830108165741,
      "learning_rate": 1.3283874528215733e-05,
      "loss": 0.4965,
      "step": 386
    },
    {
      "epoch": 0.5270684371807968,
      "grad_norm": 0.7288891673088074,
      "learning_rate": 1.3062525487937699e-05,
      "loss": 0.5127,
      "step": 387
    },
    {
      "epoch": 0.5284303711270003,
      "grad_norm": 0.7512708902359009,
      "learning_rate": 1.2842758726130283e-05,
      "loss": 0.549,
      "step": 388
    },
    {
      "epoch": 0.529792305073204,
      "grad_norm": 0.7276395559310913,
      "learning_rate": 1.2624583656870154e-05,
      "loss": 0.4332,
      "step": 389
    },
    {
      "epoch": 0.5311542390194075,
      "grad_norm": 0.7371705770492554,
      "learning_rate": 1.2408009626051137e-05,
      "loss": 0.4727,
      "step": 390
    },
    {
      "epoch": 0.5325161729656112,
      "grad_norm": 0.8400708436965942,
      "learning_rate": 1.2193045910983863e-05,
      "loss": 0.5157,
      "step": 391
    },
    {
      "epoch": 0.5338781069118148,
      "grad_norm": 0.7981722354888916,
      "learning_rate": 1.1979701719998453e-05,
      "loss": 0.4702,
      "step": 392
    },
    {
      "epoch": 0.5352400408580184,
      "grad_norm": 0.7362338900566101,
      "learning_rate": 1.1767986192049984e-05,
      "loss": 0.4614,
      "step": 393
    },
    {
      "epoch": 0.536601974804222,
      "grad_norm": 0.8625655174255371,
      "learning_rate": 1.1557908396327028e-05,
      "loss": 0.4862,
      "step": 394
    },
    {
      "epoch": 0.5379639087504257,
      "grad_norm": 0.9110238552093506,
      "learning_rate": 1.134947733186315e-05,
      "loss": 0.5014,
      "step": 395
    },
    {
      "epoch": 0.5393258426966292,
      "grad_norm": 1.0649625062942505,
      "learning_rate": 1.1142701927151456e-05,
      "loss": 0.5312,
      "step": 396
    },
    {
      "epoch": 0.5406877766428329,
      "grad_norm": 0.9835678935050964,
      "learning_rate": 1.0937591039762085e-05,
      "loss": 0.4989,
      "step": 397
    },
    {
      "epoch": 0.5420497105890364,
      "grad_norm": 1.0416983366012573,
      "learning_rate": 1.0734153455962765e-05,
      "loss": 0.5524,
      "step": 398
    },
    {
      "epoch": 0.54341164453524,
      "grad_norm": 1.4598973989486694,
      "learning_rate": 1.0532397890342505e-05,
      "loss": 0.6047,
      "step": 399
    },
    {
      "epoch": 0.5447735784814437,
      "grad_norm": 2.293917417526245,
      "learning_rate": 1.0332332985438248e-05,
      "loss": 0.7414,
      "step": 400
    },
    {
      "epoch": 0.5447735784814437,
      "eval_loss": 0.5299856066703796,
      "eval_runtime": 152.8199,
      "eval_samples_per_second": 8.094,
      "eval_steps_per_second": 2.029,
      "step": 400
    },
    {
      "epoch": 0.5461355124276472,
      "grad_norm": 0.54931640625,
      "learning_rate": 1.013396731136465e-05,
      "loss": 0.5829,
      "step": 401
    },
    {
      "epoch": 0.5474974463738509,
      "grad_norm": 0.521465539932251,
      "learning_rate": 9.937309365446973e-06,
      "loss": 0.6688,
      "step": 402
    },
    {
      "epoch": 0.5488593803200544,
      "grad_norm": 0.5131102800369263,
      "learning_rate": 9.742367571857091e-06,
      "loss": 0.581,
      "step": 403
    },
    {
      "epoch": 0.5502213142662581,
      "grad_norm": 0.44577908515930176,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.4947,
      "step": 404
    },
    {
      "epoch": 0.5515832482124617,
      "grad_norm": 0.5106862187385559,
      "learning_rate": 9.357665770419244e-06,
      "loss": 0.5882,
      "step": 405
    },
    {
      "epoch": 0.5529451821586653,
      "grad_norm": 0.5089471936225891,
      "learning_rate": 9.167922241916055e-06,
      "loss": 0.5681,
      "step": 406
    },
    {
      "epoch": 0.5543071161048689,
      "grad_norm": 1.0630840063095093,
      "learning_rate": 8.97992782372432e-06,
      "loss": 0.6777,
      "step": 407
    },
    {
      "epoch": 0.5556690500510725,
      "grad_norm": 0.5694750547409058,
      "learning_rate": 8.793690568899216e-06,
      "loss": 0.5003,
      "step": 408
    },
    {
      "epoch": 0.5570309839972761,
      "grad_norm": 0.5045074224472046,
      "learning_rate": 8.609218455224893e-06,
      "loss": 0.5008,
      "step": 409
    },
    {
      "epoch": 0.5583929179434798,
      "grad_norm": 0.5081794261932373,
      "learning_rate": 8.426519384872733e-06,
      "loss": 0.5411,
      "step": 410
    },
    {
      "epoch": 0.5597548518896833,
      "grad_norm": 0.6146826148033142,
      "learning_rate": 8.245601184062852e-06,
      "loss": 0.4463,
      "step": 411
    },
    {
      "epoch": 0.561116785835887,
      "grad_norm": 0.5130215287208557,
      "learning_rate": 8.066471602728803e-06,
      "loss": 0.4872,
      "step": 412
    },
    {
      "epoch": 0.5624787197820905,
      "grad_norm": 0.7060839533805847,
      "learning_rate": 7.889138314185678e-06,
      "loss": 0.5111,
      "step": 413
    },
    {
      "epoch": 0.5638406537282942,
      "grad_norm": 0.5828076601028442,
      "learning_rate": 7.71360891480134e-06,
      "loss": 0.4822,
      "step": 414
    },
    {
      "epoch": 0.5652025876744978,
      "grad_norm": 0.7213607430458069,
      "learning_rate": 7.539890923671062e-06,
      "loss": 0.5358,
      "step": 415
    },
    {
      "epoch": 0.5665645216207014,
      "grad_norm": 0.6452810168266296,
      "learning_rate": 7.367991782295391e-06,
      "loss": 0.5447,
      "step": 416
    },
    {
      "epoch": 0.567926455566905,
      "grad_norm": 0.7393741607666016,
      "learning_rate": 7.197918854261432e-06,
      "loss": 0.5965,
      "step": 417
    },
    {
      "epoch": 0.5692883895131086,
      "grad_norm": 0.6635223031044006,
      "learning_rate": 7.029679424927365e-06,
      "loss": 0.4964,
      "step": 418
    },
    {
      "epoch": 0.5706503234593122,
      "grad_norm": 0.6983364820480347,
      "learning_rate": 6.863280701110408e-06,
      "loss": 0.3943,
      "step": 419
    },
    {
      "epoch": 0.5720122574055159,
      "grad_norm": 0.5849578380584717,
      "learning_rate": 6.698729810778065e-06,
      "loss": 0.4307,
      "step": 420
    },
    {
      "epoch": 0.5733741913517194,
      "grad_norm": 0.703769326210022,
      "learning_rate": 6.536033802742813e-06,
      "loss": 0.5444,
      "step": 421
    },
    {
      "epoch": 0.5747361252979231,
      "grad_norm": 0.7360491752624512,
      "learning_rate": 6.375199646360142e-06,
      "loss": 0.4167,
      "step": 422
    },
    {
      "epoch": 0.5760980592441267,
      "grad_norm": 0.8468068838119507,
      "learning_rate": 6.216234231230012e-06,
      "loss": 0.5626,
      "step": 423
    },
    {
      "epoch": 0.5774599931903303,
      "grad_norm": 0.8693845272064209,
      "learning_rate": 6.059144366901736e-06,
      "loss": 0.5174,
      "step": 424
    },
    {
      "epoch": 0.5788219271365339,
      "grad_norm": 0.8013517260551453,
      "learning_rate": 5.903936782582253e-06,
      "loss": 0.5139,
      "step": 425
    },
    {
      "epoch": 0.5801838610827375,
      "grad_norm": 0.9400525093078613,
      "learning_rate": 5.750618126847912e-06,
      "loss": 0.5828,
      "step": 426
    },
    {
      "epoch": 0.5815457950289411,
      "grad_norm": 0.9872230887413025,
      "learning_rate": 5.599194967359639e-06,
      "loss": 0.4136,
      "step": 427
    },
    {
      "epoch": 0.5829077289751448,
      "grad_norm": 0.7285526990890503,
      "learning_rate": 5.449673790581611e-06,
      "loss": 0.4517,
      "step": 428
    },
    {
      "epoch": 0.5842696629213483,
      "grad_norm": 1.011879801750183,
      "learning_rate": 5.302061001503394e-06,
      "loss": 0.5392,
      "step": 429
    },
    {
      "epoch": 0.585631596867552,
      "grad_norm": 0.6948602199554443,
      "learning_rate": 5.156362923365588e-06,
      "loss": 0.4626,
      "step": 430
    },
    {
      "epoch": 0.5869935308137555,
      "grad_norm": 0.700434684753418,
      "learning_rate": 5.012585797388936e-06,
      "loss": 0.4383,
      "step": 431
    },
    {
      "epoch": 0.5883554647599591,
      "grad_norm": 0.7421513795852661,
      "learning_rate": 4.87073578250698e-06,
      "loss": 0.4413,
      "step": 432
    },
    {
      "epoch": 0.5897173987061628,
      "grad_norm": 0.8031982183456421,
      "learning_rate": 4.730818955102234e-06,
      "loss": 0.4548,
      "step": 433
    },
    {
      "epoch": 0.5910793326523663,
      "grad_norm": 0.6696775555610657,
      "learning_rate": 4.592841308745932e-06,
      "loss": 0.4283,
      "step": 434
    },
    {
      "epoch": 0.59244126659857,
      "grad_norm": 0.6684313416481018,
      "learning_rate": 4.456808753941205e-06,
      "loss": 0.4176,
      "step": 435
    },
    {
      "epoch": 0.5938032005447735,
      "grad_norm": 0.78340744972229,
      "learning_rate": 4.322727117869951e-06,
      "loss": 0.4855,
      "step": 436
    },
    {
      "epoch": 0.5951651344909772,
      "grad_norm": 0.8859238028526306,
      "learning_rate": 4.190602144143207e-06,
      "loss": 0.4494,
      "step": 437
    },
    {
      "epoch": 0.5965270684371808,
      "grad_norm": 0.7926759123802185,
      "learning_rate": 4.06043949255509e-06,
      "loss": 0.4473,
      "step": 438
    },
    {
      "epoch": 0.5978890023833844,
      "grad_norm": 0.7094496488571167,
      "learning_rate": 3.932244738840379e-06,
      "loss": 0.4022,
      "step": 439
    },
    {
      "epoch": 0.599250936329588,
      "grad_norm": 0.7825495004653931,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.4681,
      "step": 440
    },
    {
      "epoch": 0.6006128702757916,
      "grad_norm": 0.8244885802268982,
      "learning_rate": 3.681780806244095e-06,
      "loss": 0.5154,
      "step": 441
    },
    {
      "epoch": 0.6019748042219952,
      "grad_norm": 0.8276898264884949,
      "learning_rate": 3.5595223564037884e-06,
      "loss": 0.487,
      "step": 442
    },
    {
      "epoch": 0.6033367381681989,
      "grad_norm": 0.9214934706687927,
      "learning_rate": 3.4392532620598216e-06,
      "loss": 0.5613,
      "step": 443
    },
    {
      "epoch": 0.6046986721144024,
      "grad_norm": 0.9555485844612122,
      "learning_rate": 3.3209786751399187e-06,
      "loss": 0.5597,
      "step": 444
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 0.8737406134605408,
      "learning_rate": 3.2047036621337236e-06,
      "loss": 0.488,
      "step": 445
    },
    {
      "epoch": 0.6074225400068096,
      "grad_norm": 1.145499587059021,
      "learning_rate": 3.0904332038757977e-06,
      "loss": 0.4535,
      "step": 446
    },
    {
      "epoch": 0.6087844739530133,
      "grad_norm": 1.149397611618042,
      "learning_rate": 2.978172195332263e-06,
      "loss": 0.4518,
      "step": 447
    },
    {
      "epoch": 0.6101464078992169,
      "grad_norm": 1.3012440204620361,
      "learning_rate": 2.8679254453910785e-06,
      "loss": 0.5319,
      "step": 448
    },
    {
      "epoch": 0.6115083418454205,
      "grad_norm": 1.1406331062316895,
      "learning_rate": 2.759697676656098e-06,
      "loss": 0.4901,
      "step": 449
    },
    {
      "epoch": 0.6128702757916241,
      "grad_norm": 2.0314929485321045,
      "learning_rate": 2.653493525244721e-06,
      "loss": 0.6488,
      "step": 450
    },
    {
      "epoch": 0.6142322097378277,
      "grad_norm": 0.4973229765892029,
      "learning_rate": 2.549317540589308e-06,
      "loss": 0.6044,
      "step": 451
    },
    {
      "epoch": 0.6155941436840313,
      "grad_norm": 0.4732745289802551,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 0.5759,
      "step": 452
    },
    {
      "epoch": 0.616956077630235,
      "grad_norm": 0.4533315598964691,
      "learning_rate": 2.3470678346851518e-06,
      "loss": 0.5511,
      "step": 453
    },
    {
      "epoch": 0.6183180115764385,
      "grad_norm": 0.40576285123825073,
      "learning_rate": 2.2490027771406687e-06,
      "loss": 0.5447,
      "step": 454
    },
    {
      "epoch": 0.6196799455226422,
      "grad_norm": 0.48012784123420715,
      "learning_rate": 2.152983213389559e-06,
      "loss": 0.6735,
      "step": 455
    },
    {
      "epoch": 0.6210418794688458,
      "grad_norm": 0.5396324396133423,
      "learning_rate": 2.0590132565903476e-06,
      "loss": 0.6272,
      "step": 456
    },
    {
      "epoch": 0.6224038134150494,
      "grad_norm": 0.5428497791290283,
      "learning_rate": 1.9670969321032407e-06,
      "loss": 0.5316,
      "step": 457
    },
    {
      "epoch": 0.623765747361253,
      "grad_norm": 0.5881633162498474,
      "learning_rate": 1.8772381773176417e-06,
      "loss": 0.4856,
      "step": 458
    },
    {
      "epoch": 0.6251276813074566,
      "grad_norm": 0.5558289289474487,
      "learning_rate": 1.7894408414835362e-06,
      "loss": 0.5262,
      "step": 459
    },
    {
      "epoch": 0.6264896152536602,
      "grad_norm": 0.5173474550247192,
      "learning_rate": 1.70370868554659e-06,
      "loss": 0.5073,
      "step": 460
    },
    {
      "epoch": 0.6278515491998639,
      "grad_norm": 0.5354324579238892,
      "learning_rate": 1.620045381987012e-06,
      "loss": 0.4596,
      "step": 461
    },
    {
      "epoch": 0.6292134831460674,
      "grad_norm": 0.57889723777771,
      "learning_rate": 1.5384545146622852e-06,
      "loss": 0.5365,
      "step": 462
    },
    {
      "epoch": 0.630575417092271,
      "grad_norm": 0.645720899105072,
      "learning_rate": 1.4589395786535953e-06,
      "loss": 0.4665,
      "step": 463
    },
    {
      "epoch": 0.6319373510384746,
      "grad_norm": 0.7268922924995422,
      "learning_rate": 1.3815039801161721e-06,
      "loss": 0.5446,
      "step": 464
    },
    {
      "epoch": 0.6332992849846782,
      "grad_norm": 0.5946515798568726,
      "learning_rate": 1.3061510361333185e-06,
      "loss": 0.505,
      "step": 465
    },
    {
      "epoch": 0.6346612189308819,
      "grad_norm": 0.7125037312507629,
      "learning_rate": 1.232883974574367e-06,
      "loss": 0.5175,
      "step": 466
    },
    {
      "epoch": 0.6360231528770854,
      "grad_norm": 0.9209554195404053,
      "learning_rate": 1.1617059339563807e-06,
      "loss": 0.5226,
      "step": 467
    },
    {
      "epoch": 0.6373850868232891,
      "grad_norm": 0.740816056728363,
      "learning_rate": 1.0926199633097157e-06,
      "loss": 0.4817,
      "step": 468
    },
    {
      "epoch": 0.6387470207694926,
      "grad_norm": 0.70599365234375,
      "learning_rate": 1.0256290220474307e-06,
      "loss": 0.5063,
      "step": 469
    },
    {
      "epoch": 0.6401089547156963,
      "grad_norm": 0.7017062902450562,
      "learning_rate": 9.607359798384785e-07,
      "loss": 0.4441,
      "step": 470
    },
    {
      "epoch": 0.6414708886618999,
      "grad_norm": 0.8116502165794373,
      "learning_rate": 8.979436164848088e-07,
      "loss": 0.5034,
      "step": 471
    },
    {
      "epoch": 0.6428328226081035,
      "grad_norm": 0.7812343835830688,
      "learning_rate": 8.372546218022747e-07,
      "loss": 0.4646,
      "step": 472
    },
    {
      "epoch": 0.6441947565543071,
      "grad_norm": 0.813950777053833,
      "learning_rate": 7.786715955054203e-07,
      "loss": 0.4721,
      "step": 473
    },
    {
      "epoch": 0.6455566905005107,
      "grad_norm": 0.8855940699577332,
      "learning_rate": 7.221970470961125e-07,
      "loss": 0.4147,
      "step": 474
    },
    {
      "epoch": 0.6469186244467143,
      "grad_norm": 0.8395663499832153,
      "learning_rate": 6.678333957560512e-07,
      "loss": 0.4648,
      "step": 475
    },
    {
      "epoch": 0.648280558392918,
      "grad_norm": 0.6990065574645996,
      "learning_rate": 6.15582970243117e-07,
      "loss": 0.46,
      "step": 476
    },
    {
      "epoch": 0.6496424923391215,
      "grad_norm": 0.7181410789489746,
      "learning_rate": 5.654480087916303e-07,
      "loss": 0.4139,
      "step": 477
    },
    {
      "epoch": 0.6510044262853252,
      "grad_norm": 0.8109433054924011,
      "learning_rate": 5.174306590164879e-07,
      "loss": 0.4678,
      "step": 478
    },
    {
      "epoch": 0.6523663602315287,
      "grad_norm": 0.7934712171554565,
      "learning_rate": 4.715329778211375e-07,
      "loss": 0.4834,
      "step": 479
    },
    {
      "epoch": 0.6537282941777324,
      "grad_norm": 0.7957246899604797,
      "learning_rate": 4.277569313094809e-07,
      "loss": 0.439,
      "step": 480
    },
    {
      "epoch": 0.655090228123936,
      "grad_norm": 0.745269238948822,
      "learning_rate": 3.8610439470164737e-07,
      "loss": 0.4721,
      "step": 481
    },
    {
      "epoch": 0.6564521620701396,
      "grad_norm": 0.7737083435058594,
      "learning_rate": 3.465771522536854e-07,
      "loss": 0.4393,
      "step": 482
    },
    {
      "epoch": 0.6578140960163432,
      "grad_norm": 0.810192346572876,
      "learning_rate": 3.09176897181096e-07,
      "loss": 0.4851,
      "step": 483
    },
    {
      "epoch": 0.6591760299625468,
      "grad_norm": 0.7566391229629517,
      "learning_rate": 2.7390523158633554e-07,
      "loss": 0.4779,
      "step": 484
    },
    {
      "epoch": 0.6605379639087504,
      "grad_norm": 0.7294163107872009,
      "learning_rate": 2.407636663901591e-07,
      "loss": 0.4335,
      "step": 485
    },
    {
      "epoch": 0.6618998978549541,
      "grad_norm": 0.8193715810775757,
      "learning_rate": 2.0975362126691712e-07,
      "loss": 0.4461,
      "step": 486
    },
    {
      "epoch": 0.6632618318011576,
      "grad_norm": 0.7951440811157227,
      "learning_rate": 1.8087642458373134e-07,
      "loss": 0.4687,
      "step": 487
    },
    {
      "epoch": 0.6646237657473613,
      "grad_norm": 0.72347491979599,
      "learning_rate": 1.5413331334360182e-07,
      "loss": 0.4906,
      "step": 488
    },
    {
      "epoch": 0.6659856996935649,
      "grad_norm": 0.7901465892791748,
      "learning_rate": 1.2952543313240472e-07,
      "loss": 0.474,
      "step": 489
    },
    {
      "epoch": 0.6673476336397685,
      "grad_norm": 0.852858304977417,
      "learning_rate": 1.0705383806982606e-07,
      "loss": 0.5185,
      "step": 490
    },
    {
      "epoch": 0.6687095675859721,
      "grad_norm": 0.7795984148979187,
      "learning_rate": 8.671949076420882e-08,
      "loss": 0.4433,
      "step": 491
    },
    {
      "epoch": 0.6700715015321757,
      "grad_norm": 0.9037585854530334,
      "learning_rate": 6.852326227130834e-08,
      "loss": 0.5214,
      "step": 492
    },
    {
      "epoch": 0.6714334354783793,
      "grad_norm": 0.7909391522407532,
      "learning_rate": 5.246593205699424e-08,
      "loss": 0.5078,
      "step": 493
    },
    {
      "epoch": 0.672795369424583,
      "grad_norm": 0.8686991333961487,
      "learning_rate": 3.8548187963854956e-08,
      "loss": 0.4828,
      "step": 494
    },
    {
      "epoch": 0.6741573033707865,
      "grad_norm": 1.1080681085586548,
      "learning_rate": 2.6770626181715773e-08,
      "loss": 0.465,
      "step": 495
    },
    {
      "epoch": 0.6755192373169902,
      "grad_norm": 0.9239915609359741,
      "learning_rate": 1.7133751222137007e-08,
      "loss": 0.4729,
      "step": 496
    },
    {
      "epoch": 0.6768811712631937,
      "grad_norm": 1.044656753540039,
      "learning_rate": 9.637975896759077e-09,
      "loss": 0.4499,
      "step": 497
    },
    {
      "epoch": 0.6782431052093973,
      "grad_norm": 1.1255676746368408,
      "learning_rate": 4.2836212996499865e-09,
      "loss": 0.55,
      "step": 498
    },
    {
      "epoch": 0.679605039155601,
      "grad_norm": 1.5408134460449219,
      "learning_rate": 1.0709167935385455e-09,
      "loss": 0.6591,
      "step": 499
    },
    {
      "epoch": 0.6809669731018045,
      "grad_norm": 2.2999343872070312,
      "learning_rate": 0.0,
      "loss": 0.6385,
      "step": 500
    },
    {
      "epoch": 0.6809669731018045,
      "eval_loss": 0.5032929182052612,
      "eval_runtime": 152.8674,
      "eval_samples_per_second": 8.092,
      "eval_steps_per_second": 2.028,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2097578389691433e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}