{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 8784,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001138433515482696,
      "grad_norm": 1.1881835750538772,
      "learning_rate": 9.999994883476085e-06,
      "loss": 0.6587,
      "num_input_tokens_seen": 267128,
      "step": 5,
      "train_runtime": 32.1347,
      "train_tokens_per_second": 8312.746
    },
    {
      "epoch": 0.002276867030965392,
      "grad_norm": 1.0791006215860541,
      "learning_rate": 9.999974097615626e-06,
      "loss": 0.5363,
      "num_input_tokens_seen": 560400,
      "step": 10,
      "train_runtime": 63.7833,
      "train_tokens_per_second": 8785.992
    },
    {
      "epoch": 0.0034153005464480873,
      "grad_norm": 0.9611395244520237,
      "learning_rate": 9.999937322702294e-06,
      "loss": 0.4854,
      "num_input_tokens_seen": 826344,
      "step": 15,
      "train_runtime": 91.4799,
      "train_tokens_per_second": 9033.063
    },
    {
      "epoch": 0.004553734061930784,
      "grad_norm": 0.8607458659498174,
      "learning_rate": 9.999884558853694e-06,
      "loss": 0.4767,
      "num_input_tokens_seen": 1100312,
      "step": 20,
      "train_runtime": 116.3433,
      "train_tokens_per_second": 9457.458
    },
    {
      "epoch": 0.005692167577413479,
      "grad_norm": 0.9418329072988826,
      "learning_rate": 9.999815806238551e-06,
      "loss": 0.4505,
      "num_input_tokens_seen": 1384352,
      "step": 25,
      "train_runtime": 147.6706,
      "train_tokens_per_second": 9374.595
    },
    {
      "epoch": 0.006830601092896175,
      "grad_norm": 0.7628317043484723,
      "learning_rate": 9.999731065076725e-06,
      "loss": 0.4824,
      "num_input_tokens_seen": 1657408,
      "step": 30,
      "train_runtime": 180.474,
      "train_tokens_per_second": 9183.639
    },
    {
      "epoch": 0.007969034608378871,
      "grad_norm": 0.9945811869626061,
      "learning_rate": 9.999630335639206e-06,
      "loss": 0.4531,
      "num_input_tokens_seen": 1919360,
      "step": 35,
      "train_runtime": 205.1009,
      "train_tokens_per_second": 9358.127
    },
    {
      "epoch": 0.009107468123861567,
      "grad_norm": 0.968029301183335,
      "learning_rate": 9.999513618248107e-06,
      "loss": 0.449,
      "num_input_tokens_seen": 2199968,
      "step": 40,
      "train_runtime": 231.5493,
      "train_tokens_per_second": 9501.077
    },
    {
      "epoch": 0.010245901639344262,
      "grad_norm": 0.7863928869177746,
      "learning_rate": 9.99938091327667e-06,
      "loss": 0.4137,
      "num_input_tokens_seen": 2502600,
      "step": 45,
      "train_runtime": 265.022,
      "train_tokens_per_second": 9442.99
    },
    {
      "epoch": 0.011384335154826957,
      "grad_norm": 0.8288799664376203,
      "learning_rate": 9.999232221149264e-06,
      "loss": 0.4142,
      "num_input_tokens_seen": 2811872,
      "step": 50,
      "train_runtime": 292.1107,
      "train_tokens_per_second": 9626.048
    },
    {
      "epoch": 0.012522768670309653,
      "grad_norm": 0.7643852831082628,
      "learning_rate": 9.99906754234138e-06,
      "loss": 0.4172,
      "num_input_tokens_seen": 3113112,
      "step": 55,
      "train_runtime": 315.298,
      "train_tokens_per_second": 9873.556
    },
    {
      "epoch": 0.01366120218579235,
      "grad_norm": 0.7565965791337023,
      "learning_rate": 9.998886877379632e-06,
      "loss": 0.4369,
      "num_input_tokens_seen": 3398944,
      "step": 60,
      "train_runtime": 337.9627,
      "train_tokens_per_second": 10057.157
    },
    {
      "epoch": 0.014799635701275045,
      "grad_norm": 0.9947004195282368,
      "learning_rate": 9.998690226841756e-06,
      "loss": 0.3942,
      "num_input_tokens_seen": 3680072,
      "step": 65,
      "train_runtime": 369.3733,
      "train_tokens_per_second": 9963.017
    },
    {
      "epoch": 0.015938069216757743,
      "grad_norm": 1.0601337334222438,
      "learning_rate": 9.998477591356606e-06,
      "loss": 0.4303,
      "num_input_tokens_seen": 3971224,
      "step": 70,
      "train_runtime": 403.3932,
      "train_tokens_per_second": 9844.549
    },
    {
      "epoch": 0.01707650273224044,
      "grad_norm": 0.7210379149174336,
      "learning_rate": 9.998248971604155e-06,
      "loss": 0.3835,
      "num_input_tokens_seen": 4253680,
      "step": 75,
      "train_runtime": 429.704,
      "train_tokens_per_second": 9899.094
    },
    {
      "epoch": 0.018214936247723135,
      "grad_norm": 0.8297743601148291,
      "learning_rate": 9.998004368315486e-06,
      "loss": 0.3828,
      "num_input_tokens_seen": 4539752,
      "step": 80,
      "train_runtime": 462.5241,
      "train_tokens_per_second": 9815.169
    },
    {
      "epoch": 0.019353369763205827,
      "grad_norm": 0.7942177800514065,
      "learning_rate": 9.9977437822728e-06,
      "loss": 0.4004,
      "num_input_tokens_seen": 4821672,
      "step": 85,
      "train_runtime": 488.9156,
      "train_tokens_per_second": 9861.972
    },
    {
      "epoch": 0.020491803278688523,
      "grad_norm": 0.8538593725860453,
      "learning_rate": 9.997467214309407e-06,
      "loss": 0.4352,
      "num_input_tokens_seen": 5114672,
      "step": 90,
      "train_runtime": 518.9876,
      "train_tokens_per_second": 9855.094
    },
    {
      "epoch": 0.02163023679417122,
      "grad_norm": 0.9400067528601316,
      "learning_rate": 9.997174665309723e-06,
      "loss": 0.3724,
      "num_input_tokens_seen": 5414448,
      "step": 95,
      "train_runtime": 549.2035,
      "train_tokens_per_second": 9858.728
    },
    {
      "epoch": 0.022768670309653915,
      "grad_norm": 0.9675326218277203,
      "learning_rate": 9.996866136209269e-06,
      "loss": 0.4035,
      "num_input_tokens_seen": 5699640,
      "step": 100,
      "train_runtime": 580.8232,
      "train_tokens_per_second": 9813.038
    },
    {
      "epoch": 0.02390710382513661,
      "grad_norm": 0.9037779540586791,
      "learning_rate": 9.996541627994668e-06,
      "loss": 0.3991,
      "num_input_tokens_seen": 5990472,
      "step": 105,
      "train_runtime": 606.5442,
      "train_tokens_per_second": 9876.398
    },
    {
      "epoch": 0.025045537340619307,
      "grad_norm": 0.8378249315716529,
      "learning_rate": 9.996201141703639e-06,
      "loss": 0.4234,
      "num_input_tokens_seen": 6284664,
      "step": 110,
      "train_runtime": 638.254,
      "train_tokens_per_second": 9846.65
    },
    {
      "epoch": 0.026183970856102003,
      "grad_norm": 0.8030614218110967,
      "learning_rate": 9.995844678425002e-06,
      "loss": 0.4132,
      "num_input_tokens_seen": 6555168,
      "step": 115,
      "train_runtime": 670.5224,
      "train_tokens_per_second": 9776.21
    },
    {
      "epoch": 0.0273224043715847,
      "grad_norm": 0.9928435248830417,
      "learning_rate": 9.995472239298662e-06,
      "loss": 0.4342,
      "num_input_tokens_seen": 6848832,
      "step": 120,
      "train_runtime": 700.2167,
      "train_tokens_per_second": 9781.018
    },
    {
      "epoch": 0.028460837887067394,
      "grad_norm": 0.8448507237762608,
      "learning_rate": 9.995083825515616e-06,
      "loss": 0.4206,
      "num_input_tokens_seen": 7144528,
      "step": 125,
      "train_runtime": 734.177,
      "train_tokens_per_second": 9731.343
    },
    {
      "epoch": 0.02959927140255009,
      "grad_norm": 0.8575960899390431,
      "learning_rate": 9.994679438317944e-06,
      "loss": 0.4087,
      "num_input_tokens_seen": 7432608,
      "step": 130,
      "train_runtime": 761.4665,
      "train_tokens_per_second": 9760.913
    },
    {
      "epoch": 0.030737704918032786,
      "grad_norm": 0.8245852931162212,
      "learning_rate": 9.994259078998807e-06,
      "loss": 0.431,
      "num_input_tokens_seen": 7732480,
      "step": 135,
      "train_runtime": 790.0484,
      "train_tokens_per_second": 9787.35
    },
    {
      "epoch": 0.031876138433515486,
      "grad_norm": 0.755057214551077,
      "learning_rate": 9.99382274890244e-06,
      "loss": 0.4206,
      "num_input_tokens_seen": 8042040,
      "step": 140,
      "train_runtime": 816.9074,
      "train_tokens_per_second": 9844.494
    },
    {
      "epoch": 0.03301457194899818,
      "grad_norm": 1.1058401095354726,
      "learning_rate": 9.993370449424153e-06,
      "loss": 0.3749,
      "num_input_tokens_seen": 8320000,
      "step": 145,
      "train_runtime": 847.3021,
      "train_tokens_per_second": 9819.402
    },
    {
      "epoch": 0.03415300546448088,
      "grad_norm": 0.8358373324821667,
      "learning_rate": 9.99290218201032e-06,
      "loss": 0.38,
      "num_input_tokens_seen": 8610584,
      "step": 150,
      "train_runtime": 880.5487,
      "train_tokens_per_second": 9778.657
    },
    {
      "epoch": 0.03529143897996357,
      "grad_norm": 0.9306333481660239,
      "learning_rate": 9.992417948158382e-06,
      "loss": 0.3873,
      "num_input_tokens_seen": 8885472,
      "step": 155,
      "train_runtime": 906.0834,
      "train_tokens_per_second": 9806.461
    },
    {
      "epoch": 0.03642987249544627,
      "grad_norm": 0.8699026335844525,
      "learning_rate": 9.99191774941683e-06,
      "loss": 0.4021,
      "num_input_tokens_seen": 9172232,
      "step": 160,
      "train_runtime": 933.3351,
      "train_tokens_per_second": 9827.372
    },
    {
      "epoch": 0.03756830601092896,
      "grad_norm": 0.9071666363723179,
      "learning_rate": 9.99140158738522e-06,
      "loss": 0.3786,
      "num_input_tokens_seen": 9456832,
      "step": 165,
      "train_runtime": 960.5439,
      "train_tokens_per_second": 9845.289
    },
    {
      "epoch": 0.038706739526411654,
      "grad_norm": 2.0859226297599776,
      "learning_rate": 9.990869463714143e-06,
      "loss": 0.3298,
      "num_input_tokens_seen": 9764552,
      "step": 170,
      "train_runtime": 997.4321,
      "train_tokens_per_second": 9789.691
    },
    {
      "epoch": 0.039845173041894354,
      "grad_norm": 0.8931136219562178,
      "learning_rate": 9.99032138010524e-06,
      "loss": 0.4249,
      "num_input_tokens_seen": 10053240,
      "step": 175,
      "train_runtime": 1027.1104,
      "train_tokens_per_second": 9787.886
    },
    {
      "epoch": 0.040983606557377046,
      "grad_norm": 0.6737216786782502,
      "learning_rate": 9.98975733831119e-06,
      "loss": 0.3756,
      "num_input_tokens_seen": 10356376,
      "step": 180,
      "train_runtime": 1060.6056,
      "train_tokens_per_second": 9764.587
    },
    {
      "epoch": 0.042122040072859745,
      "grad_norm": 1.0470454220225136,
      "learning_rate": 9.989177340135697e-06,
      "loss": 0.4013,
      "num_input_tokens_seen": 10604872,
      "step": 185,
      "train_runtime": 1084.2929,
      "train_tokens_per_second": 9780.449
    },
    {
      "epoch": 0.04326047358834244,
      "grad_norm": 0.9288873799085844,
      "learning_rate": 9.9885813874335e-06,
      "loss": 0.3473,
      "num_input_tokens_seen": 10886168,
      "step": 190,
      "train_runtime": 1105.8273,
      "train_tokens_per_second": 9844.365
    },
    {
      "epoch": 0.04439890710382514,
      "grad_norm": 0.6496246477913175,
      "learning_rate": 9.987969482110349e-06,
      "loss": 0.4042,
      "num_input_tokens_seen": 11189768,
      "step": 195,
      "train_runtime": 1133.0316,
      "train_tokens_per_second": 9875.954
    },
    {
      "epoch": 0.04553734061930783,
      "grad_norm": 0.8170745742124479,
      "learning_rate": 9.987341626123011e-06,
      "loss": 0.372,
      "num_input_tokens_seen": 11463912,
      "step": 200,
      "train_runtime": 1163.1749,
      "train_tokens_per_second": 9855.708
    },
    {
      "epoch": 0.04667577413479053,
      "grad_norm": 0.7792273449856865,
      "learning_rate": 9.986697821479265e-06,
      "loss": 0.3619,
      "num_input_tokens_seen": 11762440,
      "step": 205,
      "train_runtime": 1191.0606,
      "train_tokens_per_second": 9875.602
    },
    {
      "epoch": 0.04781420765027322,
      "grad_norm": 0.914441249106063,
      "learning_rate": 9.986038070237884e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 12051232,
      "step": 210,
      "train_runtime": 1216.0506,
      "train_tokens_per_second": 9910.14
    },
    {
      "epoch": 0.04895264116575592,
      "grad_norm": 0.8658397287008582,
      "learning_rate": 9.985362374508638e-06,
      "loss": 0.416,
      "num_input_tokens_seen": 12325880,
      "step": 215,
      "train_runtime": 1246.4048,
      "train_tokens_per_second": 9889.147
    },
    {
      "epoch": 0.05009107468123861,
      "grad_norm": 0.740922047006598,
      "learning_rate": 9.984670736452287e-06,
      "loss": 0.367,
      "num_input_tokens_seen": 12609456,
      "step": 220,
      "train_runtime": 1280.2934,
      "train_tokens_per_second": 9848.88
    },
    {
      "epoch": 0.05122950819672131,
      "grad_norm": 0.718899936369804,
      "learning_rate": 9.983963158280568e-06,
      "loss": 0.4042,
      "num_input_tokens_seen": 12882992,
      "step": 225,
      "train_runtime": 1307.4226,
      "train_tokens_per_second": 9853.732
    },
    {
      "epoch": 0.052367941712204005,
      "grad_norm": 0.800872079489133,
      "learning_rate": 9.983239642256194e-06,
      "loss": 0.3915,
      "num_input_tokens_seen": 13176224,
      "step": 230,
      "train_runtime": 1332.2513,
      "train_tokens_per_second": 9890.194
    },
    {
      "epoch": 0.053506375227686705,
      "grad_norm": 0.7425386467078396,
      "learning_rate": 9.982500190692846e-06,
      "loss": 0.3914,
      "num_input_tokens_seen": 13437104,
      "step": 235,
      "train_runtime": 1359.2494,
      "train_tokens_per_second": 9885.68
    },
    {
      "epoch": 0.0546448087431694,
      "grad_norm": 0.8632176782949095,
      "learning_rate": 9.98174480595516e-06,
      "loss": 0.3754,
      "num_input_tokens_seen": 13730040,
      "step": 240,
      "train_runtime": 1389.907,
      "train_tokens_per_second": 9878.387
    },
    {
      "epoch": 0.055783242258652097,
      "grad_norm": 0.829497338544076,
      "learning_rate": 9.980973490458728e-06,
      "loss": 0.4031,
      "num_input_tokens_seen": 14023896,
      "step": 245,
      "train_runtime": 1419.5508,
      "train_tokens_per_second": 9879.109
    },
    {
      "epoch": 0.05692167577413479,
      "grad_norm": 1.0699694673102043,
      "learning_rate": 9.980186246670084e-06,
      "loss": 0.3763,
      "num_input_tokens_seen": 14291824,
      "step": 250,
      "train_runtime": 1450.6749,
      "train_tokens_per_second": 9851.845
    },
    {
      "epoch": 0.05806010928961749,
      "grad_norm": 0.7424080179684379,
      "learning_rate": 9.979383077106693e-06,
      "loss": 0.3992,
      "num_input_tokens_seen": 14564408,
      "step": 255,
      "train_runtime": 1474.6622,
      "train_tokens_per_second": 9876.437
    },
    {
      "epoch": 0.05919854280510018,
      "grad_norm": 0.7048533679442528,
      "learning_rate": 9.978563984336958e-06,
      "loss": 0.366,
      "num_input_tokens_seen": 14844576,
      "step": 260,
      "train_runtime": 1509.1474,
      "train_tokens_per_second": 9836.399
    },
    {
      "epoch": 0.06033697632058288,
      "grad_norm": 0.7237346525990511,
      "learning_rate": 9.977728970980193e-06,
      "loss": 0.3511,
      "num_input_tokens_seen": 15140568,
      "step": 265,
      "train_runtime": 1541.7832,
      "train_tokens_per_second": 9820.167
    },
    {
      "epoch": 0.06147540983606557,
      "grad_norm": 0.6481707796150433,
      "learning_rate": 9.976878039706625e-06,
      "loss": 0.3432,
      "num_input_tokens_seen": 15414024,
      "step": 270,
      "train_runtime": 1566.8887,
      "train_tokens_per_second": 9837.344
    },
    {
      "epoch": 0.06261384335154827,
      "grad_norm": 0.7306639931036337,
      "learning_rate": 9.97601119323739e-06,
      "loss": 0.4004,
      "num_input_tokens_seen": 15691184,
      "step": 275,
      "train_runtime": 1596.6079,
      "train_tokens_per_second": 9827.825
    },
    {
      "epoch": 0.06375227686703097,
      "grad_norm": 0.7612583354635174,
      "learning_rate": 9.975128434344508e-06,
      "loss": 0.3845,
      "num_input_tokens_seen": 15999952,
      "step": 280,
      "train_runtime": 1622.9155,
      "train_tokens_per_second": 9858.771
    },
    {
      "epoch": 0.06489071038251366,
      "grad_norm": 0.6683766879103441,
      "learning_rate": 9.974229765850892e-06,
      "loss": 0.3888,
      "num_input_tokens_seen": 16295032,
      "step": 285,
      "train_runtime": 1653.2603,
      "train_tokens_per_second": 9856.302
    },
    {
      "epoch": 0.06602914389799636,
      "grad_norm": 0.6094162936851902,
      "learning_rate": 9.973315190630329e-06,
      "loss": 0.3939,
      "num_input_tokens_seen": 16600424,
      "step": 290,
      "train_runtime": 1682.3992,
      "train_tokens_per_second": 9867.114
    },
    {
      "epoch": 0.06716757741347905,
      "grad_norm": 0.7991093074152439,
      "learning_rate": 9.97238471160747e-06,
      "loss": 0.3959,
      "num_input_tokens_seen": 16892296,
      "step": 295,
      "train_runtime": 1711.8602,
      "train_tokens_per_second": 9867.801
    },
    {
      "epoch": 0.06830601092896176,
      "grad_norm": 0.8087107459907734,
      "learning_rate": 9.97143833175783e-06,
      "loss": 0.3901,
      "num_input_tokens_seen": 17181832,
      "step": 300,
      "train_runtime": 1737.6197,
      "train_tokens_per_second": 9888.143
    },
    {
      "epoch": 0.06944444444444445,
      "grad_norm": 0.757176832123399,
      "learning_rate": 9.970476054107763e-06,
      "loss": 0.3695,
      "num_input_tokens_seen": 17442568,
      "step": 305,
      "train_runtime": 1764.949,
      "train_tokens_per_second": 9882.76
    },
    {
      "epoch": 0.07058287795992714,
      "grad_norm": 0.8332651078546838,
      "learning_rate": 9.969497881734473e-06,
      "loss": 0.3912,
      "num_input_tokens_seen": 17752528,
      "step": 310,
      "train_runtime": 1795.1205,
      "train_tokens_per_second": 9889.324
    },
    {
      "epoch": 0.07172131147540983,
      "grad_norm": 0.761569872567418,
      "learning_rate": 9.96850381776598e-06,
      "loss": 0.3771,
      "num_input_tokens_seen": 18054488,
      "step": 315,
      "train_runtime": 1821.9304,
      "train_tokens_per_second": 9909.538
    },
    {
      "epoch": 0.07285974499089254,
      "grad_norm": 0.750009387895909,
      "learning_rate": 9.967493865381131e-06,
      "loss": 0.3736,
      "num_input_tokens_seen": 18331064,
      "step": 320,
      "train_runtime": 1851.5919,
      "train_tokens_per_second": 9900.164
    },
    {
      "epoch": 0.07399817850637523,
      "grad_norm": 0.860997167365666,
      "learning_rate": 9.966468027809582e-06,
      "loss": 0.3599,
      "num_input_tokens_seen": 18627000,
      "step": 325,
      "train_runtime": 1880.4863,
      "train_tokens_per_second": 9905.417
    },
    {
      "epoch": 0.07513661202185792,
      "grad_norm": 0.8505695378549932,
      "learning_rate": 9.965426308331779e-06,
      "loss": 0.3711,
      "num_input_tokens_seen": 18904320,
      "step": 330,
      "train_runtime": 1906.1131,
      "train_tokens_per_second": 9917.733
    },
    {
      "epoch": 0.07627504553734062,
      "grad_norm": 0.7178502227070662,
      "learning_rate": 9.964368710278964e-06,
      "loss": 0.3692,
      "num_input_tokens_seen": 19184320,
      "step": 335,
      "train_runtime": 1935.1988,
      "train_tokens_per_second": 9913.359
    },
    {
      "epoch": 0.07741347905282331,
      "grad_norm": 0.7763508448563429,
      "learning_rate": 9.963295237033152e-06,
      "loss": 0.3343,
      "num_input_tokens_seen": 19479552,
      "step": 340,
      "train_runtime": 1958.5596,
      "train_tokens_per_second": 9945.856
    },
    {
      "epoch": 0.07855191256830601,
      "grad_norm": 0.7706185508215339,
      "learning_rate": 9.962205892027129e-06,
      "loss": 0.3776,
      "num_input_tokens_seen": 19769216,
      "step": 345,
      "train_runtime": 1985.39,
      "train_tokens_per_second": 9957.347
    },
    {
      "epoch": 0.07969034608378871,
      "grad_norm": 0.9607669656558637,
      "learning_rate": 9.961100678744425e-06,
      "loss": 0.3641,
      "num_input_tokens_seen": 20042240,
      "step": 350,
      "train_runtime": 2011.9222,
      "train_tokens_per_second": 9961.737
    },
    {
      "epoch": 0.0808287795992714,
      "grad_norm": 0.8408976742523321,
      "learning_rate": 9.959979600719322e-06,
      "loss": 0.38,
      "num_input_tokens_seen": 20344784,
      "step": 355,
      "train_runtime": 2039.8908,
      "train_tokens_per_second": 9973.467
    },
    {
      "epoch": 0.08196721311475409,
      "grad_norm": 0.8616351153973422,
      "learning_rate": 9.958842661536838e-06,
      "loss": 0.3792,
      "num_input_tokens_seen": 20612560,
      "step": 360,
      "train_runtime": 2068.5884,
      "train_tokens_per_second": 9964.554
    },
    {
      "epoch": 0.0831056466302368,
      "grad_norm": 1.4668899485390243,
      "learning_rate": 9.957689864832707e-06,
      "loss": 0.3658,
      "num_input_tokens_seen": 20891440,
      "step": 365,
      "train_runtime": 2093.3192,
      "train_tokens_per_second": 9980.055
    },
    {
      "epoch": 0.08424408014571949,
      "grad_norm": 0.7501078754261734,
      "learning_rate": 9.95652121429337e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 21175096,
      "step": 370,
      "train_runtime": 2121.3378,
      "train_tokens_per_second": 9981.954
    },
    {
      "epoch": 0.08538251366120218,
      "grad_norm": 0.8846652494537139,
      "learning_rate": 9.955336713655976e-06,
      "loss": 0.3544,
      "num_input_tokens_seen": 21453984,
      "step": 375,
      "train_runtime": 2149.2712,
      "train_tokens_per_second": 9981.981
    },
    {
      "epoch": 0.08652094717668488,
      "grad_norm": 0.8325406263550199,
      "learning_rate": 9.954136366708346e-06,
      "loss": 0.364,
      "num_input_tokens_seen": 21736048,
      "step": 380,
      "train_runtime": 2183.5787,
      "train_tokens_per_second": 9954.323
    },
    {
      "epoch": 0.08765938069216758,
      "grad_norm": 0.7992795663122627,
      "learning_rate": 9.952920177288985e-06,
      "loss": 0.3777,
      "num_input_tokens_seen": 22029896,
      "step": 385,
      "train_runtime": 2214.7485,
      "train_tokens_per_second": 9946.906
    },
    {
      "epoch": 0.08879781420765027,
      "grad_norm": 0.8227178297109993,
      "learning_rate": 9.951688149287059e-06,
      "loss": 0.4004,
      "num_input_tokens_seen": 22332640,
      "step": 390,
      "train_runtime": 2246.4465,
      "train_tokens_per_second": 9941.318
    },
    {
      "epoch": 0.08993624772313297,
      "grad_norm": 0.8284050546905042,
      "learning_rate": 9.950440286642377e-06,
      "loss": 0.352,
      "num_input_tokens_seen": 22609944,
      "step": 395,
      "train_runtime": 2270.2191,
      "train_tokens_per_second": 9959.367
    },
    {
      "epoch": 0.09107468123861566,
      "grad_norm": 0.6281005189348137,
      "learning_rate": 9.94917659334539e-06,
      "loss": 0.3483,
      "num_input_tokens_seen": 22917416,
      "step": 400,
      "train_runtime": 2301.1526,
      "train_tokens_per_second": 9959.103
    },
    {
      "epoch": 0.09221311475409837,
      "grad_norm": 0.7742210740648754,
      "learning_rate": 9.947897073437169e-06,
      "loss": 0.3412,
      "num_input_tokens_seen": 23208504,
      "step": 405,
      "train_runtime": 2329.1487,
      "train_tokens_per_second": 9964.372
    },
    {
      "epoch": 0.09335154826958106,
      "grad_norm": 0.776979980775816,
      "learning_rate": 9.946601731009398e-06,
      "loss": 0.3806,
      "num_input_tokens_seen": 23493840,
      "step": 410,
      "train_runtime": 2359.4517,
      "train_tokens_per_second": 9957.33
    },
    {
      "epoch": 0.09448998178506375,
      "grad_norm": 0.7331045127513112,
      "learning_rate": 9.945290570204361e-06,
      "loss": 0.358,
      "num_input_tokens_seen": 23783160,
      "step": 415,
      "train_runtime": 2393.0343,
      "train_tokens_per_second": 9938.495
    },
    {
      "epoch": 0.09562841530054644,
      "grad_norm": 0.6989705775070271,
      "learning_rate": 9.943963595214921e-06,
      "loss": 0.3675,
      "num_input_tokens_seen": 24083528,
      "step": 420,
      "train_runtime": 2419.8013,
      "train_tokens_per_second": 9952.688
    },
    {
      "epoch": 0.09676684881602915,
      "grad_norm": 0.7642768719960205,
      "learning_rate": 9.942620810284513e-06,
      "loss": 0.3877,
      "num_input_tokens_seen": 24350272,
      "step": 425,
      "train_runtime": 2449.934,
      "train_tokens_per_second": 9939.154
    },
    {
      "epoch": 0.09790528233151184,
      "grad_norm": 0.825328154545686,
      "learning_rate": 9.941262219707135e-06,
      "loss": 0.3482,
      "num_input_tokens_seen": 24620256,
      "step": 430,
      "train_runtime": 2475.2454,
      "train_tokens_per_second": 9946.592
    },
    {
      "epoch": 0.09904371584699453,
      "grad_norm": 0.920308708104242,
      "learning_rate": 9.939887827827321e-06,
      "loss": 0.3909,
      "num_input_tokens_seen": 24868488,
      "step": 435,
      "train_runtime": 2505.6185,
      "train_tokens_per_second": 9925.089
    },
    {
      "epoch": 0.10018214936247723,
      "grad_norm": 0.6478316910236817,
      "learning_rate": 9.938497639040143e-06,
      "loss": 0.3726,
      "num_input_tokens_seen": 25175208,
      "step": 440,
      "train_runtime": 2531.0558,
      "train_tokens_per_second": 9946.524
    },
    {
      "epoch": 0.10132058287795993,
      "grad_norm": 0.6659162911931353,
      "learning_rate": 9.93709165779118e-06,
      "loss": 0.3576,
      "num_input_tokens_seen": 25449288,
      "step": 445,
      "train_runtime": 2555.7349,
      "train_tokens_per_second": 9957.718
    },
    {
      "epoch": 0.10245901639344263,
      "grad_norm": 0.7118794376732661,
      "learning_rate": 9.93566988857652e-06,
      "loss": 0.4103,
      "num_input_tokens_seen": 25723960,
      "step": 450,
      "train_runtime": 2586.9549,
      "train_tokens_per_second": 9943.722
    },
    {
      "epoch": 0.10359744990892532,
      "grad_norm": 0.689192650612493,
      "learning_rate": 9.934232335942735e-06,
      "loss": 0.353,
      "num_input_tokens_seen": 26013512,
      "step": 455,
      "train_runtime": 2612.1845,
      "train_tokens_per_second": 9958.528
    },
    {
      "epoch": 0.10473588342440801,
      "grad_norm": 0.7174781362571463,
      "learning_rate": 9.932779004486867e-06,
      "loss": 0.3537,
      "num_input_tokens_seen": 26315552,
      "step": 460,
      "train_runtime": 2640.423,
      "train_tokens_per_second": 9966.415
    },
    {
      "epoch": 0.10587431693989072,
      "grad_norm": 0.7646380354525935,
      "learning_rate": 9.931309898856423e-06,
      "loss": 0.37,
      "num_input_tokens_seen": 26606784,
      "step": 465,
      "train_runtime": 2660.2192,
      "train_tokens_per_second": 10001.726
    },
    {
      "epoch": 0.10701275045537341,
      "grad_norm": 0.7902706390249465,
      "learning_rate": 9.929825023749345e-06,
      "loss": 0.3688,
      "num_input_tokens_seen": 26883568,
      "step": 470,
      "train_runtime": 2689.8178,
      "train_tokens_per_second": 9994.568
    },
    {
      "epoch": 0.1081511839708561,
      "grad_norm": 0.6889232050932863,
      "learning_rate": 9.928324383914009e-06,
      "loss": 0.3561,
      "num_input_tokens_seen": 27173024,
      "step": 475,
      "train_runtime": 2719.0191,
      "train_tokens_per_second": 9993.686
    },
    {
      "epoch": 0.1092896174863388,
      "grad_norm": 0.993657581630257,
      "learning_rate": 9.926807984149203e-06,
      "loss": 0.3677,
      "num_input_tokens_seen": 27472672,
      "step": 480,
      "train_runtime": 2737.0307,
      "train_tokens_per_second": 10037.4
    },
    {
      "epoch": 0.11042805100182149,
      "grad_norm": 0.68857075554738,
      "learning_rate": 9.925275829304107e-06,
      "loss": 0.3763,
      "num_input_tokens_seen": 27778424,
      "step": 485,
      "train_runtime": 2763.669,
      "train_tokens_per_second": 10051.285
    },
    {
      "epoch": 0.11156648451730419,
      "grad_norm": 0.7286094152758079,
      "learning_rate": 9.923727924278293e-06,
      "loss": 0.389,
      "num_input_tokens_seen": 28076792,
      "step": 490,
      "train_runtime": 2790.6535,
      "train_tokens_per_second": 10061.01
    },
    {
      "epoch": 0.11270491803278689,
      "grad_norm": 0.929300862716688,
      "learning_rate": 9.922164274021689e-06,
      "loss": 0.3676,
      "num_input_tokens_seen": 28371352,
      "step": 495,
      "train_runtime": 2817.2495,
      "train_tokens_per_second": 10070.585
    },
    {
      "epoch": 0.11384335154826958,
      "grad_norm": 0.6861078778237657,
      "learning_rate": 9.92058488353458e-06,
      "loss": 0.3384,
      "num_input_tokens_seen": 28682480,
      "step": 500,
      "train_runtime": 2849.3206,
      "train_tokens_per_second": 10066.428
    },
    {
      "epoch": 0.11498178506375227,
      "grad_norm": 0.6770434525934667,
      "learning_rate": 9.918989757867584e-06,
      "loss": 0.3536,
      "num_input_tokens_seen": 28976352,
      "step": 505,
      "train_runtime": 2879.965,
      "train_tokens_per_second": 10061.355
    },
    {
      "epoch": 0.11612021857923498,
      "grad_norm": 0.75930140205932,
      "learning_rate": 9.917378902121635e-06,
      "loss": 0.3314,
      "num_input_tokens_seen": 29270568,
      "step": 510,
      "train_runtime": 2910.3186,
      "train_tokens_per_second": 10057.513
    },
    {
      "epoch": 0.11725865209471767,
      "grad_norm": 0.6666533364669364,
      "learning_rate": 9.915752321447976e-06,
      "loss": 0.3498,
      "num_input_tokens_seen": 29588360,
      "step": 515,
      "train_runtime": 2939.1861,
      "train_tokens_per_second": 10066.855
    },
    {
      "epoch": 0.11839708561020036,
      "grad_norm": 0.6932025528861782,
      "learning_rate": 9.914110021048126e-06,
      "loss": 0.3779,
      "num_input_tokens_seen": 29884336,
      "step": 520,
      "train_runtime": 2968.3719,
      "train_tokens_per_second": 10067.585
    },
    {
      "epoch": 0.11953551912568305,
      "grad_norm": 0.8859072555071915,
      "learning_rate": 9.91245200617388e-06,
      "loss": 0.3785,
      "num_input_tokens_seen": 30167544,
      "step": 525,
      "train_runtime": 2999.0003,
      "train_tokens_per_second": 10059.2
    },
    {
      "epoch": 0.12067395264116576,
      "grad_norm": 0.7704268193321102,
      "learning_rate": 9.910778282127283e-06,
      "loss": 0.3388,
      "num_input_tokens_seen": 30452656,
      "step": 530,
      "train_runtime": 3024.6424,
      "train_tokens_per_second": 10068.184
    },
    {
      "epoch": 0.12181238615664845,
      "grad_norm": 0.8509898519550806,
      "learning_rate": 9.909088854260615e-06,
      "loss": 0.3485,
      "num_input_tokens_seen": 30727560,
      "step": 535,
      "train_runtime": 3056.5733,
      "train_tokens_per_second": 10052.944
    },
    {
      "epoch": 0.12295081967213115,
      "grad_norm": 0.6918068910145782,
      "learning_rate": 9.907383727976374e-06,
      "loss": 0.3612,
      "num_input_tokens_seen": 31023688,
      "step": 540,
      "train_runtime": 3084.4768,
      "train_tokens_per_second": 10058.007
    },
    {
      "epoch": 0.12408925318761384,
      "grad_norm": 0.6941754966485706,
      "learning_rate": 9.905662908727258e-06,
      "loss": 0.3647,
      "num_input_tokens_seen": 31313752,
      "step": 545,
      "train_runtime": 3104.3318,
      "train_tokens_per_second": 10087.115
    },
    {
      "epoch": 0.12522768670309653,
      "grad_norm": 0.6533230964683835,
      "learning_rate": 9.903926402016153e-06,
      "loss": 0.3651,
      "num_input_tokens_seen": 31612416,
      "step": 550,
      "train_runtime": 3132.9928,
      "train_tokens_per_second": 10090.166
    },
    {
      "epoch": 0.12636612021857924,
      "grad_norm": 0.7224920434414023,
      "learning_rate": 9.902174213396104e-06,
      "loss": 0.3556,
      "num_input_tokens_seen": 31890896,
      "step": 555,
      "train_runtime": 3161.384,
      "train_tokens_per_second": 10087.638
    },
    {
      "epoch": 0.12750455373406194,
      "grad_norm": 0.6241757187231222,
      "learning_rate": 9.900406348470309e-06,
      "loss": 0.3313,
      "num_input_tokens_seen": 32190912,
      "step": 560,
      "train_runtime": 3188.1532,
      "train_tokens_per_second": 10097.041
    },
    {
      "epoch": 0.12864298724954462,
      "grad_norm": 0.7161895373310507,
      "learning_rate": 9.898622812892094e-06,
      "loss": 0.3492,
      "num_input_tokens_seen": 32492224,
      "step": 565,
      "train_runtime": 3221.9966,
      "train_tokens_per_second": 10084.5
    },
    {
      "epoch": 0.12978142076502733,
      "grad_norm": 0.5799605456062173,
      "learning_rate": 9.896823612364895e-06,
      "loss": 0.3237,
      "num_input_tokens_seen": 32781440,
      "step": 570,
      "train_runtime": 3248.0179,
      "train_tokens_per_second": 10092.752
    },
    {
      "epoch": 0.13091985428051,
      "grad_norm": 0.6072589043535745,
      "learning_rate": 9.89500875264225e-06,
      "loss": 0.3227,
      "num_input_tokens_seen": 33082200,
      "step": 575,
      "train_runtime": 3280.2925,
      "train_tokens_per_second": 10085.137
    },
    {
      "epoch": 0.1320582877959927,
      "grad_norm": 0.7069010227381681,
      "learning_rate": 9.89317823952776e-06,
      "loss": 0.3673,
      "num_input_tokens_seen": 33346168,
      "step": 580,
      "train_runtime": 3307.017,
      "train_tokens_per_second": 10083.458
    },
    {
      "epoch": 0.13319672131147542,
      "grad_norm": 0.9274942603558426,
      "learning_rate": 9.891332078875094e-06,
      "loss": 0.3176,
      "num_input_tokens_seen": 33624168,
      "step": 585,
      "train_runtime": 3343.5164,
      "train_tokens_per_second": 10056.528
    },
    {
      "epoch": 0.1343351548269581,
      "grad_norm": 0.7447556444079537,
      "learning_rate": 9.889470276587954e-06,
      "loss": 0.3652,
      "num_input_tokens_seen": 33910712,
      "step": 590,
      "train_runtime": 3370.938,
      "train_tokens_per_second": 10059.726
    },
    {
      "epoch": 0.1354735883424408,
      "grad_norm": 0.8515905120713488,
      "learning_rate": 9.88759283862006e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 34160528,
      "step": 595,
      "train_runtime": 3395.1707,
      "train_tokens_per_second": 10061.505
    },
    {
      "epoch": 0.1366120218579235,
      "grad_norm": 0.6610678251710973,
      "learning_rate": 9.885699770975136e-06,
      "loss": 0.2982,
      "num_input_tokens_seen": 34462656,
      "step": 600,
      "train_runtime": 3425.4889,
      "train_tokens_per_second": 10060.653
    },
    {
      "epoch": 0.1377504553734062,
      "grad_norm": 0.661506106563137,
      "learning_rate": 9.883791079706885e-06,
      "loss": 0.3789,
      "num_input_tokens_seen": 34754120,
      "step": 605,
      "train_runtime": 3453.8726,
      "train_tokens_per_second": 10062.363
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 0.7238875299991251,
      "learning_rate": 9.881866770918971e-06,
      "loss": 0.3702,
      "num_input_tokens_seen": 35023624,
      "step": 610,
      "train_runtime": 3482.3979,
      "train_tokens_per_second": 10057.33
    },
    {
      "epoch": 0.14002732240437157,
      "grad_norm": 0.7200818218043179,
      "learning_rate": 9.879926850765001e-06,
      "loss": 0.3604,
      "num_input_tokens_seen": 35316744,
      "step": 615,
      "train_runtime": 3511.5632,
      "train_tokens_per_second": 10057.271
    },
    {
      "epoch": 0.14116575591985428,
      "grad_norm": 0.685285254771609,
      "learning_rate": 9.877971325448505e-06,
      "loss": 0.3317,
      "num_input_tokens_seen": 35618944,
      "step": 620,
      "train_runtime": 3538.3186,
      "train_tokens_per_second": 10066.63
    },
    {
      "epoch": 0.142304189435337,
      "grad_norm": 0.8330111036765857,
      "learning_rate": 9.876000201222912e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 35907648,
      "step": 625,
      "train_runtime": 3564.3461,
      "train_tokens_per_second": 10074.119
    },
    {
      "epoch": 0.14344262295081966,
      "grad_norm": 0.6695117188228695,
      "learning_rate": 9.874013484391542e-06,
      "loss": 0.3543,
      "num_input_tokens_seen": 36201064,
      "step": 630,
      "train_runtime": 3594.1203,
      "train_tokens_per_second": 10072.302
    },
    {
      "epoch": 0.14458105646630237,
      "grad_norm": 0.7061930482305514,
      "learning_rate": 9.872011181307568e-06,
      "loss": 0.3512,
      "num_input_tokens_seen": 36465392,
      "step": 635,
      "train_runtime": 3624.5428,
      "train_tokens_per_second": 10060.687
    },
    {
      "epoch": 0.14571948998178508,
      "grad_norm": 0.6312433783262524,
      "learning_rate": 9.86999329837401e-06,
      "loss": 0.3362,
      "num_input_tokens_seen": 36745624,
      "step": 640,
      "train_runtime": 3658.4404,
      "train_tokens_per_second": 10044.068
    },
    {
      "epoch": 0.14685792349726776,
      "grad_norm": 0.7733761806619178,
      "learning_rate": 9.867959842043707e-06,
      "loss": 0.3534,
      "num_input_tokens_seen": 37043760,
      "step": 645,
      "train_runtime": 3682.7133,
      "train_tokens_per_second": 10058.823
    },
    {
      "epoch": 0.14799635701275046,
      "grad_norm": 0.7174459969183352,
      "learning_rate": 9.865910818819304e-06,
      "loss": 0.3555,
      "num_input_tokens_seen": 37341392,
      "step": 650,
      "train_runtime": 3710.7905,
      "train_tokens_per_second": 10062.921
    },
    {
      "epoch": 0.14913479052823314,
      "grad_norm": 0.6448141291208076,
      "learning_rate": 9.863846235253223e-06,
      "loss": 0.333,
      "num_input_tokens_seen": 37648168,
      "step": 655,
      "train_runtime": 3735.9836,
      "train_tokens_per_second": 10077.177
    },
    {
      "epoch": 0.15027322404371585,
      "grad_norm": 0.6399097709873408,
      "learning_rate": 9.861766097947642e-06,
      "loss": 0.3648,
      "num_input_tokens_seen": 37950608,
      "step": 660,
      "train_runtime": 3772.9972,
      "train_tokens_per_second": 10058.478
    },
    {
      "epoch": 0.15141165755919855,
      "grad_norm": 0.6664771876627038,
      "learning_rate": 9.859670413554483e-06,
      "loss": 0.3729,
      "num_input_tokens_seen": 38227600,
      "step": 665,
      "train_runtime": 3795.4505,
      "train_tokens_per_second": 10071.953
    },
    {
      "epoch": 0.15255009107468123,
      "grad_norm": 0.6814021629094236,
      "learning_rate": 9.857559188775383e-06,
      "loss": 0.374,
      "num_input_tokens_seen": 38532184,
      "step": 670,
      "train_runtime": 3815.361,
      "train_tokens_per_second": 10099.224
    },
    {
      "epoch": 0.15368852459016394,
      "grad_norm": 0.7037824820438056,
      "learning_rate": 9.855432430361672e-06,
      "loss": 0.3575,
      "num_input_tokens_seen": 38804432,
      "step": 675,
      "train_runtime": 3840.5001,
      "train_tokens_per_second": 10104.005
    },
    {
      "epoch": 0.15482695810564662,
      "grad_norm": 0.6277501025518656,
      "learning_rate": 9.853290145114357e-06,
      "loss": 0.3446,
      "num_input_tokens_seen": 39104016,
      "step": 680,
      "train_runtime": 3869.0979,
      "train_tokens_per_second": 10106.753
    },
    {
      "epoch": 0.15596539162112932,
      "grad_norm": 0.6759361433072625,
      "learning_rate": 9.851132339884097e-06,
      "loss": 0.3134,
      "num_input_tokens_seen": 39390824,
      "step": 685,
      "train_runtime": 3896.2419,
      "train_tokens_per_second": 10109.953
    },
    {
      "epoch": 0.15710382513661203,
      "grad_norm": 0.7068260520380004,
      "learning_rate": 9.848959021571178e-06,
      "loss": 0.344,
      "num_input_tokens_seen": 39670168,
      "step": 690,
      "train_runtime": 3925.0705,
      "train_tokens_per_second": 10106.868
    },
    {
      "epoch": 0.1582422586520947,
      "grad_norm": 0.594698248620102,
      "learning_rate": 9.846770197125497e-06,
      "loss": 0.3205,
      "num_input_tokens_seen": 39976392,
      "step": 695,
      "train_runtime": 3951.3601,
      "train_tokens_per_second": 10117.122
    },
    {
      "epoch": 0.15938069216757741,
      "grad_norm": 0.7002311152247686,
      "learning_rate": 9.844565873546535e-06,
      "loss": 0.3386,
      "num_input_tokens_seen": 40248704,
      "step": 700,
      "train_runtime": 3976.1163,
      "train_tokens_per_second": 10122.617
    },
    {
      "epoch": 0.16051912568306012,
      "grad_norm": 0.6428667914573164,
      "learning_rate": 9.842346057883341e-06,
      "loss": 0.3231,
      "num_input_tokens_seen": 40533272,
      "step": 705,
      "train_runtime": 4005.2123,
      "train_tokens_per_second": 10120.131
    },
    {
      "epoch": 0.1616575591985428,
      "grad_norm": 0.7673360624904638,
      "learning_rate": 9.840110757234499e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 40807736,
      "step": 710,
      "train_runtime": 4034.4744,
      "train_tokens_per_second": 10114.759
    },
    {
      "epoch": 0.1627959927140255,
      "grad_norm": 0.7704603469892927,
      "learning_rate": 9.837859978748114e-06,
      "loss": 0.3589,
      "num_input_tokens_seen": 41075776,
      "step": 715,
      "train_runtime": 4064.1685,
      "train_tokens_per_second": 10106.809
    },
    {
      "epoch": 0.16393442622950818,
      "grad_norm": 0.6979529231481143,
      "learning_rate": 9.835593729621789e-06,
      "loss": 0.3529,
      "num_input_tokens_seen": 41351576,
      "step": 720,
      "train_runtime": 4085.6385,
      "train_tokens_per_second": 10121.203
    },
    {
      "epoch": 0.1650728597449909,
      "grad_norm": 0.6953437079380005,
      "learning_rate": 9.833312017102594e-06,
      "loss": 0.34,
      "num_input_tokens_seen": 41655768,
      "step": 725,
      "train_runtime": 4112.9695,
      "train_tokens_per_second": 10127.906
    },
    {
      "epoch": 0.1662112932604736,
      "grad_norm": 0.7613998917100598,
      "learning_rate": 9.831014848487053e-06,
      "loss": 0.3523,
      "num_input_tokens_seen": 41922064,
      "step": 730,
      "train_runtime": 4144.8031,
      "train_tokens_per_second": 10114.368
    },
    {
      "epoch": 0.16734972677595628,
      "grad_norm": 0.739337931243999,
      "learning_rate": 9.828702231121114e-06,
      "loss": 0.3623,
      "num_input_tokens_seen": 42223712,
      "step": 735,
      "train_runtime": 4170.7635,
      "train_tokens_per_second": 10123.737
    },
    {
      "epoch": 0.16848816029143898,
      "grad_norm": 0.6826810138678961,
      "learning_rate": 9.826374172400128e-06,
      "loss": 0.368,
      "num_input_tokens_seen": 42491248,
      "step": 740,
      "train_runtime": 4195.3856,
      "train_tokens_per_second": 10128.091
    },
    {
      "epoch": 0.1696265938069217,
      "grad_norm": 0.8758842646727611,
      "learning_rate": 9.824030679768824e-06,
      "loss": 0.3699,
      "num_input_tokens_seen": 42772616,
      "step": 745,
      "train_runtime": 4223.2779,
      "train_tokens_per_second": 10127.824
    },
    {
      "epoch": 0.17076502732240437,
      "grad_norm": 0.6749272868193374,
      "learning_rate": 9.821671760721285e-06,
      "loss": 0.3348,
      "num_input_tokens_seen": 43061864,
      "step": 750,
      "train_runtime": 4254.072,
      "train_tokens_per_second": 10122.505
    },
    {
      "epoch": 0.17190346083788707,
      "grad_norm": 0.628642212280747,
      "learning_rate": 9.819297422800927e-06,
      "loss": 0.3208,
      "num_input_tokens_seen": 43370296,
      "step": 755,
      "train_runtime": 4278.6802,
      "train_tokens_per_second": 10136.372
    },
    {
      "epoch": 0.17304189435336975,
      "grad_norm": 0.6275214163531526,
      "learning_rate": 9.816907673600475e-06,
      "loss": 0.3532,
      "num_input_tokens_seen": 43660472,
      "step": 760,
      "train_runtime": 4305.4375,
      "train_tokens_per_second": 10140.775
    },
    {
      "epoch": 0.17418032786885246,
      "grad_norm": 0.7683374055542489,
      "learning_rate": 9.814502520761929e-06,
      "loss": 0.3669,
      "num_input_tokens_seen": 43936520,
      "step": 765,
      "train_runtime": 4335.6507,
      "train_tokens_per_second": 10133.778
    },
    {
      "epoch": 0.17531876138433516,
      "grad_norm": 0.7425660678752025,
      "learning_rate": 9.812081971976555e-06,
      "loss": 0.3425,
      "num_input_tokens_seen": 44234736,
      "step": 770,
      "train_runtime": 4365.581,
      "train_tokens_per_second": 10132.611
    },
    {
      "epoch": 0.17645719489981784,
      "grad_norm": 0.6522685907397228,
      "learning_rate": 9.80964603498485e-06,
      "loss": 0.3322,
      "num_input_tokens_seen": 44542264,
      "step": 775,
      "train_runtime": 4392.344,
      "train_tokens_per_second": 10140.887
    },
    {
      "epoch": 0.17759562841530055,
      "grad_norm": 0.6069399052066745,
      "learning_rate": 9.807194717576518e-06,
      "loss": 0.3107,
      "num_input_tokens_seen": 44835440,
      "step": 780,
      "train_runtime": 4424.7563,
      "train_tokens_per_second": 10132.861
    },
    {
      "epoch": 0.17873406193078326,
      "grad_norm": 0.7582543780619266,
      "learning_rate": 9.80472802759045e-06,
      "loss": 0.361,
      "num_input_tokens_seen": 45129440,
      "step": 785,
      "train_runtime": 4457.7281,
      "train_tokens_per_second": 10123.866
    },
    {
      "epoch": 0.17987249544626593,
      "grad_norm": 0.6190210134669156,
      "learning_rate": 9.80224597291469e-06,
      "loss": 0.3167,
      "num_input_tokens_seen": 45422136,
      "step": 790,
      "train_runtime": 4486.7189,
      "train_tokens_per_second": 10123.687
    },
    {
      "epoch": 0.18101092896174864,
      "grad_norm": 0.8713152149498726,
      "learning_rate": 9.799748561486426e-06,
      "loss": 0.3512,
      "num_input_tokens_seen": 45724784,
      "step": 795,
      "train_runtime": 4515.4591,
      "train_tokens_per_second": 10126.276
    },
    {
      "epoch": 0.18214936247723132,
      "grad_norm": 0.6984182833089245,
      "learning_rate": 9.797235801291944e-06,
      "loss": 0.3599,
      "num_input_tokens_seen": 45987640,
      "step": 800,
      "train_runtime": 4546.7303,
      "train_tokens_per_second": 10114.442
    },
    {
      "epoch": 0.18328779599271403,
      "grad_norm": 0.7458139760503385,
      "learning_rate": 9.794707700366616e-06,
      "loss": 0.3244,
      "num_input_tokens_seen": 46266176,
      "step": 805,
      "train_runtime": 4574.8569,
      "train_tokens_per_second": 10113.142
    },
    {
      "epoch": 0.18442622950819673,
      "grad_norm": 0.7494078025263322,
      "learning_rate": 9.792164266794876e-06,
      "loss": 0.349,
      "num_input_tokens_seen": 46537248,
      "step": 810,
      "train_runtime": 4603.1291,
      "train_tokens_per_second": 10109.916
    },
    {
      "epoch": 0.1855646630236794,
      "grad_norm": 0.6650697037733292,
      "learning_rate": 9.78960550871018e-06,
      "loss": 0.3474,
      "num_input_tokens_seen": 46845408,
      "step": 815,
      "train_runtime": 4634.7946,
      "train_tokens_per_second": 10107.332
    },
    {
      "epoch": 0.18670309653916212,
      "grad_norm": 0.667271644684756,
      "learning_rate": 9.787031434294998e-06,
      "loss": 0.3587,
      "num_input_tokens_seen": 47121752,
      "step": 820,
      "train_runtime": 4662.2726,
      "train_tokens_per_second": 10107.035
    },
    {
      "epoch": 0.1878415300546448,
      "grad_norm": 0.715419958890253,
      "learning_rate": 9.784442051780773e-06,
      "loss": 0.3483,
      "num_input_tokens_seen": 47418616,
      "step": 825,
      "train_runtime": 4693.9741,
      "train_tokens_per_second": 10102.019
    },
    {
      "epoch": 0.1889799635701275,
      "grad_norm": 0.7187028787204937,
      "learning_rate": 9.781837369447904e-06,
      "loss": 0.3499,
      "num_input_tokens_seen": 47707256,
      "step": 830,
      "train_runtime": 4714.4066,
      "train_tokens_per_second": 10119.461
    },
    {
      "epoch": 0.1901183970856102,
      "grad_norm": 0.7063481630197213,
      "learning_rate": 9.779217395625713e-06,
      "loss": 0.3568,
      "num_input_tokens_seen": 47991840,
      "step": 835,
      "train_runtime": 4745.0986,
      "train_tokens_per_second": 10113.982
    },
    {
      "epoch": 0.1912568306010929,
      "grad_norm": 0.6300364427220898,
      "learning_rate": 9.776582138692424e-06,
      "loss": 0.3678,
      "num_input_tokens_seen": 48284640,
      "step": 840,
      "train_runtime": 4777.6582,
      "train_tokens_per_second": 10106.34
    },
    {
      "epoch": 0.1923952641165756,
      "grad_norm": 0.6116049074601653,
      "learning_rate": 9.773931607075133e-06,
      "loss": 0.3556,
      "num_input_tokens_seen": 48578120,
      "step": 845,
      "train_runtime": 4806.9757,
      "train_tokens_per_second": 10105.755
    },
    {
      "epoch": 0.1935336976320583,
      "grad_norm": 0.7299157295567328,
      "learning_rate": 9.771265809249784e-06,
      "loss": 0.3895,
      "num_input_tokens_seen": 48846368,
      "step": 850,
      "train_runtime": 4842.4197,
      "train_tokens_per_second": 10087.182
    },
    {
      "epoch": 0.19467213114754098,
      "grad_norm": 0.8355241140627414,
      "learning_rate": 9.768584753741134e-06,
      "loss": 0.351,
      "num_input_tokens_seen": 49138416,
      "step": 855,
      "train_runtime": 4869.7554,
      "train_tokens_per_second": 10090.531
    },
    {
      "epoch": 0.19581056466302368,
      "grad_norm": 0.6270207502794843,
      "learning_rate": 9.76588844912274e-06,
      "loss": 0.3505,
      "num_input_tokens_seen": 49440048,
      "step": 860,
      "train_runtime": 4895.8839,
      "train_tokens_per_second": 10098.288
    },
    {
      "epoch": 0.19694899817850636,
      "grad_norm": 0.6603178105995898,
      "learning_rate": 9.763176904016914e-06,
      "loss": 0.3677,
      "num_input_tokens_seen": 49726552,
      "step": 865,
      "train_runtime": 4927.2331,
      "train_tokens_per_second": 10092.186
    },
    {
      "epoch": 0.19808743169398907,
      "grad_norm": 0.672053015080484,
      "learning_rate": 9.76045012709471e-06,
      "loss": 0.373,
      "num_input_tokens_seen": 50007320,
      "step": 870,
      "train_runtime": 4957.7117,
      "train_tokens_per_second": 10086.775
    },
    {
      "epoch": 0.19922586520947178,
      "grad_norm": 0.7448205118126439,
      "learning_rate": 9.757708127075888e-06,
      "loss": 0.3716,
      "num_input_tokens_seen": 50283032,
      "step": 875,
      "train_runtime": 4982.7917,
      "train_tokens_per_second": 10091.337
    },
    {
      "epoch": 0.20036429872495445,
      "grad_norm": 0.6699433559468838,
      "learning_rate": 9.754950912728891e-06,
      "loss": 0.3349,
      "num_input_tokens_seen": 50595616,
      "step": 880,
      "train_runtime": 5014.1667,
      "train_tokens_per_second": 10090.533
    },
    {
      "epoch": 0.20150273224043716,
      "grad_norm": 1.1115008156200523,
      "learning_rate": 9.752178492870814e-06,
      "loss": 0.3497,
      "num_input_tokens_seen": 50882040,
      "step": 885,
      "train_runtime": 5040.9974,
      "train_tokens_per_second": 10093.645
    },
    {
      "epoch": 0.20264116575591987,
      "grad_norm": 0.6130203091550814,
      "learning_rate": 9.749390876367377e-06,
      "loss": 0.3569,
      "num_input_tokens_seen": 51143296,
      "step": 890,
      "train_runtime": 5066.179,
      "train_tokens_per_second": 10095.043
    },
    {
      "epoch": 0.20377959927140255,
      "grad_norm": 0.635778936105247,
      "learning_rate": 9.746588072132895e-06,
      "loss": 0.3564,
      "num_input_tokens_seen": 51446536,
      "step": 895,
      "train_runtime": 5096.1915,
      "train_tokens_per_second": 10095.095
    },
    {
      "epoch": 0.20491803278688525,
      "grad_norm": 0.6219906710835346,
      "learning_rate": 9.74377008913025e-06,
      "loss": 0.3121,
      "num_input_tokens_seen": 51742352,
      "step": 900,
      "train_runtime": 5117.9133,
      "train_tokens_per_second": 10110.049
    },
    {
      "epoch": 0.20605646630236793,
      "grad_norm": 0.9029842308007769,
      "learning_rate": 9.74093693637087e-06,
      "loss": 0.3756,
      "num_input_tokens_seen": 52031728,
      "step": 905,
      "train_runtime": 5147.0867,
      "train_tokens_per_second": 10108.967
    },
    {
      "epoch": 0.20719489981785064,
      "grad_norm": 0.627508622690965,
      "learning_rate": 9.738088622914682e-06,
      "loss": 0.3373,
      "num_input_tokens_seen": 52302696,
      "step": 910,
      "train_runtime": 5176.0878,
      "train_tokens_per_second": 10104.677
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 0.669721767755057,
      "learning_rate": 9.7352251578701e-06,
      "loss": 0.3635,
      "num_input_tokens_seen": 52596432,
      "step": 915,
      "train_runtime": 5205.1936,
      "train_tokens_per_second": 10104.606
    },
    {
      "epoch": 0.20947176684881602,
      "grad_norm": 0.8381451610896502,
      "learning_rate": 9.732346550393994e-06,
      "loss": 0.339,
      "num_input_tokens_seen": 52878264,
      "step": 920,
      "train_runtime": 5234.647,
      "train_tokens_per_second": 10101.591
    },
    {
      "epoch": 0.21061020036429873,
      "grad_norm": 0.676393352836907,
      "learning_rate": 9.729452809691652e-06,
      "loss": 0.3808,
      "num_input_tokens_seen": 53156072,
      "step": 925,
      "train_runtime": 5258.5311,
      "train_tokens_per_second": 10108.54
    },
    {
      "epoch": 0.21174863387978143,
      "grad_norm": 0.6690685781170483,
      "learning_rate": 9.726543945016755e-06,
      "loss": 0.3268,
      "num_input_tokens_seen": 53461072,
      "step": 930,
      "train_runtime": 5287.29,
      "train_tokens_per_second": 10111.243
    },
    {
      "epoch": 0.2128870673952641,
      "grad_norm": 0.6852750723991681,
      "learning_rate": 9.72361996567135e-06,
      "loss": 0.3503,
      "num_input_tokens_seen": 53783344,
      "step": 935,
      "train_runtime": 5313.0271,
      "train_tokens_per_second": 10122.919
    },
    {
      "epoch": 0.21402550091074682,
      "grad_norm": 0.7108898109051394,
      "learning_rate": 9.720680881005818e-06,
      "loss": 0.3343,
      "num_input_tokens_seen": 54072600,
      "step": 940,
      "train_runtime": 5341.7588,
      "train_tokens_per_second": 10122.621
    },
    {
      "epoch": 0.2151639344262295,
      "grad_norm": 0.6475628399026653,
      "learning_rate": 9.717726700418842e-06,
      "loss": 0.3254,
      "num_input_tokens_seen": 54346120,
      "step": 945,
      "train_runtime": 5367.1699,
      "train_tokens_per_second": 10125.657
    },
    {
      "epoch": 0.2163023679417122,
      "grad_norm": 0.6494308175720015,
      "learning_rate": 9.714757433357385e-06,
      "loss": 0.3661,
      "num_input_tokens_seen": 54650888,
      "step": 950,
      "train_runtime": 5393.6982,
      "train_tokens_per_second": 10132.359
    },
    {
      "epoch": 0.2174408014571949,
      "grad_norm": 0.6031042383171561,
      "learning_rate": 9.711773089316645e-06,
      "loss": 0.3193,
      "num_input_tokens_seen": 54927528,
      "step": 955,
      "train_runtime": 5423.4514,
      "train_tokens_per_second": 10127.781
    },
    {
      "epoch": 0.2185792349726776,
      "grad_norm": 0.7159178009493417,
      "learning_rate": 9.708773677840043e-06,
      "loss": 0.3505,
      "num_input_tokens_seen": 55217472,
      "step": 960,
      "train_runtime": 5457.1035,
      "train_tokens_per_second": 10118.458
    },
    {
      "epoch": 0.2197176684881603,
      "grad_norm": 0.7592955168444189,
      "learning_rate": 9.705759208519173e-06,
      "loss": 0.3644,
      "num_input_tokens_seen": 55506152,
      "step": 965,
      "train_runtime": 5480.3064,
      "train_tokens_per_second": 10128.293
    },
    {
      "epoch": 0.22085610200364297,
      "grad_norm": 0.6709662411935429,
      "learning_rate": 9.702729690993792e-06,
      "loss": 0.3232,
      "num_input_tokens_seen": 55798408,
      "step": 970,
      "train_runtime": 5507.4838,
      "train_tokens_per_second": 10131.379
    },
    {
      "epoch": 0.22199453551912568,
      "grad_norm": 0.6322761643329508,
      "learning_rate": 9.69968513495177e-06,
      "loss": 0.3007,
      "num_input_tokens_seen": 56095088,
      "step": 975,
      "train_runtime": 5530.0757,
      "train_tokens_per_second": 10143.638
    },
    {
      "epoch": 0.22313296903460839,
      "grad_norm": 0.5951804240983941,
      "learning_rate": 9.696625550129073e-06,
      "loss": 0.3448,
      "num_input_tokens_seen": 56387696,
      "step": 980,
      "train_runtime": 5557.5288,
      "train_tokens_per_second": 10146.182
    },
    {
      "epoch": 0.22427140255009106,
      "grad_norm": 0.6928947905517316,
      "learning_rate": 9.693550946309721e-06,
      "loss": 0.3576,
      "num_input_tokens_seen": 56691104,
      "step": 985,
      "train_runtime": 5585.7774,
      "train_tokens_per_second": 10149.188
    },
    {
      "epoch": 0.22540983606557377,
      "grad_norm": 0.6641647349169133,
      "learning_rate": 9.69046133332577e-06,
      "loss": 0.3343,
      "num_input_tokens_seen": 56988544,
      "step": 990,
      "train_runtime": 5614.5674,
      "train_tokens_per_second": 10150.122
    },
    {
      "epoch": 0.22654826958105648,
      "grad_norm": 0.5921985031091598,
      "learning_rate": 9.687356721057267e-06,
      "loss": 0.3256,
      "num_input_tokens_seen": 57296336,
      "step": 995,
      "train_runtime": 5645.7986,
      "train_tokens_per_second": 10148.491
    },
    {
      "epoch": 0.22768670309653916,
      "grad_norm": 0.6798623062536566,
      "learning_rate": 9.684237119432222e-06,
      "loss": 0.3574,
      "num_input_tokens_seen": 57602088,
      "step": 1000,
      "train_runtime": 5676.8173,
      "train_tokens_per_second": 10146.898
    },
    {
      "epoch": 0.22882513661202186,
      "grad_norm": 0.7002065737914681,
      "learning_rate": 9.681102538426588e-06,
      "loss": 0.3281,
      "num_input_tokens_seen": 57882600,
      "step": 1005,
      "train_runtime": 5702.1133,
      "train_tokens_per_second": 10151.079
    },
    {
      "epoch": 0.22996357012750454,
      "grad_norm": 0.711419787353518,
      "learning_rate": 9.677952988064207e-06,
      "loss": 0.324,
      "num_input_tokens_seen": 58164360,
      "step": 1010,
      "train_runtime": 5724.2842,
      "train_tokens_per_second": 10160.984
    },
    {
      "epoch": 0.23110200364298725,
      "grad_norm": 0.6966540317169728,
      "learning_rate": 9.674788478416799e-06,
      "loss": 0.3531,
      "num_input_tokens_seen": 58453800,
      "step": 1015,
      "train_runtime": 5756.8842,
      "train_tokens_per_second": 10153.722
    },
    {
      "epoch": 0.23224043715846995,
      "grad_norm": 0.6535766918324384,
      "learning_rate": 9.67160901960392e-06,
      "loss": 0.3609,
      "num_input_tokens_seen": 58760792,
      "step": 1020,
      "train_runtime": 5784.0788,
      "train_tokens_per_second": 10159.058
    },
    {
      "epoch": 0.23337887067395263,
      "grad_norm": 0.5522053368289218,
      "learning_rate": 9.668414621792928e-06,
      "loss": 0.3249,
      "num_input_tokens_seen": 59046328,
      "step": 1025,
      "train_runtime": 5812.6442,
      "train_tokens_per_second": 10158.256
    },
    {
      "epoch": 0.23451730418943534,
      "grad_norm": 0.6610078217854053,
      "learning_rate": 9.665205295198957e-06,
      "loss": 0.344,
      "num_input_tokens_seen": 59333320,
      "step": 1030,
      "train_runtime": 5846.2042,
      "train_tokens_per_second": 10149.033
    },
    {
      "epoch": 0.23565573770491804,
      "grad_norm": 0.5958916581388537,
      "learning_rate": 9.661981050084877e-06,
      "loss": 0.3264,
      "num_input_tokens_seen": 59604352,
      "step": 1035,
      "train_runtime": 5872.5533,
      "train_tokens_per_second": 10149.649
    },
    {
      "epoch": 0.23679417122040072,
      "grad_norm": 0.8286063890917794,
      "learning_rate": 9.658741896761266e-06,
      "loss": 0.3523,
      "num_input_tokens_seen": 59860696,
      "step": 1040,
      "train_runtime": 5902.3448,
      "train_tokens_per_second": 10141.85
    },
    {
      "epoch": 0.23793260473588343,
      "grad_norm": 0.6890786680003266,
      "learning_rate": 9.655487845586378e-06,
      "loss": 0.3521,
      "num_input_tokens_seen": 60143288,
      "step": 1045,
      "train_runtime": 5933.4866,
      "train_tokens_per_second": 10136.247
    },
    {
      "epoch": 0.2390710382513661,
      "grad_norm": 0.6987938665951587,
      "learning_rate": 9.652218906966105e-06,
      "loss": 0.3591,
      "num_input_tokens_seen": 60423056,
      "step": 1050,
      "train_runtime": 5961.2252,
      "train_tokens_per_second": 10136.013
    },
    {
      "epoch": 0.24020947176684881,
      "grad_norm": 0.7238796811114204,
      "learning_rate": 9.64893509135395e-06,
      "loss": 0.3278,
      "num_input_tokens_seen": 60700808,
      "step": 1055,
      "train_runtime": 5984.3103,
      "train_tokens_per_second": 10143.326
    },
    {
      "epoch": 0.24134790528233152,
      "grad_norm": 0.5969498589045185,
      "learning_rate": 9.645636409250983e-06,
      "loss": 0.345,
      "num_input_tokens_seen": 60998528,
      "step": 1060,
      "train_runtime": 6014.2792,
      "train_tokens_per_second": 10142.284
    },
    {
      "epoch": 0.2424863387978142,
      "grad_norm": 0.5908036391037592,
      "learning_rate": 9.642322871205825e-06,
      "loss": 0.3518,
      "num_input_tokens_seen": 61288904,
      "step": 1065,
      "train_runtime": 6044.446,
      "train_tokens_per_second": 10139.706
    },
    {
      "epoch": 0.2436247723132969,
      "grad_norm": 0.6201493690049832,
      "learning_rate": 9.638994487814593e-06,
      "loss": 0.3524,
      "num_input_tokens_seen": 61567608,
      "step": 1070,
      "train_runtime": 6076.6058,
      "train_tokens_per_second": 10131.907
    },
    {
      "epoch": 0.2447632058287796,
      "grad_norm": 0.7561259762502598,
      "learning_rate": 9.635651269720884e-06,
      "loss": 0.325,
      "num_input_tokens_seen": 61847400,
      "step": 1075,
      "train_runtime": 6101.4753,
      "train_tokens_per_second": 10136.467
    },
    {
      "epoch": 0.2459016393442623,
      "grad_norm": 0.6734611170663207,
      "learning_rate": 9.63229322761573e-06,
      "loss": 0.3589,
      "num_input_tokens_seen": 62159312,
      "step": 1080,
      "train_runtime": 6133.1863,
      "train_tokens_per_second": 10134.913
    },
    {
      "epoch": 0.247040072859745,
      "grad_norm": 0.8422224261934018,
      "learning_rate": 9.628920372237572e-06,
      "loss": 0.3391,
      "num_input_tokens_seen": 62448112,
      "step": 1085,
      "train_runtime": 6164.2291,
      "train_tokens_per_second": 10130.725
    },
    {
      "epoch": 0.24817850637522768,
      "grad_norm": 0.7459198378314553,
      "learning_rate": 9.625532714372216e-06,
      "loss": 0.3358,
      "num_input_tokens_seen": 62734400,
      "step": 1090,
      "train_runtime": 6191.9179,
      "train_tokens_per_second": 10131.659
    },
    {
      "epoch": 0.24931693989071038,
      "grad_norm": 0.6950539799386127,
      "learning_rate": 9.622130264852809e-06,
      "loss": 0.3595,
      "num_input_tokens_seen": 63039200,
      "step": 1095,
      "train_runtime": 6225.3481,
      "train_tokens_per_second": 10126.213
    },
    {
      "epoch": 0.25045537340619306,
      "grad_norm": 0.6349140710416101,
      "learning_rate": 9.618713034559792e-06,
      "loss": 0.3099,
      "num_input_tokens_seen": 63323576,
      "step": 1100,
      "train_runtime": 6249.4568,
      "train_tokens_per_second": 10132.653
    },
    {
      "epoch": 0.25159380692167577,
      "grad_norm": 0.5834764638631854,
      "learning_rate": 9.615281034420882e-06,
      "loss": 0.3437,
      "num_input_tokens_seen": 63615552,
      "step": 1105,
      "train_runtime": 6277.36,
      "train_tokens_per_second": 10134.125
    },
    {
      "epoch": 0.2527322404371585,
      "grad_norm": 0.7266387496187275,
      "learning_rate": 9.611834275411018e-06,
      "loss": 0.3314,
      "num_input_tokens_seen": 63918752,
      "step": 1110,
      "train_runtime": 6298.504,
      "train_tokens_per_second": 10148.243
    },
    {
      "epoch": 0.2538706739526412,
      "grad_norm": 0.7960820575410902,
      "learning_rate": 9.60837276855234e-06,
      "loss": 0.3156,
      "num_input_tokens_seen": 64200872,
      "step": 1115,
      "train_runtime": 6328.36,
      "train_tokens_per_second": 10144.946
    },
    {
      "epoch": 0.2550091074681239,
      "grad_norm": 0.7604761080439629,
      "learning_rate": 9.604896524914152e-06,
      "loss": 0.3672,
      "num_input_tokens_seen": 64473088,
      "step": 1120,
      "train_runtime": 6355.3061,
      "train_tokens_per_second": 10144.765
    },
    {
      "epoch": 0.25614754098360654,
      "grad_norm": 0.7081441203722075,
      "learning_rate": 9.601405555612874e-06,
      "loss": 0.3195,
      "num_input_tokens_seen": 64745296,
      "step": 1125,
      "train_runtime": 6382.063,
      "train_tokens_per_second": 10144.885
    },
    {
      "epoch": 0.25728597449908924,
      "grad_norm": 0.6672991623564763,
      "learning_rate": 9.59789987181203e-06,
      "loss": 0.3401,
      "num_input_tokens_seen": 65021904,
      "step": 1130,
      "train_runtime": 6410.5898,
      "train_tokens_per_second": 10142.889
    },
    {
      "epoch": 0.25842440801457195,
      "grad_norm": 0.6771286591660082,
      "learning_rate": 9.594379484722185e-06,
      "loss": 0.3178,
      "num_input_tokens_seen": 65307040,
      "step": 1135,
      "train_runtime": 6436.1118,
      "train_tokens_per_second": 10146.971
    },
    {
      "epoch": 0.25956284153005466,
      "grad_norm": 0.9268779341344628,
      "learning_rate": 9.59084440560093e-06,
      "loss": 0.3284,
      "num_input_tokens_seen": 65559736,
      "step": 1140,
      "train_runtime": 6458.3769,
      "train_tokens_per_second": 10151.117
    },
    {
      "epoch": 0.26070127504553736,
      "grad_norm": 0.6533538047039629,
      "learning_rate": 9.587294645752839e-06,
      "loss": 0.3504,
      "num_input_tokens_seen": 65853568,
      "step": 1145,
      "train_runtime": 6488.6993,
      "train_tokens_per_second": 10148.963
    },
    {
      "epoch": 0.26183970856102,
      "grad_norm": 0.5914775247509314,
      "learning_rate": 9.583730216529428e-06,
      "loss": 0.3432,
      "num_input_tokens_seen": 66154104,
      "step": 1150,
      "train_runtime": 6522.2425,
      "train_tokens_per_second": 10142.846
    },
    {
      "epoch": 0.2629781420765027,
      "grad_norm": 0.6234893965090609,
      "learning_rate": 9.580151129329127e-06,
      "loss": 0.3032,
      "num_input_tokens_seen": 66433552,
      "step": 1155,
      "train_runtime": 6547.0686,
      "train_tokens_per_second": 10147.068
    },
    {
      "epoch": 0.2641165755919854,
      "grad_norm": 0.6097690974386082,
      "learning_rate": 9.576557395597237e-06,
      "loss": 0.327,
      "num_input_tokens_seen": 66729616,
      "step": 1160,
      "train_runtime": 6576.8647,
      "train_tokens_per_second": 10146.113
    },
    {
      "epoch": 0.26525500910746813,
      "grad_norm": 0.6903908767208108,
      "learning_rate": 9.572949026825898e-06,
      "loss": 0.3414,
      "num_input_tokens_seen": 67036096,
      "step": 1165,
      "train_runtime": 6606.9702,
      "train_tokens_per_second": 10146.269
    },
    {
      "epoch": 0.26639344262295084,
      "grad_norm": 0.6655944965090947,
      "learning_rate": 9.569326034554048e-06,
      "loss": 0.3325,
      "num_input_tokens_seen": 67333336,
      "step": 1170,
      "train_runtime": 6639.2199,
      "train_tokens_per_second": 10141.754
    },
    {
      "epoch": 0.2675318761384335,
      "grad_norm": 0.679327998149404,
      "learning_rate": 9.565688430367392e-06,
      "loss": 0.3301,
      "num_input_tokens_seen": 67616048,
      "step": 1175,
      "train_runtime": 6670.429,
      "train_tokens_per_second": 10136.687
    },
    {
      "epoch": 0.2686703096539162,
      "grad_norm": 0.5817623463684942,
      "learning_rate": 9.562036225898359e-06,
      "loss": 0.3221,
      "num_input_tokens_seen": 67918224,
      "step": 1180,
      "train_runtime": 6699.1845,
      "train_tokens_per_second": 10138.282
    },
    {
      "epoch": 0.2698087431693989,
      "grad_norm": 0.5972448118077021,
      "learning_rate": 9.558369432826067e-06,
      "loss": 0.3259,
      "num_input_tokens_seen": 68194608,
      "step": 1185,
      "train_runtime": 6726.5188,
      "train_tokens_per_second": 10138.173
    },
    {
      "epoch": 0.2709471766848816,
      "grad_norm": 0.8689834800811271,
      "learning_rate": 9.554688062876288e-06,
      "loss": 0.33,
      "num_input_tokens_seen": 68498288,
      "step": 1190,
      "train_runtime": 6755.2611,
      "train_tokens_per_second": 10139.991
    },
    {
      "epoch": 0.2720856102003643,
      "grad_norm": 0.6586485741581459,
      "learning_rate": 9.550992127821404e-06,
      "loss": 0.3527,
      "num_input_tokens_seen": 68774024,
      "step": 1195,
      "train_runtime": 6781.6271,
      "train_tokens_per_second": 10141.228
    },
    {
      "epoch": 0.273224043715847,
      "grad_norm": 0.6215363907215131,
      "learning_rate": 9.547281639480377e-06,
      "loss": 0.3363,
      "num_input_tokens_seen": 69058848,
      "step": 1200,
      "train_runtime": 6814.8617,
      "train_tokens_per_second": 10133.566
    },
    {
      "epoch": 0.27436247723132967,
      "grad_norm": 0.7156801167789186,
      "learning_rate": 9.54355660971871e-06,
      "loss": 0.3866,
      "num_input_tokens_seen": 69348744,
      "step": 1205,
      "train_runtime": 6846.153,
      "train_tokens_per_second": 10129.593
    },
    {
      "epoch": 0.2755009107468124,
      "grad_norm": 0.7091760654775738,
      "learning_rate": 9.539817050448401e-06,
      "loss": 0.3219,
      "num_input_tokens_seen": 69629384,
      "step": 1210,
      "train_runtime": 6881.2884,
      "train_tokens_per_second": 10118.655
    },
    {
      "epoch": 0.2766393442622951,
      "grad_norm": 0.5156885547490322,
      "learning_rate": 9.536062973627916e-06,
      "loss": 0.3193,
      "num_input_tokens_seen": 69897624,
      "step": 1215,
      "train_runtime": 6913.393,
      "train_tokens_per_second": 10110.466
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 0.7048550855096908,
      "learning_rate": 9.532294391262144e-06,
      "loss": 0.3005,
      "num_input_tokens_seen": 70171928,
      "step": 1220,
      "train_runtime": 6945.0051,
      "train_tokens_per_second": 10103.942
    },
    {
      "epoch": 0.2789162112932605,
      "grad_norm": 0.6733857799303422,
      "learning_rate": 9.528511315402358e-06,
      "loss": 0.3177,
      "num_input_tokens_seen": 70464440,
      "step": 1225,
      "train_runtime": 6970.933,
      "train_tokens_per_second": 10108.323
    },
    {
      "epoch": 0.28005464480874315,
      "grad_norm": 0.6973126771796135,
      "learning_rate": 9.524713758146182e-06,
      "loss": 0.3115,
      "num_input_tokens_seen": 70755592,
      "step": 1230,
      "train_runtime": 6996.2779,
      "train_tokens_per_second": 10113.319
    },
    {
      "epoch": 0.28119307832422585,
      "grad_norm": 0.6369755591241731,
      "learning_rate": 9.52090173163755e-06,
      "loss": 0.3777,
      "num_input_tokens_seen": 71046048,
      "step": 1235,
      "train_runtime": 7025.8044,
      "train_tokens_per_second": 10112.159
    },
    {
      "epoch": 0.28233151183970856,
      "grad_norm": 0.5622822320212815,
      "learning_rate": 9.517075248066659e-06,
      "loss": 0.3264,
      "num_input_tokens_seen": 71340240,
      "step": 1240,
      "train_runtime": 7058.9231,
      "train_tokens_per_second": 10106.391
    },
    {
      "epoch": 0.28346994535519127,
      "grad_norm": 0.6962317541390451,
      "learning_rate": 9.513234319669945e-06,
      "loss": 0.3492,
      "num_input_tokens_seen": 71615184,
      "step": 1245,
      "train_runtime": 7090.4978,
      "train_tokens_per_second": 10100.163
    },
    {
      "epoch": 0.284608378870674,
      "grad_norm": 0.5861030945951193,
      "learning_rate": 9.509378958730034e-06,
      "loss": 0.354,
      "num_input_tokens_seen": 71905928,
      "step": 1250,
      "train_runtime": 7120.0551,
      "train_tokens_per_second": 10099.069
    },
    {
      "epoch": 0.2857468123861566,
      "grad_norm": 0.6163306303365098,
      "learning_rate": 9.5055091775757e-06,
      "loss": 0.3497,
      "num_input_tokens_seen": 72208368,
      "step": 1255,
      "train_runtime": 7149.8653,
      "train_tokens_per_second": 10099.263
    },
    {
      "epoch": 0.28688524590163933,
      "grad_norm": 0.6487109140901113,
      "learning_rate": 9.501624988581837e-06,
      "loss": 0.3434,
      "num_input_tokens_seen": 72505744,
      "step": 1260,
      "train_runtime": 7181.6157,
      "train_tokens_per_second": 10096.021
    },
    {
      "epoch": 0.28802367941712204,
      "grad_norm": 0.6603464868133113,
      "learning_rate": 9.497726404169411e-06,
      "loss": 0.3025,
      "num_input_tokens_seen": 72792368,
      "step": 1265,
      "train_runtime": 7201.149,
      "train_tokens_per_second": 10108.438
    },
    {
      "epoch": 0.28916211293260474,
      "grad_norm": 4.990032169689239,
      "learning_rate": 9.493813436805419e-06,
      "loss": 0.354,
      "num_input_tokens_seen": 73084784,
      "step": 1270,
      "train_runtime": 7231.9475,
      "train_tokens_per_second": 10105.823
    },
    {
      "epoch": 0.29030054644808745,
      "grad_norm": 0.673825850960624,
      "learning_rate": 9.489886099002855e-06,
      "loss": 0.3703,
      "num_input_tokens_seen": 73376736,
      "step": 1275,
      "train_runtime": 7262.6192,
      "train_tokens_per_second": 10103.343
    },
    {
      "epoch": 0.29143897996357016,
      "grad_norm": 0.7102758834971896,
      "learning_rate": 9.485944403320666e-06,
      "loss": 0.3351,
      "num_input_tokens_seen": 73672488,
      "step": 1280,
      "train_runtime": 7287.9405,
      "train_tokens_per_second": 10108.821
    },
    {
      "epoch": 0.2925774134790528,
      "grad_norm": 0.7426023994716382,
      "learning_rate": 9.481988362363713e-06,
      "loss": 0.3756,
      "num_input_tokens_seen": 73964664,
      "step": 1285,
      "train_runtime": 7317.8586,
      "train_tokens_per_second": 10107.419
    },
    {
      "epoch": 0.2937158469945355,
      "grad_norm": 0.7317604890558955,
      "learning_rate": 9.478017988782733e-06,
      "loss": 0.3379,
      "num_input_tokens_seen": 74239592,
      "step": 1290,
      "train_runtime": 7341.8122,
      "train_tokens_per_second": 10111.889
    },
    {
      "epoch": 0.2948542805100182,
      "grad_norm": 0.7015592660676804,
      "learning_rate": 9.474033295274293e-06,
      "loss": 0.3421,
      "num_input_tokens_seen": 74527184,
      "step": 1295,
      "train_runtime": 7365.9078,
      "train_tokens_per_second": 10117.855
    },
    {
      "epoch": 0.2959927140255009,
      "grad_norm": 0.6195244926602753,
      "learning_rate": 9.470034294580753e-06,
      "loss": 0.3347,
      "num_input_tokens_seen": 74832920,
      "step": 1300,
      "train_runtime": 7395.7307,
      "train_tokens_per_second": 10118.394
    },
    {
      "epoch": 0.29713114754098363,
      "grad_norm": 0.7202488902508078,
      "learning_rate": 9.466020999490228e-06,
      "loss": 0.3406,
      "num_input_tokens_seen": 75129304,
      "step": 1305,
      "train_runtime": 7428.8377,
      "train_tokens_per_second": 10113.198
    },
    {
      "epoch": 0.2982695810564663,
      "grad_norm": 0.7709543941244598,
      "learning_rate": 9.461993422836541e-06,
      "loss": 0.3389,
      "num_input_tokens_seen": 75424960,
      "step": 1310,
      "train_runtime": 7462.6982,
      "train_tokens_per_second": 10106.929
    },
    {
      "epoch": 0.299408014571949,
      "grad_norm": 0.5928988862420836,
      "learning_rate": 9.457951577499187e-06,
      "loss": 0.3377,
      "num_input_tokens_seen": 75700736,
      "step": 1315,
      "train_runtime": 7497.2056,
      "train_tokens_per_second": 10097.194
    },
    {
      "epoch": 0.3005464480874317,
      "grad_norm": 0.7540116092421518,
      "learning_rate": 9.453895476403287e-06,
      "loss": 0.3689,
      "num_input_tokens_seen": 75968280,
      "step": 1320,
      "train_runtime": 7527.4348,
      "train_tokens_per_second": 10092.187
    },
    {
      "epoch": 0.3016848816029144,
      "grad_norm": 0.6408773841469751,
      "learning_rate": 9.449825132519552e-06,
      "loss": 0.3155,
      "num_input_tokens_seen": 76268768,
      "step": 1325,
      "train_runtime": 7550.7243,
      "train_tokens_per_second": 10100.855
    },
    {
      "epoch": 0.3028233151183971,
      "grad_norm": 0.7045139387255736,
      "learning_rate": 9.445740558864239e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 76565864,
      "step": 1330,
      "train_runtime": 7578.4206,
      "train_tokens_per_second": 10103.143
    },
    {
      "epoch": 0.30396174863387976,
      "grad_norm": 0.6033794688114738,
      "learning_rate": 9.441641768499108e-06,
      "loss": 0.2682,
      "num_input_tokens_seen": 76852288,
      "step": 1335,
      "train_runtime": 7607.2803,
      "train_tokens_per_second": 10102.466
    },
    {
      "epoch": 0.30510018214936246,
      "grad_norm": 0.6243057173325028,
      "learning_rate": 9.43752877453138e-06,
      "loss": 0.3447,
      "num_input_tokens_seen": 77144488,
      "step": 1340,
      "train_runtime": 7636.5917,
      "train_tokens_per_second": 10101.953
    },
    {
      "epoch": 0.30623861566484517,
      "grad_norm": 0.6562622019046708,
      "learning_rate": 9.433401590113702e-06,
      "loss": 0.3699,
      "num_input_tokens_seen": 77414568,
      "step": 1345,
      "train_runtime": 7670.0348,
      "train_tokens_per_second": 10093.118
    },
    {
      "epoch": 0.3073770491803279,
      "grad_norm": 0.5842193787962752,
      "learning_rate": 9.429260228444093e-06,
      "loss": 0.3259,
      "num_input_tokens_seen": 77708904,
      "step": 1350,
      "train_runtime": 7703.4528,
      "train_tokens_per_second": 10087.542
    },
    {
      "epoch": 0.3085154826958106,
      "grad_norm": 0.5865620104789527,
      "learning_rate": 9.425104702765914e-06,
      "loss": 0.3072,
      "num_input_tokens_seen": 78001168,
      "step": 1355,
      "train_runtime": 7734.2973,
      "train_tokens_per_second": 10085.101
    },
    {
      "epoch": 0.30965391621129323,
      "grad_norm": 0.5422592913575153,
      "learning_rate": 9.42093502636782e-06,
      "loss": 0.2942,
      "num_input_tokens_seen": 78302152,
      "step": 1360,
      "train_runtime": 7754.6762,
      "train_tokens_per_second": 10097.411
    },
    {
      "epoch": 0.31079234972677594,
      "grad_norm": 0.6335458411917911,
      "learning_rate": 9.41675121258371e-06,
      "loss": 0.3261,
      "num_input_tokens_seen": 78579616,
      "step": 1365,
      "train_runtime": 7784.9439,
      "train_tokens_per_second": 10093.794
    },
    {
      "epoch": 0.31193078324225865,
      "grad_norm": 0.5770929955309289,
      "learning_rate": 9.4125532747927e-06,
      "loss": 0.3245,
      "num_input_tokens_seen": 78856888,
      "step": 1370,
      "train_runtime": 7819.1403,
      "train_tokens_per_second": 10085.11
    },
    {
      "epoch": 0.31306921675774135,
      "grad_norm": 0.6760930017129636,
      "learning_rate": 9.40834122641907e-06,
      "loss": 0.351,
      "num_input_tokens_seen": 79134568,
      "step": 1375,
      "train_runtime": 7839.8153,
      "train_tokens_per_second": 10093.933
    },
    {
      "epoch": 0.31420765027322406,
      "grad_norm": 0.6495154846387725,
      "learning_rate": 9.40411508093222e-06,
      "loss": 0.3553,
      "num_input_tokens_seen": 79421328,
      "step": 1380,
      "train_runtime": 7875.6275,
      "train_tokens_per_second": 10084.444
    },
    {
      "epoch": 0.31534608378870677,
      "grad_norm": 0.6246774831210908,
      "learning_rate": 9.399874851846637e-06,
      "loss": 0.3459,
      "num_input_tokens_seen": 79717208,
      "step": 1385,
      "train_runtime": 7912.8089,
      "train_tokens_per_second": 10074.451
    },
    {
      "epoch": 0.3164845173041894,
      "grad_norm": 0.6415175973513221,
      "learning_rate": 9.395620552721838e-06,
      "loss": 0.3531,
      "num_input_tokens_seen": 80005112,
      "step": 1390,
      "train_runtime": 7938.7169,
      "train_tokens_per_second": 10077.839
    },
    {
      "epoch": 0.3176229508196721,
      "grad_norm": 0.6339501849578959,
      "learning_rate": 9.391352197162336e-06,
      "loss": 0.349,
      "num_input_tokens_seen": 80285560,
      "step": 1395,
      "train_runtime": 7963.8161,
      "train_tokens_per_second": 10081.293
    },
    {
      "epoch": 0.31876138433515483,
      "grad_norm": 0.6582489258718245,
      "learning_rate": 9.387069798817596e-06,
      "loss": 0.3488,
      "num_input_tokens_seen": 80593048,
      "step": 1400,
      "train_runtime": 7993.7007,
      "train_tokens_per_second": 10082.07
    },
    {
      "epoch": 0.31989981785063754,
      "grad_norm": 0.7016541524460496,
      "learning_rate": 9.382773371381986e-06,
      "loss": 0.3459,
      "num_input_tokens_seen": 80878368,
      "step": 1405,
      "train_runtime": 8023.2638,
      "train_tokens_per_second": 10080.482
    },
    {
      "epoch": 0.32103825136612024,
      "grad_norm": 0.7050782858229081,
      "learning_rate": 9.378462928594739e-06,
      "loss": 0.3544,
      "num_input_tokens_seen": 81170512,
      "step": 1410,
      "train_runtime": 8050.5757,
      "train_tokens_per_second": 10082.572
    },
    {
      "epoch": 0.3221766848816029,
      "grad_norm": 0.7460136482334833,
      "learning_rate": 9.374138484239903e-06,
      "loss": 0.3652,
      "num_input_tokens_seen": 81459704,
      "step": 1415,
      "train_runtime": 8077.1798,
      "train_tokens_per_second": 10085.167
    },
    {
      "epoch": 0.3233151183970856,
      "grad_norm": 0.7576130394036666,
      "learning_rate": 9.369800052146309e-06,
      "loss": 0.3114,
      "num_input_tokens_seen": 81739472,
      "step": 1420,
      "train_runtime": 8105.7104,
      "train_tokens_per_second": 10084.184
    },
    {
      "epoch": 0.3244535519125683,
      "grad_norm": 0.6464169676727298,
      "learning_rate": 9.365447646187509e-06,
      "loss": 0.3254,
      "num_input_tokens_seen": 82041312,
      "step": 1425,
      "train_runtime": 8131.4895,
      "train_tokens_per_second": 10089.334
    },
    {
      "epoch": 0.325591985428051,
      "grad_norm": 0.683806939024785,
      "learning_rate": 9.361081280281746e-06,
      "loss": 0.3503,
      "num_input_tokens_seen": 82338952,
      "step": 1430,
      "train_runtime": 8157.9377,
      "train_tokens_per_second": 10093.109
    },
    {
      "epoch": 0.3267304189435337,
      "grad_norm": 0.6691718611421726,
      "learning_rate": 9.356700968391904e-06,
      "loss": 0.3157,
      "num_input_tokens_seen": 82625208,
      "step": 1435,
      "train_runtime": 8180.8819,
      "train_tokens_per_second": 10099.792
    },
    {
      "epoch": 0.32786885245901637,
      "grad_norm": 0.6327472338466507,
      "learning_rate": 9.352306724525462e-06,
      "loss": 0.3208,
      "num_input_tokens_seen": 82927520,
      "step": 1440,
      "train_runtime": 8211.5766,
      "train_tokens_per_second": 10098.855
    },
    {
      "epoch": 0.3290072859744991,
      "grad_norm": 0.6617807220447115,
      "learning_rate": 9.347898562734453e-06,
      "loss": 0.329,
      "num_input_tokens_seen": 83224192,
      "step": 1445,
      "train_runtime": 8243.2232,
      "train_tokens_per_second": 10096.074
    },
    {
      "epoch": 0.3301457194899818,
      "grad_norm": 0.5811117867168573,
      "learning_rate": 9.343476497115416e-06,
      "loss": 0.3641,
      "num_input_tokens_seen": 83522024,
      "step": 1450,
      "train_runtime": 8271.2529,
      "train_tokens_per_second": 10097.868
    },
    {
      "epoch": 0.3312841530054645,
      "grad_norm": 0.652972084954556,
      "learning_rate": 9.339040541809354e-06,
      "loss": 0.3434,
      "num_input_tokens_seen": 83801664,
      "step": 1455,
      "train_runtime": 8304.3496,
      "train_tokens_per_second": 10091.298
    },
    {
      "epoch": 0.3324225865209472,
      "grad_norm": 0.6516833702718732,
      "learning_rate": 9.334590711001683e-06,
      "loss": 0.3419,
      "num_input_tokens_seen": 84084536,
      "step": 1460,
      "train_runtime": 8327.9847,
      "train_tokens_per_second": 10096.625
    },
    {
      "epoch": 0.33356102003642984,
      "grad_norm": 0.6135308505791475,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.3256,
      "num_input_tokens_seen": 84378104,
      "step": 1465,
      "train_runtime": 8354.4373,
      "train_tokens_per_second": 10099.795
    },
    {
      "epoch": 0.33469945355191255,
      "grad_norm": 0.585032330243435,
      "learning_rate": 9.325649479845003e-06,
      "loss": 0.3427,
      "num_input_tokens_seen": 84690776,
      "step": 1470,
      "train_runtime": 8388.66,
      "train_tokens_per_second": 10095.865
    },
    {
      "epoch": 0.33583788706739526,
      "grad_norm": 0.6671825607234503,
      "learning_rate": 9.321158108088504e-06,
      "loss": 0.3276,
      "num_input_tokens_seen": 84974296,
      "step": 1475,
      "train_runtime": 8421.2805,
      "train_tokens_per_second": 10090.425
    },
    {
      "epoch": 0.33697632058287796,
      "grad_norm": 0.7838521695481813,
      "learning_rate": 9.31665291801533e-06,
      "loss": 0.359,
      "num_input_tokens_seen": 85261840,
      "step": 1480,
      "train_runtime": 8447.808,
      "train_tokens_per_second": 10092.777
    },
    {
      "epoch": 0.33811475409836067,
      "grad_norm": 0.7420939582116096,
      "learning_rate": 9.3121339240323e-06,
      "loss": 0.3417,
      "num_input_tokens_seen": 85559272,
      "step": 1485,
      "train_runtime": 8470.9271,
      "train_tokens_per_second": 10100.343
    },
    {
      "epoch": 0.3392531876138434,
      "grad_norm": 1.7090657374261298,
      "learning_rate": 9.307601140590373e-06,
      "loss": 0.3408,
      "num_input_tokens_seen": 85871312,
      "step": 1490,
      "train_runtime": 8494.4611,
      "train_tokens_per_second": 10109.095
    },
    {
      "epoch": 0.340391621129326,
      "grad_norm": 0.580159020019491,
      "learning_rate": 9.30305458218461e-06,
      "loss": 0.3328,
      "num_input_tokens_seen": 86177648,
      "step": 1495,
      "train_runtime": 8524.917,
      "train_tokens_per_second": 10108.913
    },
    {
      "epoch": 0.34153005464480873,
      "grad_norm": 0.7161777250590846,
      "learning_rate": 9.298494263354117e-06,
      "loss": 0.3171,
      "num_input_tokens_seen": 86457064,
      "step": 1500,
      "train_runtime": 8553.167,
      "train_tokens_per_second": 10108.193
    },
    {
      "epoch": 0.34266848816029144,
      "grad_norm": 0.6768828341900396,
      "learning_rate": 9.293920198682004e-06,
      "loss": 0.3336,
      "num_input_tokens_seen": 86774464,
      "step": 1505,
      "train_runtime": 8579.388,
      "train_tokens_per_second": 10114.295
    },
    {
      "epoch": 0.34380692167577415,
      "grad_norm": 0.5839382112340825,
      "learning_rate": 9.28933240279534e-06,
      "loss": 0.2895,
      "num_input_tokens_seen": 87066768,
      "step": 1510,
      "train_runtime": 8606.7703,
      "train_tokens_per_second": 10116.079
    },
    {
      "epoch": 0.34494535519125685,
      "grad_norm": 0.7002586866255833,
      "learning_rate": 9.284730890365108e-06,
      "loss": 0.3384,
      "num_input_tokens_seen": 87355504,
      "step": 1515,
      "train_runtime": 8630.501,
      "train_tokens_per_second": 10121.719
    },
    {
      "epoch": 0.3460837887067395,
      "grad_norm": 0.6086572055567179,
      "learning_rate": 9.280115676106144e-06,
      "loss": 0.3297,
      "num_input_tokens_seen": 87641728,
      "step": 1520,
      "train_runtime": 8662.8844,
      "train_tokens_per_second": 10116.922
    },
    {
      "epoch": 0.3472222222222222,
      "grad_norm": 0.651220616818131,
      "learning_rate": 9.275486774777106e-06,
      "loss": 0.3331,
      "num_input_tokens_seen": 87932944,
      "step": 1525,
      "train_runtime": 8688.3341,
      "train_tokens_per_second": 10120.806
    },
    {
      "epoch": 0.3483606557377049,
      "grad_norm": 0.6381736935041677,
      "learning_rate": 9.270844201180423e-06,
      "loss": 0.3253,
      "num_input_tokens_seen": 88208712,
      "step": 1530,
      "train_runtime": 8714.3107,
      "train_tokens_per_second": 10122.282
    },
    {
      "epoch": 0.3494990892531876,
      "grad_norm": 0.7312889502880634,
      "learning_rate": 9.26618797016224e-06,
      "loss": 0.3226,
      "num_input_tokens_seen": 88486384,
      "step": 1535,
      "train_runtime": 8743.0722,
      "train_tokens_per_second": 10120.743
    },
    {
      "epoch": 0.35063752276867033,
      "grad_norm": 0.5976741417836186,
      "learning_rate": 9.261518096612378e-06,
      "loss": 0.3229,
      "num_input_tokens_seen": 88780848,
      "step": 1540,
      "train_runtime": 8770.5208,
      "train_tokens_per_second": 10122.643
    },
    {
      "epoch": 0.351775956284153,
      "grad_norm": 0.7188969405474744,
      "learning_rate": 9.25683459546429e-06,
      "loss": 0.3506,
      "num_input_tokens_seen": 89059504,
      "step": 1545,
      "train_runtime": 8800.1228,
      "train_tokens_per_second": 10120.257
    },
    {
      "epoch": 0.3529143897996357,
      "grad_norm": 0.766128807737002,
      "learning_rate": 9.252137481694998e-06,
      "loss": 0.3786,
      "num_input_tokens_seen": 89323736,
      "step": 1550,
      "train_runtime": 8823.8915,
      "train_tokens_per_second": 10122.941
    },
    {
      "epoch": 0.3540528233151184,
      "grad_norm": 0.6843157031133122,
      "learning_rate": 9.247426770325064e-06,
      "loss": 0.3292,
      "num_input_tokens_seen": 89626048,
      "step": 1555,
      "train_runtime": 8851.1478,
      "train_tokens_per_second": 10125.924
    },
    {
      "epoch": 0.3551912568306011,
      "grad_norm": 0.7392629016004882,
      "learning_rate": 9.242702476418525e-06,
      "loss": 0.3488,
      "num_input_tokens_seen": 89901280,
      "step": 1560,
      "train_runtime": 8884.2923,
      "train_tokens_per_second": 10119.127
    },
    {
      "epoch": 0.3563296903460838,
      "grad_norm": 0.6147794113970831,
      "learning_rate": 9.237964615082861e-06,
      "loss": 0.3186,
      "num_input_tokens_seen": 90187872,
      "step": 1565,
      "train_runtime": 8917.3623,
      "train_tokens_per_second": 10113.739
    },
    {
      "epoch": 0.3574681238615665,
      "grad_norm": 0.5870179934560074,
      "learning_rate": 9.233213201468928e-06,
      "loss": 0.3547,
      "num_input_tokens_seen": 90485680,
      "step": 1570,
      "train_runtime": 8940.4869,
      "train_tokens_per_second": 10120.89
    },
    {
      "epoch": 0.35860655737704916,
      "grad_norm": 0.6478268653182029,
      "learning_rate": 9.228448250770932e-06,
      "loss": 0.33,
      "num_input_tokens_seen": 90762624,
      "step": 1575,
      "train_runtime": 8974.7882,
      "train_tokens_per_second": 10113.066
    },
    {
      "epoch": 0.35974499089253187,
      "grad_norm": 0.6454429026802917,
      "learning_rate": 9.223669778226353e-06,
      "loss": 0.3398,
      "num_input_tokens_seen": 91047496,
      "step": 1580,
      "train_runtime": 9009.0598,
      "train_tokens_per_second": 10106.215
    },
    {
      "epoch": 0.3608834244080146,
      "grad_norm": 4.1679845156648865,
      "learning_rate": 9.218877799115929e-06,
      "loss": 0.3425,
      "num_input_tokens_seen": 91344224,
      "step": 1585,
      "train_runtime": 9037.2776,
      "train_tokens_per_second": 10107.493
    },
    {
      "epoch": 0.3620218579234973,
      "grad_norm": 0.7470598541765979,
      "learning_rate": 9.214072328763574e-06,
      "loss": 0.3504,
      "num_input_tokens_seen": 91631000,
      "step": 1590,
      "train_runtime": 9066.3593,
      "train_tokens_per_second": 10106.703
    },
    {
      "epoch": 0.36316029143898,
      "grad_norm": 0.5871500254077562,
      "learning_rate": 9.209253382536359e-06,
      "loss": 0.3206,
      "num_input_tokens_seen": 91932720,
      "step": 1595,
      "train_runtime": 9086.2413,
      "train_tokens_per_second": 10117.794
    },
    {
      "epoch": 0.36429872495446264,
      "grad_norm": 0.6811896358720889,
      "learning_rate": 9.204420975844432e-06,
      "loss": 0.328,
      "num_input_tokens_seen": 92237208,
      "step": 1600,
      "train_runtime": 9114.0835,
      "train_tokens_per_second": 10120.294
    },
    {
      "epoch": 0.36543715846994534,
      "grad_norm": 0.6598213989508094,
      "learning_rate": 9.199575124141e-06,
      "loss": 0.318,
      "num_input_tokens_seen": 92529648,
      "step": 1605,
      "train_runtime": 9140.6915,
      "train_tokens_per_second": 10122.828
    },
    {
      "epoch": 0.36657559198542805,
      "grad_norm": 1.3352870810230588,
      "learning_rate": 9.194715842922257e-06,
      "loss": 0.2893,
      "num_input_tokens_seen": 92817992,
      "step": 1610,
      "train_runtime": 9165.8083,
      "train_tokens_per_second": 10126.547
    },
    {
      "epoch": 0.36771402550091076,
      "grad_norm": 0.6370231614039147,
      "learning_rate": 9.189843147727346e-06,
      "loss": 0.345,
      "num_input_tokens_seen": 93105072,
      "step": 1615,
      "train_runtime": 9194.9944,
      "train_tokens_per_second": 10125.626
    },
    {
      "epoch": 0.36885245901639346,
      "grad_norm": 0.6583144910366853,
      "learning_rate": 9.184957054138304e-06,
      "loss": 0.3446,
      "num_input_tokens_seen": 93378592,
      "step": 1620,
      "train_runtime": 9228.6134,
      "train_tokens_per_second": 10118.377
    },
    {
      "epoch": 0.3699908925318761,
      "grad_norm": 0.6168757349658985,
      "learning_rate": 9.180057577780012e-06,
      "loss": 0.3272,
      "num_input_tokens_seen": 93687288,
      "step": 1625,
      "train_runtime": 9258.3195,
      "train_tokens_per_second": 10119.254
    },
    {
      "epoch": 0.3711293260473588,
      "grad_norm": 0.7164154445481559,
      "learning_rate": 9.17514473432015e-06,
      "loss": 0.3419,
      "num_input_tokens_seen": 93977752,
      "step": 1630,
      "train_runtime": 9285.0616,
      "train_tokens_per_second": 10121.392
    },
    {
      "epoch": 0.3722677595628415,
      "grad_norm": 0.5822237977357027,
      "learning_rate": 9.17021853946914e-06,
      "loss": 0.3194,
      "num_input_tokens_seen": 94263200,
      "step": 1635,
      "train_runtime": 9320.0858,
      "train_tokens_per_second": 10113.984
    },
    {
      "epoch": 0.37340619307832423,
      "grad_norm": 0.651871030175223,
      "learning_rate": 9.165279008980104e-06,
      "loss": 0.2932,
      "num_input_tokens_seen": 94539792,
      "step": 1640,
      "train_runtime": 9348.2385,
      "train_tokens_per_second": 10113.113
    },
    {
      "epoch": 0.37454462659380694,
      "grad_norm": 0.49425718396709856,
      "learning_rate": 9.160326158648804e-06,
      "loss": 0.2977,
      "num_input_tokens_seen": 94827624,
      "step": 1645,
      "train_runtime": 9376.2807,
      "train_tokens_per_second": 10113.565
    },
    {
      "epoch": 0.3756830601092896,
      "grad_norm": 0.6833463037509703,
      "learning_rate": 9.155360004313604e-06,
      "loss": 0.3379,
      "num_input_tokens_seen": 95110664,
      "step": 1650,
      "train_runtime": 9409.7087,
      "train_tokens_per_second": 10107.716
    },
    {
      "epoch": 0.3768214936247723,
      "grad_norm": 0.6951944980452301,
      "learning_rate": 9.150380561855402e-06,
      "loss": 0.3416,
      "num_input_tokens_seen": 95404280,
      "step": 1655,
      "train_runtime": 9429.9306,
      "train_tokens_per_second": 10117.177
    },
    {
      "epoch": 0.377959927140255,
      "grad_norm": 0.5933806958056198,
      "learning_rate": 9.145387847197597e-06,
      "loss": 0.329,
      "num_input_tokens_seen": 95695216,
      "step": 1660,
      "train_runtime": 9459.024,
      "train_tokens_per_second": 10116.817
    },
    {
      "epoch": 0.3790983606557377,
      "grad_norm": 0.6700621844825558,
      "learning_rate": 9.140381876306026e-06,
      "loss": 0.352,
      "num_input_tokens_seen": 95954584,
      "step": 1665,
      "train_runtime": 9483.1317,
      "train_tokens_per_second": 10118.449
    },
    {
      "epoch": 0.3802367941712204,
      "grad_norm": 0.5782654189838977,
      "learning_rate": 9.135362665188919e-06,
      "loss": 0.2966,
      "num_input_tokens_seen": 96228112,
      "step": 1670,
      "train_runtime": 9510.903,
      "train_tokens_per_second": 10117.663
    },
    {
      "epoch": 0.3813752276867031,
      "grad_norm": 0.6227113543553889,
      "learning_rate": 9.130330229896846e-06,
      "loss": 0.3223,
      "num_input_tokens_seen": 96527568,
      "step": 1675,
      "train_runtime": 9530.0097,
      "train_tokens_per_second": 10128.801
    },
    {
      "epoch": 0.3825136612021858,
      "grad_norm": 0.7487922010344855,
      "learning_rate": 9.125284586522667e-06,
      "loss": 0.3226,
      "num_input_tokens_seen": 96816648,
      "step": 1680,
      "train_runtime": 9555.0578,
      "train_tokens_per_second": 10132.503
    },
    {
      "epoch": 0.3836520947176685,
      "grad_norm": 0.5752625653206056,
      "learning_rate": 9.120225751201472e-06,
      "loss": 0.3216,
      "num_input_tokens_seen": 97113656,
      "step": 1685,
      "train_runtime": 9583.5371,
      "train_tokens_per_second": 10133.383
    },
    {
      "epoch": 0.3847905282331512,
      "grad_norm": 0.7645587560252761,
      "learning_rate": 9.115153740110547e-06,
      "loss": 0.3611,
      "num_input_tokens_seen": 97385496,
      "step": 1690,
      "train_runtime": 9608.5202,
      "train_tokens_per_second": 10135.327
    },
    {
      "epoch": 0.3859289617486339,
      "grad_norm": 0.6841663039222732,
      "learning_rate": 9.110068569469305e-06,
      "loss": 0.3706,
      "num_input_tokens_seen": 97648320,
      "step": 1695,
      "train_runtime": 9640.2771,
      "train_tokens_per_second": 10129.203
    },
    {
      "epoch": 0.3870673952641166,
      "grad_norm": 0.6497406059778047,
      "learning_rate": 9.104970255539243e-06,
      "loss": 0.3222,
      "num_input_tokens_seen": 97941136,
      "step": 1700,
      "train_runtime": 9667.8929,
      "train_tokens_per_second": 10130.557
    },
    {
      "epoch": 0.38820582877959925,
      "grad_norm": 0.658058101648747,
      "learning_rate": 9.099858814623887e-06,
      "loss": 0.3011,
      "num_input_tokens_seen": 98228680,
      "step": 1705,
      "train_runtime": 9694.3971,
      "train_tokens_per_second": 10132.521
    },
    {
      "epoch": 0.38934426229508196,
      "grad_norm": 0.5967343780209451,
      "learning_rate": 9.09473426306874e-06,
      "loss": 0.3246,
      "num_input_tokens_seen": 98539944,
      "step": 1710,
      "train_runtime": 9725.0962,
      "train_tokens_per_second": 10132.542
    },
    {
      "epoch": 0.39048269581056466,
      "grad_norm": 0.6516885981307624,
      "learning_rate": 9.089596617261236e-06,
      "loss": 0.3545,
      "num_input_tokens_seen": 98844856,
      "step": 1715,
      "train_runtime": 9753.2207,
      "train_tokens_per_second": 10134.586
    },
    {
      "epoch": 0.39162112932604737,
      "grad_norm": 0.6977693306736401,
      "learning_rate": 9.084445893630676e-06,
      "loss": 0.3169,
      "num_input_tokens_seen": 99121968,
      "step": 1720,
      "train_runtime": 9781.8844,
      "train_tokens_per_second": 10133.218
    },
    {
      "epoch": 0.3927595628415301,
      "grad_norm": 0.6208440980787817,
      "learning_rate": 9.079282108648181e-06,
      "loss": 0.3178,
      "num_input_tokens_seen": 99428224,
      "step": 1725,
      "train_runtime": 9812.282,
      "train_tokens_per_second": 10133.038
    },
    {
      "epoch": 0.3938979963570127,
      "grad_norm": 0.578556621143594,
      "learning_rate": 9.074105278826648e-06,
      "loss": 0.3367,
      "num_input_tokens_seen": 99722856,
      "step": 1730,
      "train_runtime": 9839.2373,
      "train_tokens_per_second": 10135.222
    },
    {
      "epoch": 0.39503642987249543,
      "grad_norm": 0.6064624973078327,
      "learning_rate": 9.06891542072068e-06,
      "loss": 0.358,
      "num_input_tokens_seen": 99988384,
      "step": 1735,
      "train_runtime": 9865.0907,
      "train_tokens_per_second": 10135.577
    },
    {
      "epoch": 0.39617486338797814,
      "grad_norm": 0.6046766575527213,
      "learning_rate": 9.06371255092655e-06,
      "loss": 0.3387,
      "num_input_tokens_seen": 100265640,
      "step": 1740,
      "train_runtime": 9892.8272,
      "train_tokens_per_second": 10135.186
    },
    {
      "epoch": 0.39731329690346084,
      "grad_norm": 0.623737949359369,
      "learning_rate": 9.058496686082132e-06,
      "loss": 0.349,
      "num_input_tokens_seen": 100547792,
      "step": 1745,
      "train_runtime": 9923.5269,
      "train_tokens_per_second": 10132.264
    },
    {
      "epoch": 0.39845173041894355,
      "grad_norm": 0.6461555383017957,
      "learning_rate": 9.053267842866861e-06,
      "loss": 0.3529,
      "num_input_tokens_seen": 100839528,
      "step": 1750,
      "train_runtime": 9955.6823,
      "train_tokens_per_second": 10128.841
    },
    {
      "epoch": 0.39959016393442626,
      "grad_norm": 0.5806469523070346,
      "learning_rate": 9.048026038001679e-06,
      "loss": 0.3017,
      "num_input_tokens_seen": 101114600,
      "step": 1755,
      "train_runtime": 9986.9535,
      "train_tokens_per_second": 10124.669
    },
    {
      "epoch": 0.4007285974499089,
      "grad_norm": 0.6818285407410173,
      "learning_rate": 9.042771288248964e-06,
      "loss": 0.3203,
      "num_input_tokens_seen": 101400920,
      "step": 1760,
      "train_runtime": 10009.241,
      "train_tokens_per_second": 10130.73
    },
    {
      "epoch": 0.4018670309653916,
      "grad_norm": 0.6427801433570235,
      "learning_rate": 9.037503610412502e-06,
      "loss": 0.3524,
      "num_input_tokens_seen": 101697568,
      "step": 1765,
      "train_runtime": 10033.5495,
      "train_tokens_per_second": 10135.752
    },
    {
      "epoch": 0.4030054644808743,
      "grad_norm": 0.6579298779030517,
      "learning_rate": 9.032223021337415e-06,
      "loss": 0.3597,
      "num_input_tokens_seen": 101979536,
      "step": 1770,
      "train_runtime": 10058.7869,
      "train_tokens_per_second": 10138.353
    },
    {
      "epoch": 0.404143897996357,
      "grad_norm": 0.6668459213607965,
      "learning_rate": 9.026929537910112e-06,
      "loss": 0.3714,
      "num_input_tokens_seen": 102272960,
      "step": 1775,
      "train_runtime": 10085.8264,
      "train_tokens_per_second": 10140.266
    },
    {
      "epoch": 0.40528233151183973,
      "grad_norm": 0.7279180790467177,
      "learning_rate": 9.021623177058243e-06,
      "loss": 0.318,
      "num_input_tokens_seen": 102577928,
      "step": 1780,
      "train_runtime": 10115.9111,
      "train_tokens_per_second": 10140.256
    },
    {
      "epoch": 0.4064207650273224,
      "grad_norm": 0.839611100828474,
      "learning_rate": 9.01630395575063e-06,
      "loss": 0.3216,
      "num_input_tokens_seen": 102864648,
      "step": 1785,
      "train_runtime": 10138.4832,
      "train_tokens_per_second": 10145.96
    },
    {
      "epoch": 0.4075591985428051,
      "grad_norm": 0.508705675419614,
      "learning_rate": 9.01097189099722e-06,
      "loss": 0.2981,
      "num_input_tokens_seen": 103167648,
      "step": 1790,
      "train_runtime": 10169.891,
      "train_tokens_per_second": 10144.42
    },
    {
      "epoch": 0.4086976320582878,
      "grad_norm": 0.6461808334943038,
      "learning_rate": 9.00562699984904e-06,
      "loss": 0.3168,
      "num_input_tokens_seen": 103465560,
      "step": 1795,
      "train_runtime": 10202.9519,
      "train_tokens_per_second": 10140.748
    },
    {
      "epoch": 0.4098360655737705,
      "grad_norm": 0.7172457569032722,
      "learning_rate": 9.000269299398125e-06,
      "loss": 0.3221,
      "num_input_tokens_seen": 103748456,
      "step": 1800,
      "train_runtime": 10230.5019,
      "train_tokens_per_second": 10141.091
    },
    {
      "epoch": 0.4109744990892532,
      "grad_norm": 0.6082673016654961,
      "learning_rate": 8.994898806777476e-06,
      "loss": 0.3444,
      "num_input_tokens_seen": 104044904,
      "step": 1805,
      "train_runtime": 10259.4456,
      "train_tokens_per_second": 10141.377
    },
    {
      "epoch": 0.41211293260473586,
      "grad_norm": 0.6458188701667111,
      "learning_rate": 8.989515539161e-06,
      "loss": 0.2846,
      "num_input_tokens_seen": 104329456,
      "step": 1810,
      "train_runtime": 10285.5424,
      "train_tokens_per_second": 10143.311
    },
    {
      "epoch": 0.41325136612021857,
      "grad_norm": 0.6285912191832762,
      "learning_rate": 8.984119513763455e-06,
      "loss": 0.3189,
      "num_input_tokens_seen": 104611800,
      "step": 1815,
      "train_runtime": 10309.8505,
      "train_tokens_per_second": 10146.781
    },
    {
      "epoch": 0.4143897996357013,
      "grad_norm": 0.5657694477267762,
      "learning_rate": 8.978710747840397e-06,
      "loss": 0.3284,
      "num_input_tokens_seen": 104906848,
      "step": 1820,
      "train_runtime": 10333.7605,
      "train_tokens_per_second": 10151.856
    },
    {
      "epoch": 0.415528233151184,
      "grad_norm": 0.6216438449445301,
      "learning_rate": 8.973289258688125e-06,
      "loss": 0.2998,
      "num_input_tokens_seen": 105190576,
      "step": 1825,
      "train_runtime": 10361.5833,
      "train_tokens_per_second": 10151.979
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.6970423052967764,
      "learning_rate": 8.967855063643628e-06,
      "loss": 0.3199,
      "num_input_tokens_seen": 105469712,
      "step": 1830,
      "train_runtime": 10392.6184,
      "train_tokens_per_second": 10148.522
    },
    {
      "epoch": 0.41780510018214934,
      "grad_norm": 0.7114256757925131,
      "learning_rate": 8.962408180084518e-06,
      "loss": 0.3101,
      "num_input_tokens_seen": 105754000,
      "step": 1835,
      "train_runtime": 10422.8867,
      "train_tokens_per_second": 10146.325
    },
    {
      "epoch": 0.41894353369763204,
      "grad_norm": 0.5529922293075925,
      "learning_rate": 8.956948625428988e-06,
      "loss": 0.3142,
      "num_input_tokens_seen": 106041944,
      "step": 1840,
      "train_runtime": 10453.5894,
      "train_tokens_per_second": 10144.07
    },
    {
      "epoch": 0.42008196721311475,
      "grad_norm": 0.5669538120431907,
      "learning_rate": 8.95147641713575e-06,
      "loss": 0.3029,
      "num_input_tokens_seen": 106342480,
      "step": 1845,
      "train_runtime": 10479.529,
      "train_tokens_per_second": 10147.639
    },
    {
      "epoch": 0.42122040072859745,
      "grad_norm": 1.7938694084400448,
      "learning_rate": 8.94599157270398e-06,
      "loss": 0.3097,
      "num_input_tokens_seen": 106628616,
      "step": 1850,
      "train_runtime": 10505.3167,
      "train_tokens_per_second": 10149.967
    },
    {
      "epoch": 0.42235883424408016,
      "grad_norm": 0.6432148890395983,
      "learning_rate": 8.940494109673266e-06,
      "loss": 0.3005,
      "num_input_tokens_seen": 106917008,
      "step": 1855,
      "train_runtime": 10538.1076,
      "train_tokens_per_second": 10145.75
    },
    {
      "epoch": 0.42349726775956287,
      "grad_norm": 0.6400855380598857,
      "learning_rate": 8.93498404562354e-06,
      "loss": 0.3166,
      "num_input_tokens_seen": 107214000,
      "step": 1860,
      "train_runtime": 10564.1549,
      "train_tokens_per_second": 10148.848
    },
    {
      "epoch": 0.4246357012750455,
      "grad_norm": 0.541294900784237,
      "learning_rate": 8.929461398175035e-06,
      "loss": 0.3401,
      "num_input_tokens_seen": 107519072,
      "step": 1865,
      "train_runtime": 10593.1139,
      "train_tokens_per_second": 10149.902
    },
    {
      "epoch": 0.4257741347905282,
      "grad_norm": 0.6228443169862793,
      "learning_rate": 8.923926184988224e-06,
      "loss": 0.327,
      "num_input_tokens_seen": 107803080,
      "step": 1870,
      "train_runtime": 10617.6824,
      "train_tokens_per_second": 10153.165
    },
    {
      "epoch": 0.42691256830601093,
      "grad_norm": 0.7837138308732265,
      "learning_rate": 8.918378423763762e-06,
      "loss": 0.3552,
      "num_input_tokens_seen": 108082864,
      "step": 1875,
      "train_runtime": 10648.8783,
      "train_tokens_per_second": 10149.695
    },
    {
      "epoch": 0.42805100182149364,
      "grad_norm": 0.5608464511154588,
      "learning_rate": 8.912818132242429e-06,
      "loss": 0.3186,
      "num_input_tokens_seen": 108384072,
      "step": 1880,
      "train_runtime": 10673.8929,
      "train_tokens_per_second": 10154.128
    },
    {
      "epoch": 0.42918943533697634,
      "grad_norm": 0.6535524499040326,
      "learning_rate": 8.907245328205079e-06,
      "loss": 0.2945,
      "num_input_tokens_seen": 108660264,
      "step": 1885,
      "train_runtime": 10704.5369,
      "train_tokens_per_second": 10150.861
    },
    {
      "epoch": 0.430327868852459,
      "grad_norm": 0.548192256285291,
      "learning_rate": 8.901660029472573e-06,
      "loss": 0.2986,
      "num_input_tokens_seen": 108954000,
      "step": 1890,
      "train_runtime": 10734.8598,
      "train_tokens_per_second": 10149.55
    },
    {
      "epoch": 0.4314663023679417,
      "grad_norm": 0.552254240910733,
      "learning_rate": 8.896062253905733e-06,
      "loss": 0.325,
      "num_input_tokens_seen": 109235752,
      "step": 1895,
      "train_runtime": 10766.6487,
      "train_tokens_per_second": 10145.752
    },
    {
      "epoch": 0.4326047358834244,
      "grad_norm": 0.7134280439869302,
      "learning_rate": 8.890452019405275e-06,
      "loss": 0.3321,
      "num_input_tokens_seen": 109507328,
      "step": 1900,
      "train_runtime": 10800.7459,
      "train_tokens_per_second": 10138.867
    },
    {
      "epoch": 0.4337431693989071,
      "grad_norm": 0.6458401996903107,
      "learning_rate": 8.884829343911763e-06,
      "loss": 0.343,
      "num_input_tokens_seen": 109808744,
      "step": 1905,
      "train_runtime": 10829.1257,
      "train_tokens_per_second": 10140.13
    },
    {
      "epoch": 0.4348816029143898,
      "grad_norm": 0.7084135801433841,
      "learning_rate": 8.879194245405538e-06,
      "loss": 0.3395,
      "num_input_tokens_seen": 110112752,
      "step": 1910,
      "train_runtime": 10863.1674,
      "train_tokens_per_second": 10136.339
    },
    {
      "epoch": 0.43602003642987247,
      "grad_norm": 0.7213615687388539,
      "learning_rate": 8.873546741906672e-06,
      "loss": 0.3514,
      "num_input_tokens_seen": 110394376,
      "step": 1915,
      "train_runtime": 10890.9936,
      "train_tokens_per_second": 10136.3
    },
    {
      "epoch": 0.4371584699453552,
      "grad_norm": 0.633568677659359,
      "learning_rate": 8.867886851474905e-06,
      "loss": 0.3278,
      "num_input_tokens_seen": 110692808,
      "step": 1920,
      "train_runtime": 10919.1727,
      "train_tokens_per_second": 10137.472
    },
    {
      "epoch": 0.4382969034608379,
      "grad_norm": 0.6571446710743096,
      "learning_rate": 8.86221459220959e-06,
      "loss": 0.3288,
      "num_input_tokens_seen": 110959080,
      "step": 1925,
      "train_runtime": 10947.3233,
      "train_tokens_per_second": 10135.727
    },
    {
      "epoch": 0.4394353369763206,
      "grad_norm": 0.6153463837733593,
      "learning_rate": 8.856529982249628e-06,
      "loss": 0.3091,
      "num_input_tokens_seen": 111259368,
      "step": 1930,
      "train_runtime": 10978.2314,
      "train_tokens_per_second": 10134.544
    },
    {
      "epoch": 0.4405737704918033,
      "grad_norm": 0.6140685887515922,
      "learning_rate": 8.850833039773422e-06,
      "loss": 0.3355,
      "num_input_tokens_seen": 111558496,
      "step": 1935,
      "train_runtime": 11007.9858,
      "train_tokens_per_second": 10134.324
    },
    {
      "epoch": 0.44171220400728595,
      "grad_norm": 0.7583552612158686,
      "learning_rate": 8.845123782998807e-06,
      "loss": 0.3679,
      "num_input_tokens_seen": 111831600,
      "step": 1940,
      "train_runtime": 11030.8345,
      "train_tokens_per_second": 10138.091
    },
    {
      "epoch": 0.44285063752276865,
      "grad_norm": 0.6125138004544626,
      "learning_rate": 8.839402230183e-06,
      "loss": 0.3173,
      "num_input_tokens_seen": 112129560,
      "step": 1945,
      "train_runtime": 11055.5693,
      "train_tokens_per_second": 10142.36
    },
    {
      "epoch": 0.44398907103825136,
      "grad_norm": 0.5702043073197857,
      "learning_rate": 8.83366839962254e-06,
      "loss": 0.315,
      "num_input_tokens_seen": 112442104,
      "step": 1950,
      "train_runtime": 11085.7125,
      "train_tokens_per_second": 10142.975
    },
    {
      "epoch": 0.44512750455373407,
      "grad_norm": 0.6583074439582252,
      "learning_rate": 8.827922309653225e-06,
      "loss": 0.3576,
      "num_input_tokens_seen": 112691976,
      "step": 1955,
      "train_runtime": 11108.899,
      "train_tokens_per_second": 10144.298
    },
    {
      "epoch": 0.44626593806921677,
      "grad_norm": 0.5684632215627043,
      "learning_rate": 8.822163978650055e-06,
      "loss": 0.345,
      "num_input_tokens_seen": 112992360,
      "step": 1960,
      "train_runtime": 11143.7688,
      "train_tokens_per_second": 10139.51
    },
    {
      "epoch": 0.4474043715846995,
      "grad_norm": 0.5751967122182151,
      "learning_rate": 8.81639342502718e-06,
      "loss": 0.293,
      "num_input_tokens_seen": 113261712,
      "step": 1965,
      "train_runtime": 11178.5753,
      "train_tokens_per_second": 10132.035
    },
    {
      "epoch": 0.44854280510018213,
      "grad_norm": 0.6234096693592964,
      "learning_rate": 8.810610667237831e-06,
      "loss": 0.3286,
      "num_input_tokens_seen": 113545648,
      "step": 1970,
      "train_runtime": 11210.9396,
      "train_tokens_per_second": 10128.112
    },
    {
      "epoch": 0.44968123861566484,
      "grad_norm": 0.6691897396393492,
      "learning_rate": 8.80481572377427e-06,
      "loss": 0.3369,
      "num_input_tokens_seen": 113813736,
      "step": 1975,
      "train_runtime": 11241.2248,
      "train_tokens_per_second": 10124.674
    },
    {
      "epoch": 0.45081967213114754,
      "grad_norm": 0.8276108814684624,
      "learning_rate": 8.799008613167723e-06,
      "loss": 0.3332,
      "num_input_tokens_seen": 114122488,
      "step": 1980,
      "train_runtime": 11276.5309,
      "train_tokens_per_second": 10120.354
    },
    {
      "epoch": 0.45195810564663025,
      "grad_norm": 0.5856810860434037,
      "learning_rate": 8.793189353988325e-06,
      "loss": 0.3244,
      "num_input_tokens_seen": 114423880,
      "step": 1985,
      "train_runtime": 11307.9592,
      "train_tokens_per_second": 10118.88
    },
    {
      "epoch": 0.45309653916211295,
      "grad_norm": 0.6357173808755625,
      "learning_rate": 8.787357964845065e-06,
      "loss": 0.3436,
      "num_input_tokens_seen": 114707792,
      "step": 1990,
      "train_runtime": 11334.825,
      "train_tokens_per_second": 10119.944
    },
    {
      "epoch": 0.4542349726775956,
      "grad_norm": 0.6352147360048155,
      "learning_rate": 8.78151446438571e-06,
      "loss": 0.3259,
      "num_input_tokens_seen": 114981400,
      "step": 1995,
      "train_runtime": 11357.0138,
      "train_tokens_per_second": 10124.263
    },
    {
      "epoch": 0.4553734061930783,
      "grad_norm": 0.5747172954796154,
      "learning_rate": 8.775658871296773e-06,
      "loss": 0.2938,
      "num_input_tokens_seen": 115271224,
      "step": 2000,
      "train_runtime": 11385.7905,
      "train_tokens_per_second": 10124.13
    },
    {
      "epoch": 0.456511839708561,
      "grad_norm": 0.5362104854383153,
      "learning_rate": 8.769791204303421e-06,
      "loss": 0.2969,
      "num_input_tokens_seen": 115561424,
      "step": 2005,
      "train_runtime": 11415.3209,
      "train_tokens_per_second": 10123.362
    },
    {
      "epoch": 0.4576502732240437,
      "grad_norm": 0.658899772996399,
      "learning_rate": 8.763911482169446e-06,
      "loss": 0.3157,
      "num_input_tokens_seen": 115855536,
      "step": 2010,
      "train_runtime": 11444.3393,
      "train_tokens_per_second": 10123.392
    },
    {
      "epoch": 0.45878870673952643,
      "grad_norm": 0.6944263963974879,
      "learning_rate": 8.758019723697176e-06,
      "loss": 0.343,
      "num_input_tokens_seen": 116177648,
      "step": 2015,
      "train_runtime": 11476.616,
      "train_tokens_per_second": 10122.988
    },
    {
      "epoch": 0.4599271402550091,
      "grad_norm": 0.6489540121638205,
      "learning_rate": 8.752115947727441e-06,
      "loss": 0.3639,
      "num_input_tokens_seen": 116451488,
      "step": 2020,
      "train_runtime": 11504.8771,
      "train_tokens_per_second": 10121.924
    },
    {
      "epoch": 0.4610655737704918,
      "grad_norm": 0.6293889886929444,
      "learning_rate": 8.746200173139494e-06,
      "loss": 0.2955,
      "num_input_tokens_seen": 116745896,
      "step": 2025,
      "train_runtime": 11540.2926,
      "train_tokens_per_second": 10116.372
    },
    {
      "epoch": 0.4622040072859745,
      "grad_norm": 0.7928403892501844,
      "learning_rate": 8.74027241885096e-06,
      "loss": 0.3436,
      "num_input_tokens_seen": 117021040,
      "step": 2030,
      "train_runtime": 11573.3681,
      "train_tokens_per_second": 10111.235
    },
    {
      "epoch": 0.4633424408014572,
      "grad_norm": 0.6260708726602756,
      "learning_rate": 8.734332703817771e-06,
      "loss": 0.3118,
      "num_input_tokens_seen": 117294152,
      "step": 2035,
      "train_runtime": 11601.1207,
      "train_tokens_per_second": 10110.588
    },
    {
      "epoch": 0.4644808743169399,
      "grad_norm": 0.5350417450892574,
      "learning_rate": 8.728381047034114e-06,
      "loss": 0.3097,
      "num_input_tokens_seen": 117580288,
      "step": 2040,
      "train_runtime": 11634.0143,
      "train_tokens_per_second": 10106.597
    },
    {
      "epoch": 0.4656193078324226,
      "grad_norm": 0.6544629607349358,
      "learning_rate": 8.722417467532352e-06,
      "loss": 0.3294,
      "num_input_tokens_seen": 117876208,
      "step": 2045,
      "train_runtime": 11670.2254,
      "train_tokens_per_second": 10100.594
    },
    {
      "epoch": 0.46675774134790526,
      "grad_norm": 0.6010992614378249,
      "learning_rate": 8.716441984382986e-06,
      "loss": 0.3406,
      "num_input_tokens_seen": 118181952,
      "step": 2050,
      "train_runtime": 11702.8095,
      "train_tokens_per_second": 10098.597
    },
    {
      "epoch": 0.46789617486338797,
      "grad_norm": 0.5404731873011414,
      "learning_rate": 8.710454616694575e-06,
      "loss": 0.337,
      "num_input_tokens_seen": 118453736,
      "step": 2055,
      "train_runtime": 11730.5889,
      "train_tokens_per_second": 10097.851
    },
    {
      "epoch": 0.4690346083788707,
      "grad_norm": 0.7127646010839479,
      "learning_rate": 8.704455383613688e-06,
      "loss": 0.3228,
      "num_input_tokens_seen": 118730648,
      "step": 2060,
      "train_runtime": 11756.8705,
      "train_tokens_per_second": 10098.831
    },
    {
      "epoch": 0.4701730418943534,
      "grad_norm": 0.6056126693739166,
      "learning_rate": 8.698444304324837e-06,
      "loss": 0.3269,
      "num_input_tokens_seen": 119025832,
      "step": 2065,
      "train_runtime": 11783.9731,
      "train_tokens_per_second": 10100.654
    },
    {
      "epoch": 0.4713114754098361,
      "grad_norm": 0.6457232968891127,
      "learning_rate": 8.692421398050409e-06,
      "loss": 0.2859,
      "num_input_tokens_seen": 119316416,
      "step": 2070,
      "train_runtime": 11811.1456,
      "train_tokens_per_second": 10102.019
    },
    {
      "epoch": 0.47244990892531874,
      "grad_norm": 0.6766926154594638,
      "learning_rate": 8.68638668405062e-06,
      "loss": 0.3069,
      "num_input_tokens_seen": 119616632,
      "step": 2075,
      "train_runtime": 11837.1983,
      "train_tokens_per_second": 10105.147
    },
    {
      "epoch": 0.47358834244080145,
      "grad_norm": 0.672287763379747,
      "learning_rate": 8.680340181623446e-06,
      "loss": 0.3548,
      "num_input_tokens_seen": 119880224,
      "step": 2080,
      "train_runtime": 11864.1788,
      "train_tokens_per_second": 10104.384
    },
    {
      "epoch": 0.47472677595628415,
      "grad_norm": 0.6591051871968795,
      "learning_rate": 8.674281910104551e-06,
      "loss": 0.343,
      "num_input_tokens_seen": 120146968,
      "step": 2085,
      "train_runtime": 11889.3134,
      "train_tokens_per_second": 10105.459
    },
    {
      "epoch": 0.47586520947176686,
      "grad_norm": 0.7313344682070815,
      "learning_rate": 8.668211888867243e-06,
      "loss": 0.3482,
      "num_input_tokens_seen": 120426152,
      "step": 2090,
      "train_runtime": 11914.5097,
      "train_tokens_per_second": 10107.521
    },
    {
      "epoch": 0.47700364298724957,
      "grad_norm": 0.7068757949929048,
      "learning_rate": 8.662130137322401e-06,
      "loss": 0.346,
      "num_input_tokens_seen": 120714800,
      "step": 2095,
      "train_runtime": 11947.6144,
      "train_tokens_per_second": 10103.674
    },
    {
      "epoch": 0.4781420765027322,
      "grad_norm": 0.5916332681579296,
      "learning_rate": 8.656036674918414e-06,
      "loss": 0.3413,
      "num_input_tokens_seen": 120999704,
      "step": 2100,
      "train_runtime": 11974.6803,
      "train_tokens_per_second": 10104.629
    },
    {
      "epoch": 0.4792805100182149,
      "grad_norm": 0.5242261124492325,
      "learning_rate": 8.64993152114112e-06,
      "loss": 0.2894,
      "num_input_tokens_seen": 121288896,
      "step": 2105,
      "train_runtime": 11995.9824,
      "train_tokens_per_second": 10110.793
    },
    {
      "epoch": 0.48041894353369763,
      "grad_norm": 0.6341654060825161,
      "learning_rate": 8.643814695513747e-06,
      "loss": 0.3317,
      "num_input_tokens_seen": 121566792,
      "step": 2110,
      "train_runtime": 12019.65,
      "train_tokens_per_second": 10114.004
    },
    {
      "epoch": 0.48155737704918034,
      "grad_norm": 0.5612450740764198,
      "learning_rate": 8.637686217596847e-06,
      "loss": 0.3102,
      "num_input_tokens_seen": 121859104,
      "step": 2115,
      "train_runtime": 12048.1062,
      "train_tokens_per_second": 10114.378
    },
    {
      "epoch": 0.48269581056466304,
      "grad_norm": 0.6541646224275115,
      "learning_rate": 8.631546106988233e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 122160792,
      "step": 2120,
      "train_runtime": 12073.1991,
      "train_tokens_per_second": 10118.345
    },
    {
      "epoch": 0.4838342440801457,
      "grad_norm": 0.6132972633910501,
      "learning_rate": 8.625394383322914e-06,
      "loss": 0.304,
      "num_input_tokens_seen": 122442480,
      "step": 2125,
      "train_runtime": 12098.34,
      "train_tokens_per_second": 10120.602
    },
    {
      "epoch": 0.4849726775956284,
      "grad_norm": 0.6454197471280586,
      "learning_rate": 8.619231066273044e-06,
      "loss": 0.3387,
      "num_input_tokens_seen": 122713056,
      "step": 2130,
      "train_runtime": 12125.4784,
      "train_tokens_per_second": 10120.265
    },
    {
      "epoch": 0.4861111111111111,
      "grad_norm": 0.7101600601420914,
      "learning_rate": 8.613056175547841e-06,
      "loss": 0.351,
      "num_input_tokens_seen": 122989048,
      "step": 2135,
      "train_runtime": 12154.4607,
      "train_tokens_per_second": 10118.84
    },
    {
      "epoch": 0.4872495446265938,
      "grad_norm": 0.6158626771626245,
      "learning_rate": 8.606869730893542e-06,
      "loss": 0.3237,
      "num_input_tokens_seen": 123268416,
      "step": 2140,
      "train_runtime": 12179.1652,
      "train_tokens_per_second": 10121.253
    },
    {
      "epoch": 0.4883879781420765,
      "grad_norm": 0.7615688907278779,
      "learning_rate": 8.600671752093323e-06,
      "loss": 0.3228,
      "num_input_tokens_seen": 123551840,
      "step": 2145,
      "train_runtime": 12211.5727,
      "train_tokens_per_second": 10117.603
    },
    {
      "epoch": 0.4895264116575592,
      "grad_norm": 0.640653791774208,
      "learning_rate": 8.594462258967254e-06,
      "loss": 0.2982,
      "num_input_tokens_seen": 123835904,
      "step": 2150,
      "train_runtime": 12243.668,
      "train_tokens_per_second": 10114.281
    },
    {
      "epoch": 0.4906648451730419,
      "grad_norm": 0.6082442394151018,
      "learning_rate": 8.588241271372215e-06,
      "loss": 0.3063,
      "num_input_tokens_seen": 124106024,
      "step": 2155,
      "train_runtime": 12280.1701,
      "train_tokens_per_second": 10106.214
    },
    {
      "epoch": 0.4918032786885246,
      "grad_norm": 0.6476905589741084,
      "learning_rate": 8.582008809201854e-06,
      "loss": 0.2967,
      "num_input_tokens_seen": 124377568,
      "step": 2160,
      "train_runtime": 12307.6002,
      "train_tokens_per_second": 10105.753
    },
    {
      "epoch": 0.4929417122040073,
      "grad_norm": 0.5275532867098846,
      "learning_rate": 8.575764892386505e-06,
      "loss": 0.3201,
      "num_input_tokens_seen": 124696528,
      "step": 2165,
      "train_runtime": 12331.3284,
      "train_tokens_per_second": 10112.173
    },
    {
      "epoch": 0.49408014571949,
      "grad_norm": 0.5290540355526318,
      "learning_rate": 8.569509540893135e-06,
      "loss": 0.2943,
      "num_input_tokens_seen": 125002664,
      "step": 2170,
      "train_runtime": 12364.511,
      "train_tokens_per_second": 10109.794
    },
    {
      "epoch": 0.4952185792349727,
      "grad_norm": 0.6042699145971743,
      "learning_rate": 8.56324277472528e-06,
      "loss": 0.321,
      "num_input_tokens_seen": 125289160,
      "step": 2175,
      "train_runtime": 12393.1625,
      "train_tokens_per_second": 10109.539
    },
    {
      "epoch": 0.49635701275045535,
      "grad_norm": 0.6151318817217737,
      "learning_rate": 8.55696461392297e-06,
      "loss": 0.3163,
      "num_input_tokens_seen": 125561824,
      "step": 2180,
      "train_runtime": 12422.5204,
      "train_tokens_per_second": 10107.597
    },
    {
      "epoch": 0.49749544626593806,
      "grad_norm": 0.5612653767622556,
      "learning_rate": 8.55067507856268e-06,
      "loss": 0.3042,
      "num_input_tokens_seen": 125825304,
      "step": 2185,
      "train_runtime": 12457.7627,
      "train_tokens_per_second": 10100.153
    },
    {
      "epoch": 0.49863387978142076,
      "grad_norm": 0.5170162979646165,
      "learning_rate": 8.544374188757257e-06,
      "loss": 0.2778,
      "num_input_tokens_seen": 126097080,
      "step": 2190,
      "train_runtime": 12480.08,
      "train_tokens_per_second": 10103.868
    },
    {
      "epoch": 0.49977231329690347,
      "grad_norm": 0.6243126254015997,
      "learning_rate": 8.538061964655858e-06,
      "loss": 0.323,
      "num_input_tokens_seen": 126403784,
      "step": 2195,
      "train_runtime": 12507.9689,
      "train_tokens_per_second": 10105.86
    },
    {
      "epoch": 0.5009107468123861,
      "grad_norm": 0.6746359774832735,
      "learning_rate": 8.531738426443883e-06,
      "loss": 0.3407,
      "num_input_tokens_seen": 126701288,
      "step": 2200,
      "train_runtime": 12531.0799,
      "train_tokens_per_second": 10110.963
    },
    {
      "epoch": 0.5020491803278688,
      "grad_norm": 0.6419652156878918,
      "learning_rate": 8.525403594342915e-06,
      "loss": 0.314,
      "num_input_tokens_seen": 126990904,
      "step": 2205,
      "train_runtime": 12550.9497,
      "train_tokens_per_second": 10118.031
    },
    {
      "epoch": 0.5031876138433515,
      "grad_norm": 0.539159347791682,
      "learning_rate": 8.519057488610654e-06,
      "loss": 0.3248,
      "num_input_tokens_seen": 127302216,
      "step": 2210,
      "train_runtime": 12580.4566,
      "train_tokens_per_second": 10119.046
    },
    {
      "epoch": 0.5043260473588342,
      "grad_norm": 0.48956299755212457,
      "learning_rate": 8.512700129540847e-06,
      "loss": 0.3216,
      "num_input_tokens_seen": 127589592,
      "step": 2215,
      "train_runtime": 12605.1561,
      "train_tokens_per_second": 10122.016
    },
    {
      "epoch": 0.505464480874317,
      "grad_norm": 0.5254536320148653,
      "learning_rate": 8.506331537463228e-06,
      "loss": 0.2874,
      "num_input_tokens_seen": 127871360,
      "step": 2220,
      "train_runtime": 12638.2007,
      "train_tokens_per_second": 10117.845
    },
    {
      "epoch": 0.5066029143897997,
      "grad_norm": 0.6281120035703216,
      "learning_rate": 8.499951732743457e-06,
      "loss": 0.2887,
      "num_input_tokens_seen": 128154792,
      "step": 2225,
      "train_runtime": 12666.9338,
      "train_tokens_per_second": 10117.27
    },
    {
      "epoch": 0.5077413479052824,
      "grad_norm": 0.53536529837043,
      "learning_rate": 8.493560735783044e-06,
      "loss": 0.3311,
      "num_input_tokens_seen": 128444008,
      "step": 2230,
      "train_runtime": 12698.2789,
      "train_tokens_per_second": 10115.072
    },
    {
      "epoch": 0.5088797814207651,
      "grad_norm": 0.5596887687245609,
      "learning_rate": 8.487158567019292e-06,
      "loss": 0.2987,
      "num_input_tokens_seen": 128745848,
      "step": 2235,
      "train_runtime": 12727.2834,
      "train_tokens_per_second": 10115.737
    },
    {
      "epoch": 0.5100182149362478,
      "grad_norm": 0.8309210903251103,
      "learning_rate": 8.480745246925234e-06,
      "loss": 0.3159,
      "num_input_tokens_seen": 129002616,
      "step": 2240,
      "train_runtime": 12750.9345,
      "train_tokens_per_second": 10117.111
    },
    {
      "epoch": 0.5111566484517304,
      "grad_norm": 0.8133456806874431,
      "learning_rate": 8.474320796009554e-06,
      "loss": 0.3352,
      "num_input_tokens_seen": 129274424,
      "step": 2245,
      "train_runtime": 12771.6378,
      "train_tokens_per_second": 10121.993
    },
    {
      "epoch": 0.5122950819672131,
      "grad_norm": 0.5154589417885288,
      "learning_rate": 8.46788523481654e-06,
      "loss": 0.2844,
      "num_input_tokens_seen": 129582560,
      "step": 2250,
      "train_runtime": 12790.5884,
      "train_tokens_per_second": 10131.087
    },
    {
      "epoch": 0.5134335154826958,
      "grad_norm": 0.6684546214201272,
      "learning_rate": 8.461438583926e-06,
      "loss": 0.3073,
      "num_input_tokens_seen": 129858592,
      "step": 2255,
      "train_runtime": 12822.0883,
      "train_tokens_per_second": 10127.726
    },
    {
      "epoch": 0.5145719489981785,
      "grad_norm": 0.5652006830124429,
      "learning_rate": 8.454980863953213e-06,
      "loss": 0.3286,
      "num_input_tokens_seen": 130123536,
      "step": 2260,
      "train_runtime": 12843.5027,
      "train_tokens_per_second": 10131.468
    },
    {
      "epoch": 0.5157103825136612,
      "grad_norm": 0.6331165255397602,
      "learning_rate": 8.44851209554885e-06,
      "loss": 0.2995,
      "num_input_tokens_seen": 130418760,
      "step": 2265,
      "train_runtime": 12868.3465,
      "train_tokens_per_second": 10134.85
    },
    {
      "epoch": 0.5168488160291439,
      "grad_norm": 0.6148162896655509,
      "learning_rate": 8.442032299398914e-06,
      "loss": 0.3254,
      "num_input_tokens_seen": 130716160,
      "step": 2270,
      "train_runtime": 12901.1384,
      "train_tokens_per_second": 10132.142
    },
    {
      "epoch": 0.5179872495446266,
      "grad_norm": 0.658205510556745,
      "learning_rate": 8.435541496224674e-06,
      "loss": 0.3409,
      "num_input_tokens_seen": 130970728,
      "step": 2275,
      "train_runtime": 12933.5976,
      "train_tokens_per_second": 10126.396
    },
    {
      "epoch": 0.5191256830601093,
      "grad_norm": 0.5275043649652696,
      "learning_rate": 8.42903970678259e-06,
      "loss": 0.3024,
      "num_input_tokens_seen": 131277440,
      "step": 2280,
      "train_runtime": 12962.175,
      "train_tokens_per_second": 10127.732
    },
    {
      "epoch": 0.520264116575592,
      "grad_norm": 0.7509767918860819,
      "learning_rate": 8.422526951864271e-06,
      "loss": 0.311,
      "num_input_tokens_seen": 131565840,
      "step": 2285,
      "train_runtime": 12984.2311,
      "train_tokens_per_second": 10132.74
    },
    {
      "epoch": 0.5214025500910747,
      "grad_norm": 0.5819934511558694,
      "learning_rate": 8.416003252296377e-06,
      "loss": 0.3301,
      "num_input_tokens_seen": 131865088,
      "step": 2290,
      "train_runtime": 13012.623,
      "train_tokens_per_second": 10133.629
    },
    {
      "epoch": 0.5225409836065574,
      "grad_norm": 0.5564482921108145,
      "learning_rate": 8.409468628940568e-06,
      "loss": 0.3112,
      "num_input_tokens_seen": 132161328,
      "step": 2295,
      "train_runtime": 13034.1103,
      "train_tokens_per_second": 10139.651
    },
    {
      "epoch": 0.52367941712204,
      "grad_norm": 0.6196302741925617,
      "learning_rate": 8.402923102693446e-06,
      "loss": 0.3215,
      "num_input_tokens_seen": 132457288,
      "step": 2300,
      "train_runtime": 13063.3414,
      "train_tokens_per_second": 10139.618
    },
    {
      "epoch": 0.5248178506375227,
      "grad_norm": 0.6521657145211633,
      "learning_rate": 8.396366694486466e-06,
      "loss": 0.3297,
      "num_input_tokens_seen": 132741600,
      "step": 2305,
      "train_runtime": 13089.7255,
      "train_tokens_per_second": 10140.9
    },
    {
      "epoch": 0.5259562841530054,
      "grad_norm": 0.47493304518796103,
      "learning_rate": 8.389799425285894e-06,
      "loss": 0.2904,
      "num_input_tokens_seen": 133036840,
      "step": 2310,
      "train_runtime": 13115.8436,
      "train_tokens_per_second": 10143.216
    },
    {
      "epoch": 0.5270947176684881,
      "grad_norm": 0.5946108109757725,
      "learning_rate": 8.383221316092718e-06,
      "loss": 0.2999,
      "num_input_tokens_seen": 133334544,
      "step": 2315,
      "train_runtime": 13136.9131,
      "train_tokens_per_second": 10149.61
    },
    {
      "epoch": 0.5282331511839709,
      "grad_norm": 0.6904613702819206,
      "learning_rate": 8.376632387942595e-06,
      "loss": 0.3234,
      "num_input_tokens_seen": 133611872,
      "step": 2320,
      "train_runtime": 13165.1986,
      "train_tokens_per_second": 10148.869
    },
    {
      "epoch": 0.5293715846994536,
      "grad_norm": 0.6190372243316938,
      "learning_rate": 8.37003266190578e-06,
      "loss": 0.2901,
      "num_input_tokens_seen": 133896048,
      "step": 2325,
      "train_runtime": 13196.655,
      "train_tokens_per_second": 10146.211
    },
    {
      "epoch": 0.5305100182149363,
      "grad_norm": 0.5792892894818183,
      "learning_rate": 8.363422159087055e-06,
      "loss": 0.3057,
      "num_input_tokens_seen": 134177000,
      "step": 2330,
      "train_runtime": 13220.9541,
      "train_tokens_per_second": 10148.814
    },
    {
      "epoch": 0.531648451730419,
      "grad_norm": 0.5938627491089601,
      "learning_rate": 8.356800900625666e-06,
      "loss": 0.3224,
      "num_input_tokens_seen": 134455720,
      "step": 2335,
      "train_runtime": 13252.0097,
      "train_tokens_per_second": 10146.063
    },
    {
      "epoch": 0.5327868852459017,
      "grad_norm": 0.6806466958201689,
      "learning_rate": 8.350168907695249e-06,
      "loss": 0.2998,
      "num_input_tokens_seen": 134747576,
      "step": 2340,
      "train_runtime": 13277.0446,
      "train_tokens_per_second": 10148.913
    },
    {
      "epoch": 0.5339253187613844,
      "grad_norm": 0.6818894356073065,
      "learning_rate": 8.343526201503775e-06,
      "loss": 0.3251,
      "num_input_tokens_seen": 135041392,
      "step": 2345,
      "train_runtime": 13302.7072,
      "train_tokens_per_second": 10151.422
    },
    {
      "epoch": 0.535063752276867,
      "grad_norm": 0.7050659369104935,
      "learning_rate": 8.33687280329347e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 135352360,
      "step": 2350,
      "train_runtime": 13334.9649,
      "train_tokens_per_second": 10150.185
    },
    {
      "epoch": 0.5362021857923497,
      "grad_norm": 0.5816732708433414,
      "learning_rate": 8.33020873434075e-06,
      "loss": 0.3408,
      "num_input_tokens_seen": 135642960,
      "step": 2355,
      "train_runtime": 13368.2904,
      "train_tokens_per_second": 10146.62
    },
    {
      "epoch": 0.5373406193078324,
      "grad_norm": 0.9637792967700637,
      "learning_rate": 8.323534015956156e-06,
      "loss": 0.3076,
      "num_input_tokens_seen": 135933408,
      "step": 2360,
      "train_runtime": 13403.0445,
      "train_tokens_per_second": 10141.98
    },
    {
      "epoch": 0.5384790528233151,
      "grad_norm": 0.5618849724108224,
      "learning_rate": 8.316848669484283e-06,
      "loss": 0.3132,
      "num_input_tokens_seen": 136231448,
      "step": 2365,
      "train_runtime": 13427.7637,
      "train_tokens_per_second": 10145.505
    },
    {
      "epoch": 0.5396174863387978,
      "grad_norm": 0.7733593661216478,
      "learning_rate": 8.310152716303716e-06,
      "loss": 0.3007,
      "num_input_tokens_seen": 136514840,
      "step": 2370,
      "train_runtime": 13456.5889,
      "train_tokens_per_second": 10144.832
    },
    {
      "epoch": 0.5407559198542805,
      "grad_norm": 0.582283146519985,
      "learning_rate": 8.303446177826952e-06,
      "loss": 0.3157,
      "num_input_tokens_seen": 136787288,
      "step": 2375,
      "train_runtime": 13481.4151,
      "train_tokens_per_second": 10146.36
    },
    {
      "epoch": 0.5418943533697632,
      "grad_norm": 0.6039683224473668,
      "learning_rate": 8.296729075500345e-06,
      "loss": 0.3053,
      "num_input_tokens_seen": 137074976,
      "step": 2380,
      "train_runtime": 13510.2441,
      "train_tokens_per_second": 10146.003
    },
    {
      "epoch": 0.5430327868852459,
      "grad_norm": 0.6060518581118172,
      "learning_rate": 8.290001430804026e-06,
      "loss": 0.3499,
      "num_input_tokens_seen": 137373856,
      "step": 2385,
      "train_runtime": 13539.6954,
      "train_tokens_per_second": 10146.008
    },
    {
      "epoch": 0.5441712204007286,
      "grad_norm": 0.5888285275483841,
      "learning_rate": 8.283263265251838e-06,
      "loss": 0.3274,
      "num_input_tokens_seen": 137637416,
      "step": 2390,
      "train_runtime": 13567.3159,
      "train_tokens_per_second": 10144.779
    },
    {
      "epoch": 0.5453096539162113,
      "grad_norm": 0.5746356965813089,
      "learning_rate": 8.276514600391272e-06,
      "loss": 0.2936,
      "num_input_tokens_seen": 137936080,
      "step": 2395,
      "train_runtime": 13598.323,
      "train_tokens_per_second": 10143.61
    },
    {
      "epoch": 0.546448087431694,
      "grad_norm": 0.5397941086102784,
      "learning_rate": 8.269755457803392e-06,
      "loss": 0.3098,
      "num_input_tokens_seen": 138224680,
      "step": 2400,
      "train_runtime": 13629.9585,
      "train_tokens_per_second": 10141.24
    },
    {
      "epoch": 0.5475865209471766,
      "grad_norm": 0.5893122795724934,
      "learning_rate": 8.262985859102765e-06,
      "loss": 0.2858,
      "num_input_tokens_seen": 138507864,
      "step": 2405,
      "train_runtime": 13659.9315,
      "train_tokens_per_second": 10139.719
    },
    {
      "epoch": 0.5487249544626593,
      "grad_norm": 0.6034825161665428,
      "learning_rate": 8.2562058259374e-06,
      "loss": 0.2897,
      "num_input_tokens_seen": 138811176,
      "step": 2410,
      "train_runtime": 13690.1837,
      "train_tokens_per_second": 10139.468
    },
    {
      "epoch": 0.549863387978142,
      "grad_norm": 0.638238195416211,
      "learning_rate": 8.249415379988667e-06,
      "loss": 0.3221,
      "num_input_tokens_seen": 139081944,
      "step": 2415,
      "train_runtime": 13717.7566,
      "train_tokens_per_second": 10138.826
    },
    {
      "epoch": 0.5510018214936248,
      "grad_norm": 0.6696015630192675,
      "learning_rate": 8.242614542971246e-06,
      "loss": 0.2977,
      "num_input_tokens_seen": 139372592,
      "step": 2420,
      "train_runtime": 13747.7091,
      "train_tokens_per_second": 10137.878
    },
    {
      "epoch": 0.5521402550091075,
      "grad_norm": 0.6251115744591602,
      "learning_rate": 8.235803336633033e-06,
      "loss": 0.336,
      "num_input_tokens_seen": 139669592,
      "step": 2425,
      "train_runtime": 13777.0839,
      "train_tokens_per_second": 10137.82
    },
    {
      "epoch": 0.5532786885245902,
      "grad_norm": 0.5836991041261375,
      "learning_rate": 8.228981782755086e-06,
      "loss": 0.3308,
      "num_input_tokens_seen": 139979776,
      "step": 2430,
      "train_runtime": 13803.198,
      "train_tokens_per_second": 10141.112
    },
    {
      "epoch": 0.5544171220400729,
      "grad_norm": 0.6153464787522137,
      "learning_rate": 8.22214990315156e-06,
      "loss": 0.3092,
      "num_input_tokens_seen": 140279240,
      "step": 2435,
      "train_runtime": 13830.4865,
      "train_tokens_per_second": 10142.755
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.6193161205358845,
      "learning_rate": 8.215307719669622e-06,
      "loss": 0.3186,
      "num_input_tokens_seen": 140574312,
      "step": 2440,
      "train_runtime": 13866.1467,
      "train_tokens_per_second": 10137.951
    },
    {
      "epoch": 0.5566939890710383,
      "grad_norm": 0.5892663614345347,
      "learning_rate": 8.208455254189397e-06,
      "loss": 0.3476,
      "num_input_tokens_seen": 140866384,
      "step": 2445,
      "train_runtime": 13890.0426,
      "train_tokens_per_second": 10141.537
    },
    {
      "epoch": 0.557832422586521,
      "grad_norm": 0.6386342486469806,
      "learning_rate": 8.201592528623877e-06,
      "loss": 0.3465,
      "num_input_tokens_seen": 141166384,
      "step": 2450,
      "train_runtime": 13915.3967,
      "train_tokens_per_second": 10144.618
    },
    {
      "epoch": 0.5589708561020036,
      "grad_norm": 0.6583952391095774,
      "learning_rate": 8.194719564918878e-06,
      "loss": 0.3305,
      "num_input_tokens_seen": 141454584,
      "step": 2455,
      "train_runtime": 13946.366,
      "train_tokens_per_second": 10142.756
    },
    {
      "epoch": 0.5601092896174863,
      "grad_norm": 0.8840215225932602,
      "learning_rate": 8.187836385052948e-06,
      "loss": 0.2961,
      "num_input_tokens_seen": 141745328,
      "step": 2460,
      "train_runtime": 13974.146,
      "train_tokens_per_second": 10143.398
    },
    {
      "epoch": 0.561247723132969,
      "grad_norm": 0.6469660287416618,
      "learning_rate": 8.180943011037306e-06,
      "loss": 0.3433,
      "num_input_tokens_seen": 142019208,
      "step": 2465,
      "train_runtime": 14002.2357,
      "train_tokens_per_second": 10142.609
    },
    {
      "epoch": 0.5623861566484517,
      "grad_norm": 0.4849003406886445,
      "learning_rate": 8.174039464915773e-06,
      "loss": 0.3138,
      "num_input_tokens_seen": 142310648,
      "step": 2470,
      "train_runtime": 14029.5128,
      "train_tokens_per_second": 10143.663
    },
    {
      "epoch": 0.5635245901639344,
      "grad_norm": 0.5124082172158423,
      "learning_rate": 8.167125768764694e-06,
      "loss": 0.292,
      "num_input_tokens_seen": 142606088,
      "step": 2475,
      "train_runtime": 14052.3269,
      "train_tokens_per_second": 10148.219
    },
    {
      "epoch": 0.5646630236794171,
      "grad_norm": 0.664526377982131,
      "learning_rate": 8.160201944692874e-06,
      "loss": 0.3355,
      "num_input_tokens_seen": 142910016,
      "step": 2480,
      "train_runtime": 14079.7553,
      "train_tokens_per_second": 10150.035
    },
    {
      "epoch": 0.5658014571948998,
      "grad_norm": 0.6653671581567494,
      "learning_rate": 8.153268014841507e-06,
      "loss": 0.339,
      "num_input_tokens_seen": 143197368,
      "step": 2485,
      "train_runtime": 14113.3357,
      "train_tokens_per_second": 10146.245
    },
    {
      "epoch": 0.5669398907103825,
      "grad_norm": 0.5753449698202951,
      "learning_rate": 8.1463240013841e-06,
      "loss": 0.2996,
      "num_input_tokens_seen": 143471872,
      "step": 2490,
      "train_runtime": 14141.3989,
      "train_tokens_per_second": 10145.522
    },
    {
      "epoch": 0.5680783242258652,
      "grad_norm": 0.7490389918413549,
      "learning_rate": 8.139369926526409e-06,
      "loss": 0.3185,
      "num_input_tokens_seen": 143762016,
      "step": 2495,
      "train_runtime": 14175.1321,
      "train_tokens_per_second": 10141.847
    },
    {
      "epoch": 0.569216757741348,
      "grad_norm": 0.6220342037775106,
      "learning_rate": 8.132405812506364e-06,
      "loss": 0.357,
      "num_input_tokens_seen": 144060624,
      "step": 2500,
      "train_runtime": 14204.9792,
      "train_tokens_per_second": 10141.558
    },
    {
      "epoch": 0.5703551912568307,
      "grad_norm": 0.6259029124098564,
      "learning_rate": 8.125431681593996e-06,
      "loss": 0.2967,
      "num_input_tokens_seen": 144355648,
      "step": 2505,
      "train_runtime": 14233.9354,
      "train_tokens_per_second": 10141.654
    },
    {
      "epoch": 0.5714936247723132,
      "grad_norm": 0.5881551659775729,
      "learning_rate": 8.11844755609137e-06,
      "loss": 0.324,
      "num_input_tokens_seen": 144628200,
      "step": 2510,
      "train_runtime": 14262.7668,
      "train_tokens_per_second": 10140.263
    },
    {
      "epoch": 0.572632058287796,
      "grad_norm": 0.7229224614271025,
      "learning_rate": 8.111453458332516e-06,
      "loss": 0.31,
      "num_input_tokens_seen": 144880624,
      "step": 2515,
      "train_runtime": 14290.0412,
      "train_tokens_per_second": 10138.573
    },
    {
      "epoch": 0.5737704918032787,
      "grad_norm": 0.6013586860681409,
      "learning_rate": 8.104449410683346e-06,
      "loss": 0.3053,
      "num_input_tokens_seen": 145178344,
      "step": 2520,
      "train_runtime": 14316.5758,
      "train_tokens_per_second": 10140.577
    },
    {
      "epoch": 0.5749089253187614,
      "grad_norm": 0.6436524670377838,
      "learning_rate": 8.097435435541593e-06,
      "loss": 0.3022,
      "num_input_tokens_seen": 145460072,
      "step": 2525,
      "train_runtime": 14344.1209,
      "train_tokens_per_second": 10140.745
    },
    {
      "epoch": 0.5760473588342441,
      "grad_norm": 0.693528242762699,
      "learning_rate": 8.09041155533674e-06,
      "loss": 0.3312,
      "num_input_tokens_seen": 145750904,
      "step": 2530,
      "train_runtime": 14370.091,
      "train_tokens_per_second": 10142.657
    },
    {
      "epoch": 0.5771857923497268,
      "grad_norm": 0.5990202731692641,
      "learning_rate": 8.08337779252994e-06,
      "loss": 0.2993,
      "num_input_tokens_seen": 146049240,
      "step": 2535,
      "train_runtime": 14398.8788,
      "train_tokens_per_second": 10143.098
    },
    {
      "epoch": 0.5783242258652095,
      "grad_norm": 0.6822880365624007,
      "learning_rate": 8.07633416961395e-06,
      "loss": 0.3271,
      "num_input_tokens_seen": 146339352,
      "step": 2540,
      "train_runtime": 14427.8577,
      "train_tokens_per_second": 10142.833
    },
    {
      "epoch": 0.5794626593806922,
      "grad_norm": 0.6172655261336489,
      "learning_rate": 8.06928070911306e-06,
      "loss": 0.3246,
      "num_input_tokens_seen": 146639440,
      "step": 2545,
      "train_runtime": 14457.5048,
      "train_tokens_per_second": 10142.79
    },
    {
      "epoch": 0.5806010928961749,
      "grad_norm": 0.6846965214608297,
      "learning_rate": 8.062217433583017e-06,
      "loss": 0.3318,
      "num_input_tokens_seen": 146910904,
      "step": 2550,
      "train_runtime": 14490.1588,
      "train_tokens_per_second": 10138.668
    },
    {
      "epoch": 0.5817395264116576,
      "grad_norm": 0.5983216087022071,
      "learning_rate": 8.055144365610956e-06,
      "loss": 0.2945,
      "num_input_tokens_seen": 147205168,
      "step": 2555,
      "train_runtime": 14516.9602,
      "train_tokens_per_second": 10140.22
    },
    {
      "epoch": 0.5828779599271403,
      "grad_norm": 0.5929854350332707,
      "learning_rate": 8.048061527815323e-06,
      "loss": 0.2855,
      "num_input_tokens_seen": 147491752,
      "step": 2560,
      "train_runtime": 14543.502,
      "train_tokens_per_second": 10141.419
    },
    {
      "epoch": 0.5840163934426229,
      "grad_norm": 0.600878215032278,
      "learning_rate": 8.040968942845811e-06,
      "loss": 0.316,
      "num_input_tokens_seen": 147783656,
      "step": 2565,
      "train_runtime": 14576.45,
      "train_tokens_per_second": 10138.522
    },
    {
      "epoch": 0.5851548269581056,
      "grad_norm": 0.727040403053856,
      "learning_rate": 8.03386663338328e-06,
      "loss": 0.3331,
      "num_input_tokens_seen": 148077904,
      "step": 2570,
      "train_runtime": 14607.0445,
      "train_tokens_per_second": 10137.431
    },
    {
      "epoch": 0.5862932604735883,
      "grad_norm": 0.6186517276852056,
      "learning_rate": 8.026754622139691e-06,
      "loss": 0.3262,
      "num_input_tokens_seen": 148364464,
      "step": 2575,
      "train_runtime": 14635.7949,
      "train_tokens_per_second": 10137.096
    },
    {
      "epoch": 0.587431693989071,
      "grad_norm": 0.6163389096674644,
      "learning_rate": 8.019632931858021e-06,
      "loss": 0.3149,
      "num_input_tokens_seen": 148664328,
      "step": 2580,
      "train_runtime": 14662.7281,
      "train_tokens_per_second": 10138.927
    },
    {
      "epoch": 0.5885701275045537,
      "grad_norm": 0.6072709479850795,
      "learning_rate": 8.01250158531221e-06,
      "loss": 0.3164,
      "num_input_tokens_seen": 148946672,
      "step": 2585,
      "train_runtime": 14692.6548,
      "train_tokens_per_second": 10137.492
    },
    {
      "epoch": 0.5897085610200364,
      "grad_norm": 0.5415321839314251,
      "learning_rate": 8.005360605307068e-06,
      "loss": 0.3128,
      "num_input_tokens_seen": 149250120,
      "step": 2590,
      "train_runtime": 14719.2547,
      "train_tokens_per_second": 10139.788
    },
    {
      "epoch": 0.5908469945355191,
      "grad_norm": 0.6093813280302901,
      "learning_rate": 7.998210014678215e-06,
      "loss": 0.3166,
      "num_input_tokens_seen": 149535904,
      "step": 2595,
      "train_runtime": 14746.3245,
      "train_tokens_per_second": 10140.554
    },
    {
      "epoch": 0.5919854280510018,
      "grad_norm": 0.6015618813224947,
      "learning_rate": 7.991049836292005e-06,
      "loss": 0.3256,
      "num_input_tokens_seen": 149816344,
      "step": 2600,
      "train_runtime": 14774.1929,
      "train_tokens_per_second": 10140.408
    },
    {
      "epoch": 0.5931238615664846,
      "grad_norm": 0.6519270221374796,
      "learning_rate": 7.983880093045448e-06,
      "loss": 0.3211,
      "num_input_tokens_seen": 150117064,
      "step": 2605,
      "train_runtime": 14807.7079,
      "train_tokens_per_second": 10137.765
    },
    {
      "epoch": 0.5942622950819673,
      "grad_norm": 0.673120103752737,
      "learning_rate": 7.976700807866148e-06,
      "loss": 0.3639,
      "num_input_tokens_seen": 150426472,
      "step": 2610,
      "train_runtime": 14840.9411,
      "train_tokens_per_second": 10135.912
    },
    {
      "epoch": 0.5954007285974499,
      "grad_norm": 0.5987493895374171,
      "learning_rate": 7.969512003712217e-06,
      "loss": 0.3158,
      "num_input_tokens_seen": 150720056,
      "step": 2615,
      "train_runtime": 14870.6751,
      "train_tokens_per_second": 10135.388
    },
    {
      "epoch": 0.5965391621129326,
      "grad_norm": 0.5920954862856739,
      "learning_rate": 7.962313703572204e-06,
      "loss": 0.3082,
      "num_input_tokens_seen": 151012784,
      "step": 2620,
      "train_runtime": 14904.9574,
      "train_tokens_per_second": 10131.715
    },
    {
      "epoch": 0.5976775956284153,
      "grad_norm": 0.6551836936562058,
      "learning_rate": 7.955105930465032e-06,
      "loss": 0.3266,
      "num_input_tokens_seen": 151292464,
      "step": 2625,
      "train_runtime": 14936.1809,
      "train_tokens_per_second": 10129.26
    },
    {
      "epoch": 0.598816029143898,
      "grad_norm": 0.552925719420294,
      "learning_rate": 7.947888707439913e-06,
      "loss": 0.3009,
      "num_input_tokens_seen": 151597248,
      "step": 2630,
      "train_runtime": 14969.2506,
      "train_tokens_per_second": 10127.244
    },
    {
      "epoch": 0.5999544626593807,
      "grad_norm": 0.5453331215315572,
      "learning_rate": 7.940662057576278e-06,
      "loss": 0.3269,
      "num_input_tokens_seen": 151874088,
      "step": 2635,
      "train_runtime": 15001.3387,
      "train_tokens_per_second": 10124.036
    },
    {
      "epoch": 0.6010928961748634,
      "grad_norm": 0.628112776291925,
      "learning_rate": 7.933426003983706e-06,
      "loss": 0.2878,
      "num_input_tokens_seen": 152162816,
      "step": 2640,
      "train_runtime": 15026.1823,
      "train_tokens_per_second": 10126.512
    },
    {
      "epoch": 0.6022313296903461,
      "grad_norm": 0.6298430905162717,
      "learning_rate": 7.926180569801843e-06,
      "loss": 0.3559,
      "num_input_tokens_seen": 152430952,
      "step": 2645,
      "train_runtime": 15052.2087,
      "train_tokens_per_second": 10126.816
    },
    {
      "epoch": 0.6033697632058288,
      "grad_norm": 0.6094137513234976,
      "learning_rate": 7.918925778200338e-06,
      "loss": 0.3067,
      "num_input_tokens_seen": 152724064,
      "step": 2650,
      "train_runtime": 15079.5462,
      "train_tokens_per_second": 10127.895
    },
    {
      "epoch": 0.6045081967213115,
      "grad_norm": 0.6785701838835134,
      "learning_rate": 7.911661652378754e-06,
      "loss": 0.3082,
      "num_input_tokens_seen": 152993432,
      "step": 2655,
      "train_runtime": 15110.4607,
      "train_tokens_per_second": 10125.001
    },
    {
      "epoch": 0.6056466302367942,
      "grad_norm": 0.6923346477339406,
      "learning_rate": 7.904388215566518e-06,
      "loss": 0.3297,
      "num_input_tokens_seen": 153259912,
      "step": 2660,
      "train_runtime": 15137.7637,
      "train_tokens_per_second": 10124.343
    },
    {
      "epoch": 0.6067850637522769,
      "grad_norm": 0.5310577520518257,
      "learning_rate": 7.897105491022819e-06,
      "loss": 0.3033,
      "num_input_tokens_seen": 153552328,
      "step": 2665,
      "train_runtime": 15164.0804,
      "train_tokens_per_second": 10126.056
    },
    {
      "epoch": 0.6079234972677595,
      "grad_norm": 0.6834558841840119,
      "learning_rate": 7.889813502036552e-06,
      "loss": 0.3455,
      "num_input_tokens_seen": 153828856,
      "step": 2670,
      "train_runtime": 15193.0846,
      "train_tokens_per_second": 10124.926
    },
    {
      "epoch": 0.6090619307832422,
      "grad_norm": 0.5757002288765387,
      "learning_rate": 7.882512271926239e-06,
      "loss": 0.3431,
      "num_input_tokens_seen": 154139720,
      "step": 2675,
      "train_runtime": 15227.4368,
      "train_tokens_per_second": 10122.499
    },
    {
      "epoch": 0.6102003642987249,
      "grad_norm": 0.6739466467516281,
      "learning_rate": 7.875201824039948e-06,
      "loss": 0.3302,
      "num_input_tokens_seen": 154406152,
      "step": 2680,
      "train_runtime": 15259.7951,
      "train_tokens_per_second": 10118.494
    },
    {
      "epoch": 0.6113387978142076,
      "grad_norm": 0.5922660773899515,
      "learning_rate": 7.86788218175523e-06,
      "loss": 0.3202,
      "num_input_tokens_seen": 154683784,
      "step": 2685,
      "train_runtime": 15288.4173,
      "train_tokens_per_second": 10117.711
    },
    {
      "epoch": 0.6124772313296903,
      "grad_norm": 0.7083875148574846,
      "learning_rate": 7.860553368479038e-06,
      "loss": 0.2865,
      "num_input_tokens_seen": 154936664,
      "step": 2690,
      "train_runtime": 15317.5083,
      "train_tokens_per_second": 10115.004
    },
    {
      "epoch": 0.613615664845173,
      "grad_norm": 0.6793893170851258,
      "learning_rate": 7.853215407647646e-06,
      "loss": 0.3294,
      "num_input_tokens_seen": 155216440,
      "step": 2695,
      "train_runtime": 15341.6038,
      "train_tokens_per_second": 10117.354
    },
    {
      "epoch": 0.6147540983606558,
      "grad_norm": 0.5201032960423712,
      "learning_rate": 7.845868322726586e-06,
      "loss": 0.2835,
      "num_input_tokens_seen": 155505000,
      "step": 2700,
      "train_runtime": 15371.7655,
      "train_tokens_per_second": 10116.275
    },
    {
      "epoch": 0.6158925318761385,
      "grad_norm": 0.6136960631564077,
      "learning_rate": 7.838512137210565e-06,
      "loss": 0.3065,
      "num_input_tokens_seen": 155813024,
      "step": 2705,
      "train_runtime": 15394.83,
      "train_tokens_per_second": 10121.127
    },
    {
      "epoch": 0.6170309653916212,
      "grad_norm": 0.6788503796178108,
      "learning_rate": 7.831146874623395e-06,
      "loss": 0.3176,
      "num_input_tokens_seen": 156123912,
      "step": 2710,
      "train_runtime": 15421.5916,
      "train_tokens_per_second": 10123.722
    },
    {
      "epoch": 0.6181693989071039,
      "grad_norm": 0.5820229124855997,
      "learning_rate": 7.82377255851791e-06,
      "loss": 0.288,
      "num_input_tokens_seen": 156429968,
      "step": 2715,
      "train_runtime": 15449.5864,
      "train_tokens_per_second": 10125.188
    },
    {
      "epoch": 0.6193078324225865,
      "grad_norm": 0.5465986624844412,
      "learning_rate": 7.8163892124759e-06,
      "loss": 0.3375,
      "num_input_tokens_seen": 156722568,
      "step": 2720,
      "train_runtime": 15469.9487,
      "train_tokens_per_second": 10130.775
    },
    {
      "epoch": 0.6204462659380692,
      "grad_norm": 0.6479701425994632,
      "learning_rate": 7.808996860108027e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 157019232,
      "step": 2725,
      "train_runtime": 15495.0814,
      "train_tokens_per_second": 10133.489
    },
    {
      "epoch": 0.6215846994535519,
      "grad_norm": 0.5878639952276918,
      "learning_rate": 7.801595525053757e-06,
      "loss": 0.3092,
      "num_input_tokens_seen": 157322584,
      "step": 2730,
      "train_runtime": 15519.7284,
      "train_tokens_per_second": 10136.942
    },
    {
      "epoch": 0.6227231329690346,
      "grad_norm": 0.5888833669895458,
      "learning_rate": 7.794185230981282e-06,
      "loss": 0.3049,
      "num_input_tokens_seen": 157614920,
      "step": 2735,
      "train_runtime": 15549.5817,
      "train_tokens_per_second": 10136.28
    },
    {
      "epoch": 0.6238615664845173,
      "grad_norm": 0.5804515340937297,
      "learning_rate": 7.786766001587438e-06,
      "loss": 0.3113,
      "num_input_tokens_seen": 157917632,
      "step": 2740,
      "train_runtime": 15583.9306,
      "train_tokens_per_second": 10133.363
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.5807671272788282,
      "learning_rate": 7.779337860597643e-06,
      "loss": 0.315,
      "num_input_tokens_seen": 158172944,
      "step": 2745,
      "train_runtime": 15611.8102,
      "train_tokens_per_second": 10131.621
    },
    {
      "epoch": 0.6261384335154827,
      "grad_norm": 0.5940283270250079,
      "learning_rate": 7.771900831765803e-06,
      "loss": 0.2921,
      "num_input_tokens_seen": 158467064,
      "step": 2750,
      "train_runtime": 15638.5997,
      "train_tokens_per_second": 10133.072
    },
    {
      "epoch": 0.6272768670309654,
      "grad_norm": 0.6461671347278248,
      "learning_rate": 7.764454938874252e-06,
      "loss": 0.3274,
      "num_input_tokens_seen": 158783080,
      "step": 2755,
      "train_runtime": 15675.0615,
      "train_tokens_per_second": 10129.662
    },
    {
      "epoch": 0.6284153005464481,
      "grad_norm": 0.6166018179516913,
      "learning_rate": 7.757000205733671e-06,
      "loss": 0.3388,
      "num_input_tokens_seen": 159090880,
      "step": 2760,
      "train_runtime": 15702.5676,
      "train_tokens_per_second": 10131.52
    },
    {
      "epoch": 0.6295537340619308,
      "grad_norm": 0.6913871956155007,
      "learning_rate": 7.749536656183005e-06,
      "loss": 0.276,
      "num_input_tokens_seen": 159380512,
      "step": 2765,
      "train_runtime": 15726.7833,
      "train_tokens_per_second": 10134.336
    },
    {
      "epoch": 0.6306921675774135,
      "grad_norm": 0.6476055949655263,
      "learning_rate": 7.742064314089397e-06,
      "loss": 0.3209,
      "num_input_tokens_seen": 159666840,
      "step": 2770,
      "train_runtime": 15755.8974,
      "train_tokens_per_second": 10133.783
    },
    {
      "epoch": 0.6318306010928961,
      "grad_norm": 0.5788791999757573,
      "learning_rate": 7.734583203348103e-06,
      "loss": 0.3288,
      "num_input_tokens_seen": 159966216,
      "step": 2775,
      "train_runtime": 15784.019,
      "train_tokens_per_second": 10134.695
    },
    {
      "epoch": 0.6329690346083788,
      "grad_norm": 0.7188005964447671,
      "learning_rate": 7.727093347882425e-06,
      "loss": 0.3197,
      "num_input_tokens_seen": 160239288,
      "step": 2780,
      "train_runtime": 15816.4426,
      "train_tokens_per_second": 10131.184
    },
    {
      "epoch": 0.6341074681238615,
      "grad_norm": 0.49979177950097403,
      "learning_rate": 7.719594771643623e-06,
      "loss": 0.2843,
      "num_input_tokens_seen": 160539440,
      "step": 2785,
      "train_runtime": 15845.7207,
      "train_tokens_per_second": 10131.407
    },
    {
      "epoch": 0.6352459016393442,
      "grad_norm": 0.661322373963692,
      "learning_rate": 7.712087498610852e-06,
      "loss": 0.292,
      "num_input_tokens_seen": 160819928,
      "step": 2790,
      "train_runtime": 15870.7667,
      "train_tokens_per_second": 10133.091
    },
    {
      "epoch": 0.636384335154827,
      "grad_norm": 0.5510876084310095,
      "learning_rate": 7.704571552791069e-06,
      "loss": 0.2982,
      "num_input_tokens_seen": 161113320,
      "step": 2795,
      "train_runtime": 15905.4087,
      "train_tokens_per_second": 10129.467
    },
    {
      "epoch": 0.6375227686703097,
      "grad_norm": 0.517744009695649,
      "learning_rate": 7.697046958218973e-06,
      "loss": 0.2839,
      "num_input_tokens_seen": 161421704,
      "step": 2800,
      "train_runtime": 15932.4593,
      "train_tokens_per_second": 10131.625
    },
    {
      "epoch": 0.6386612021857924,
      "grad_norm": 0.6031310823519022,
      "learning_rate": 7.689513738956914e-06,
      "loss": 0.3148,
      "num_input_tokens_seen": 161708936,
      "step": 2805,
      "train_runtime": 15961.6426,
      "train_tokens_per_second": 10131.096
    },
    {
      "epoch": 0.6397996357012751,
      "grad_norm": 0.6078276132457395,
      "learning_rate": 7.681971919094825e-06,
      "loss": 0.3388,
      "num_input_tokens_seen": 162001344,
      "step": 2810,
      "train_runtime": 15989.6425,
      "train_tokens_per_second": 10131.643
    },
    {
      "epoch": 0.6409380692167578,
      "grad_norm": 0.5997235578488721,
      "learning_rate": 7.674421522750144e-06,
      "loss": 0.3261,
      "num_input_tokens_seen": 162324096,
      "step": 2815,
      "train_runtime": 16012.8611,
      "train_tokens_per_second": 10137.108
    },
    {
      "epoch": 0.6420765027322405,
      "grad_norm": 0.5619436837632302,
      "learning_rate": 7.666862574067732e-06,
      "loss": 0.3146,
      "num_input_tokens_seen": 162610184,
      "step": 2820,
      "train_runtime": 16045.0195,
      "train_tokens_per_second": 10134.621
    },
    {
      "epoch": 0.6432149362477231,
      "grad_norm": 0.5572382400877691,
      "learning_rate": 7.659295097219802e-06,
      "loss": 0.2884,
      "num_input_tokens_seen": 162900632,
      "step": 2825,
      "train_runtime": 16075.5932,
      "train_tokens_per_second": 10133.413
    },
    {
      "epoch": 0.6443533697632058,
      "grad_norm": 0.6518133793433621,
      "learning_rate": 7.651719116405836e-06,
      "loss": 0.3425,
      "num_input_tokens_seen": 163200104,
      "step": 2830,
      "train_runtime": 16109.6442,
      "train_tokens_per_second": 10130.584
    },
    {
      "epoch": 0.6454918032786885,
      "grad_norm": 0.6383156796788078,
      "learning_rate": 7.64413465585251e-06,
      "loss": 0.3215,
      "num_input_tokens_seen": 163482032,
      "step": 2835,
      "train_runtime": 16143.8354,
      "train_tokens_per_second": 10126.592
    },
    {
      "epoch": 0.6466302367941712,
      "grad_norm": 0.5619824161520329,
      "learning_rate": 7.636541739813619e-06,
      "loss": 0.2899,
      "num_input_tokens_seen": 163772800,
      "step": 2840,
      "train_runtime": 16178.4095,
      "train_tokens_per_second": 10122.923
    },
    {
      "epoch": 0.6477686703096539,
      "grad_norm": 0.6102419900559597,
      "learning_rate": 7.628940392569995e-06,
      "loss": 0.3047,
      "num_input_tokens_seen": 164058312,
      "step": 2845,
      "train_runtime": 16206.9412,
      "train_tokens_per_second": 10122.719
    },
    {
      "epoch": 0.6489071038251366,
      "grad_norm": 0.7562027671392288,
      "learning_rate": 7.621330638429434e-06,
      "loss": 0.304,
      "num_input_tokens_seen": 164348776,
      "step": 2850,
      "train_runtime": 16239.8242,
      "train_tokens_per_second": 10120.108
    },
    {
      "epoch": 0.6500455373406193,
      "grad_norm": 0.6217967338311484,
      "learning_rate": 7.613712501726614e-06,
      "loss": 0.3297,
      "num_input_tokens_seen": 164636224,
      "step": 2855,
      "train_runtime": 16267.3004,
      "train_tokens_per_second": 10120.685
    },
    {
      "epoch": 0.651183970856102,
      "grad_norm": 1.1620140919143793,
      "learning_rate": 7.606086006823017e-06,
      "loss": 0.3048,
      "num_input_tokens_seen": 164915040,
      "step": 2860,
      "train_runtime": 16297.4883,
      "train_tokens_per_second": 10119.046
    },
    {
      "epoch": 0.6523224043715847,
      "grad_norm": 0.7041899282473861,
      "learning_rate": 7.598451178106857e-06,
      "loss": 0.2878,
      "num_input_tokens_seen": 165188752,
      "step": 2865,
      "train_runtime": 16329.8849,
      "train_tokens_per_second": 10115.733
    },
    {
      "epoch": 0.6534608378870674,
      "grad_norm": 0.5512243753776052,
      "learning_rate": 7.590808039992998e-06,
      "loss": 0.3485,
      "num_input_tokens_seen": 165469200,
      "step": 2870,
      "train_runtime": 16358.3631,
      "train_tokens_per_second": 10115.266
    },
    {
      "epoch": 0.6545992714025501,
      "grad_norm": 0.7353800815321357,
      "learning_rate": 7.5831566169228736e-06,
      "loss": 0.3116,
      "num_input_tokens_seen": 165770376,
      "step": 2875,
      "train_runtime": 16386.9466,
      "train_tokens_per_second": 10116.001
    },
    {
      "epoch": 0.6557377049180327,
      "grad_norm": 0.6481448896611102,
      "learning_rate": 7.57549693336441e-06,
      "loss": 0.3018,
      "num_input_tokens_seen": 166050328,
      "step": 2880,
      "train_runtime": 16411.6784,
      "train_tokens_per_second": 10117.815
    },
    {
      "epoch": 0.6568761384335154,
      "grad_norm": 0.6885465659938241,
      "learning_rate": 7.567829013811953e-06,
      "loss": 0.3136,
      "num_input_tokens_seen": 166328248,
      "step": 2885,
      "train_runtime": 16438.9427,
      "train_tokens_per_second": 10117.941
    },
    {
      "epoch": 0.6580145719489982,
      "grad_norm": 0.5909647811509834,
      "learning_rate": 7.560152882786184e-06,
      "loss": 0.3232,
      "num_input_tokens_seen": 166600688,
      "step": 2890,
      "train_runtime": 16469.3455,
      "train_tokens_per_second": 10115.805
    },
    {
      "epoch": 0.6591530054644809,
      "grad_norm": 0.6527180414898944,
      "learning_rate": 7.552468564834044e-06,
      "loss": 0.3122,
      "num_input_tokens_seen": 166895184,
      "step": 2895,
      "train_runtime": 16498.7941,
      "train_tokens_per_second": 10115.599
    },
    {
      "epoch": 0.6602914389799636,
      "grad_norm": 0.5975184270304442,
      "learning_rate": 7.54477608452865e-06,
      "loss": 0.3135,
      "num_input_tokens_seen": 167212104,
      "step": 2900,
      "train_runtime": 16523.6424,
      "train_tokens_per_second": 10119.567
    },
    {
      "epoch": 0.6614298724954463,
      "grad_norm": 0.5367622050770575,
      "learning_rate": 7.537075466469228e-06,
      "loss": 0.3342,
      "num_input_tokens_seen": 167505800,
      "step": 2905,
      "train_runtime": 16550.0819,
      "train_tokens_per_second": 10121.146
    },
    {
      "epoch": 0.662568306010929,
      "grad_norm": 0.6621156921948675,
      "learning_rate": 7.529366735281023e-06,
      "loss": 0.3268,
      "num_input_tokens_seen": 167795768,
      "step": 2910,
      "train_runtime": 16578.3743,
      "train_tokens_per_second": 10121.364
    },
    {
      "epoch": 0.6637067395264117,
      "grad_norm": 0.7451878843517429,
      "learning_rate": 7.521649915615222e-06,
      "loss": 0.3318,
      "num_input_tokens_seen": 168055592,
      "step": 2915,
      "train_runtime": 16606.6148,
      "train_tokens_per_second": 10119.798
    },
    {
      "epoch": 0.6648451730418944,
      "grad_norm": 0.5561964519085411,
      "learning_rate": 7.513925032148883e-06,
      "loss": 0.31,
      "num_input_tokens_seen": 168343328,
      "step": 2920,
      "train_runtime": 16640.2585,
      "train_tokens_per_second": 10116.629
    },
    {
      "epoch": 0.6659836065573771,
      "grad_norm": 0.6649794809165146,
      "learning_rate": 7.506192109584847e-06,
      "loss": 0.3285,
      "num_input_tokens_seen": 168592000,
      "step": 2925,
      "train_runtime": 16668.2645,
      "train_tokens_per_second": 10114.55
    },
    {
      "epoch": 0.6671220400728597,
      "grad_norm": 0.607138183072641,
      "learning_rate": 7.498451172651664e-06,
      "loss": 0.3204,
      "num_input_tokens_seen": 168885864,
      "step": 2930,
      "train_runtime": 16701.7194,
      "train_tokens_per_second": 10111.885
    },
    {
      "epoch": 0.6682604735883424,
      "grad_norm": 0.6414949932942994,
      "learning_rate": 7.4907022461035125e-06,
      "loss": 0.3647,
      "num_input_tokens_seen": 169176608,
      "step": 2935,
      "train_runtime": 16738.0875,
      "train_tokens_per_second": 10107.284
    },
    {
      "epoch": 0.6693989071038251,
      "grad_norm": 0.642139054569875,
      "learning_rate": 7.48294535472012e-06,
      "loss": 0.3094,
      "num_input_tokens_seen": 169484696,
      "step": 2940,
      "train_runtime": 16763.6809,
      "train_tokens_per_second": 10110.232
    },
    {
      "epoch": 0.6705373406193078,
      "grad_norm": 0.6326603345178383,
      "learning_rate": 7.475180523306684e-06,
      "loss": 0.3069,
      "num_input_tokens_seen": 169761824,
      "step": 2945,
      "train_runtime": 16790.1836,
      "train_tokens_per_second": 10110.778
    },
    {
      "epoch": 0.6716757741347905,
      "grad_norm": 0.6354686733224235,
      "learning_rate": 7.467407776693791e-06,
      "loss": 0.3242,
      "num_input_tokens_seen": 170049328,
      "step": 2950,
      "train_runtime": 16812.7757,
      "train_tokens_per_second": 10114.292
    },
    {
      "epoch": 0.6728142076502732,
      "grad_norm": 0.5282575412910032,
      "learning_rate": 7.459627139737346e-06,
      "loss": 0.2819,
      "num_input_tokens_seen": 170345040,
      "step": 2955,
      "train_runtime": 16846.9739,
      "train_tokens_per_second": 10111.314
    },
    {
      "epoch": 0.6739526411657559,
      "grad_norm": 0.6060014313273004,
      "learning_rate": 7.451838637318478e-06,
      "loss": 0.3209,
      "num_input_tokens_seen": 170651832,
      "step": 2960,
      "train_runtime": 16878.2633,
      "train_tokens_per_second": 10110.746
    },
    {
      "epoch": 0.6750910746812386,
      "grad_norm": 0.6668172928663915,
      "learning_rate": 7.44404229434347e-06,
      "loss": 0.3195,
      "num_input_tokens_seen": 170936736,
      "step": 2965,
      "train_runtime": 16905.223,
      "train_tokens_per_second": 10111.475
    },
    {
      "epoch": 0.6762295081967213,
      "grad_norm": 0.6175392944291547,
      "learning_rate": 7.436238135743682e-06,
      "loss": 0.3165,
      "num_input_tokens_seen": 171226928,
      "step": 2970,
      "train_runtime": 16937.7721,
      "train_tokens_per_second": 10109.177
    },
    {
      "epoch": 0.677367941712204,
      "grad_norm": 0.6203596927268931,
      "learning_rate": 7.428426186475462e-06,
      "loss": 0.3261,
      "num_input_tokens_seen": 171521736,
      "step": 2975,
      "train_runtime": 16964.6981,
      "train_tokens_per_second": 10110.509
    },
    {
      "epoch": 0.6785063752276868,
      "grad_norm": 0.5629437290089434,
      "learning_rate": 7.420606471520076e-06,
      "loss": 0.3061,
      "num_input_tokens_seen": 171823896,
      "step": 2980,
      "train_runtime": 16994.7973,
      "train_tokens_per_second": 10110.382
    },
    {
      "epoch": 0.6796448087431693,
      "grad_norm": 0.5878152260896369,
      "learning_rate": 7.412779015883617e-06,
      "loss": 0.3067,
      "num_input_tokens_seen": 172128280,
      "step": 2985,
      "train_runtime": 17023.4894,
      "train_tokens_per_second": 10111.222
    },
    {
      "epoch": 0.680783242258652,
      "grad_norm": 0.4971911666611716,
      "learning_rate": 7.404943844596939e-06,
      "loss": 0.2849,
      "num_input_tokens_seen": 172419792,
      "step": 2990,
      "train_runtime": 17050.7486,
      "train_tokens_per_second": 10112.154
    },
    {
      "epoch": 0.6819216757741348,
      "grad_norm": 0.5445981238527163,
      "learning_rate": 7.397100982715563e-06,
      "loss": 0.2879,
      "num_input_tokens_seen": 172704192,
      "step": 2995,
      "train_runtime": 17076.8855,
      "train_tokens_per_second": 10113.331
    },
    {
      "epoch": 0.6830601092896175,
      "grad_norm": 0.6652179881991633,
      "learning_rate": 7.389250455319605e-06,
      "loss": 0.3079,
      "num_input_tokens_seen": 172988016,
      "step": 3000,
      "train_runtime": 17107.3039,
      "train_tokens_per_second": 10111.939
    },
    {
      "epoch": 0.6841985428051002,
      "grad_norm": 0.6543396659077138,
      "learning_rate": 7.381392287513699e-06,
      "loss": 0.3188,
      "num_input_tokens_seen": 173280856,
      "step": 3005,
      "train_runtime": 17135.8319,
      "train_tokens_per_second": 10112.194
    },
    {
      "epoch": 0.6853369763205829,
      "grad_norm": 0.6223450993796074,
      "learning_rate": 7.3735265044269014e-06,
      "loss": 0.3233,
      "num_input_tokens_seen": 173536368,
      "step": 3010,
      "train_runtime": 17161.8815,
      "train_tokens_per_second": 10111.733
    },
    {
      "epoch": 0.6864754098360656,
      "grad_norm": 0.5884268108326638,
      "learning_rate": 7.365653131212629e-06,
      "loss": 0.3038,
      "num_input_tokens_seen": 173825440,
      "step": 3015,
      "train_runtime": 17189.3168,
      "train_tokens_per_second": 10112.411
    },
    {
      "epoch": 0.6876138433515483,
      "grad_norm": 0.6521143046305868,
      "learning_rate": 7.35777219304857e-06,
      "loss": 0.3025,
      "num_input_tokens_seen": 174125680,
      "step": 3020,
      "train_runtime": 17215.8322,
      "train_tokens_per_second": 10114.276
    },
    {
      "epoch": 0.688752276867031,
      "grad_norm": 0.5696352190433651,
      "learning_rate": 7.349883715136601e-06,
      "loss": 0.2921,
      "num_input_tokens_seen": 174422040,
      "step": 3025,
      "train_runtime": 17244.785,
      "train_tokens_per_second": 10114.48
    },
    {
      "epoch": 0.6898907103825137,
      "grad_norm": 0.5757744417598726,
      "learning_rate": 7.341987722702711e-06,
      "loss": 0.3104,
      "num_input_tokens_seen": 174721576,
      "step": 3030,
      "train_runtime": 17271.5105,
      "train_tokens_per_second": 10116.172
    },
    {
      "epoch": 0.6910291438979964,
      "grad_norm": 0.5028019034918819,
      "learning_rate": 7.334084240996918e-06,
      "loss": 0.294,
      "num_input_tokens_seen": 175029488,
      "step": 3035,
      "train_runtime": 17296.5249,
      "train_tokens_per_second": 10119.344
    },
    {
      "epoch": 0.692167577413479,
      "grad_norm": 0.5731843604609205,
      "learning_rate": 7.326173295293192e-06,
      "loss": 0.2854,
      "num_input_tokens_seen": 175315544,
      "step": 3040,
      "train_runtime": 17328.0268,
      "train_tokens_per_second": 10117.456
    },
    {
      "epoch": 0.6933060109289617,
      "grad_norm": 0.5902765683371588,
      "learning_rate": 7.3182549108893685e-06,
      "loss": 0.3043,
      "num_input_tokens_seen": 175603728,
      "step": 3045,
      "train_runtime": 17351.7691,
      "train_tokens_per_second": 10120.22
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 0.6544281626349107,
      "learning_rate": 7.310329113107073e-06,
      "loss": 0.3264,
      "num_input_tokens_seen": 175880000,
      "step": 3050,
      "train_runtime": 17385.4056,
      "train_tokens_per_second": 10116.531
    },
    {
      "epoch": 0.6955828779599271,
      "grad_norm": 0.7180798139404928,
      "learning_rate": 7.302395927291636e-06,
      "loss": 0.3206,
      "num_input_tokens_seen": 176190936,
      "step": 3055,
      "train_runtime": 17413.013,
      "train_tokens_per_second": 10118.349
    },
    {
      "epoch": 0.6967213114754098,
      "grad_norm": 0.6710467109971736,
      "learning_rate": 7.294455378812016e-06,
      "loss": 0.3043,
      "num_input_tokens_seen": 176441568,
      "step": 3060,
      "train_runtime": 17443.0954,
      "train_tokens_per_second": 10115.267
    },
    {
      "epoch": 0.6978597449908925,
      "grad_norm": 0.635515161504378,
      "learning_rate": 7.286507493060712e-06,
      "loss": 0.3095,
      "num_input_tokens_seen": 176738544,
      "step": 3065,
      "train_runtime": 17473.0097,
      "train_tokens_per_second": 10114.946
    },
    {
      "epoch": 0.6989981785063752,
      "grad_norm": 0.505756659125469,
      "learning_rate": 7.27855229545369e-06,
      "loss": 0.2996,
      "num_input_tokens_seen": 177053368,
      "step": 3070,
      "train_runtime": 17498.8397,
      "train_tokens_per_second": 10118.006
    },
    {
      "epoch": 0.700136612021858,
      "grad_norm": 0.6131637139833875,
      "learning_rate": 7.270589811430296e-06,
      "loss": 0.2894,
      "num_input_tokens_seen": 177361512,
      "step": 3075,
      "train_runtime": 17527.785,
      "train_tokens_per_second": 10118.878
    },
    {
      "epoch": 0.7012750455373407,
      "grad_norm": 0.6609405539424278,
      "learning_rate": 7.262620066453179e-06,
      "loss": 0.3172,
      "num_input_tokens_seen": 177615736,
      "step": 3080,
      "train_runtime": 17551.4944,
      "train_tokens_per_second": 10119.693
    },
    {
      "epoch": 0.7024134790528234,
      "grad_norm": 0.7457578371710164,
      "learning_rate": 7.254643086008206e-06,
      "loss": 0.3112,
      "num_input_tokens_seen": 177885856,
      "step": 3085,
      "train_runtime": 17574.9354,
      "train_tokens_per_second": 10121.565
    },
    {
      "epoch": 0.703551912568306,
      "grad_norm": 0.6627220275815645,
      "learning_rate": 7.246658895604379e-06,
      "loss": 0.2974,
      "num_input_tokens_seen": 178185104,
      "step": 3090,
      "train_runtime": 17601.8805,
      "train_tokens_per_second": 10123.072
    },
    {
      "epoch": 0.7046903460837887,
      "grad_norm": 0.6440271370388726,
      "learning_rate": 7.238667520773761e-06,
      "loss": 0.3039,
      "num_input_tokens_seen": 178487600,
      "step": 3095,
      "train_runtime": 17627.0792,
      "train_tokens_per_second": 10125.761
    },
    {
      "epoch": 0.7058287795992714,
      "grad_norm": 0.5418919114572633,
      "learning_rate": 7.230668987071387e-06,
      "loss": 0.2888,
      "num_input_tokens_seen": 178776616,
      "step": 3100,
      "train_runtime": 17648.3147,
      "train_tokens_per_second": 10129.954
    },
    {
      "epoch": 0.7069672131147541,
      "grad_norm": 0.578493456928482,
      "learning_rate": 7.222663320075185e-06,
      "loss": 0.3012,
      "num_input_tokens_seen": 179053952,
      "step": 3105,
      "train_runtime": 17675.2044,
      "train_tokens_per_second": 10130.234
    },
    {
      "epoch": 0.7081056466302368,
      "grad_norm": 0.5670756799278618,
      "learning_rate": 7.214650545385893e-06,
      "loss": 0.3247,
      "num_input_tokens_seen": 179356120,
      "step": 3110,
      "train_runtime": 17702.9595,
      "train_tokens_per_second": 10131.42
    },
    {
      "epoch": 0.7092440801457195,
      "grad_norm": 0.5898246832354967,
      "learning_rate": 7.206630688626981e-06,
      "loss": 0.327,
      "num_input_tokens_seen": 179663008,
      "step": 3115,
      "train_runtime": 17733.3417,
      "train_tokens_per_second": 10131.368
    },
    {
      "epoch": 0.7103825136612022,
      "grad_norm": 0.5518816036611334,
      "learning_rate": 7.198603775444565e-06,
      "loss": 0.3119,
      "num_input_tokens_seen": 179939264,
      "step": 3120,
      "train_runtime": 17757.9093,
      "train_tokens_per_second": 10132.908
    },
    {
      "epoch": 0.7115209471766849,
      "grad_norm": 0.7513881064035125,
      "learning_rate": 7.190569831507323e-06,
      "loss": 0.3012,
      "num_input_tokens_seen": 180213680,
      "step": 3125,
      "train_runtime": 17788.2066,
      "train_tokens_per_second": 10131.076
    },
    {
      "epoch": 0.7126593806921676,
      "grad_norm": 0.6142080166622692,
      "learning_rate": 7.182528882506423e-06,
      "loss": 0.3089,
      "num_input_tokens_seen": 180506320,
      "step": 3130,
      "train_runtime": 17819.5058,
      "train_tokens_per_second": 10129.704
    },
    {
      "epoch": 0.7137978142076503,
      "grad_norm": 0.4885780840289198,
      "learning_rate": 7.174480954155427e-06,
      "loss": 0.2927,
      "num_input_tokens_seen": 180772088,
      "step": 3135,
      "train_runtime": 17850.9852,
      "train_tokens_per_second": 10126.729
    },
    {
      "epoch": 0.714936247723133,
      "grad_norm": 0.515658401697138,
      "learning_rate": 7.166426072190218e-06,
      "loss": 0.2886,
      "num_input_tokens_seen": 181081720,
      "step": 3140,
      "train_runtime": 17880.5714,
      "train_tokens_per_second": 10127.289
    },
    {
      "epoch": 0.7160746812386156,
      "grad_norm": 0.6104495392213235,
      "learning_rate": 7.158364262368919e-06,
      "loss": 0.2976,
      "num_input_tokens_seen": 181370568,
      "step": 3145,
      "train_runtime": 17912.2646,
      "train_tokens_per_second": 10125.496
    },
    {
      "epoch": 0.7172131147540983,
      "grad_norm": 0.5938566563205764,
      "learning_rate": 7.150295550471803e-06,
      "loss": 0.3359,
      "num_input_tokens_seen": 181672584,
      "step": 3150,
      "train_runtime": 17934.7578,
      "train_tokens_per_second": 10129.637
    },
    {
      "epoch": 0.718351548269581,
      "grad_norm": 0.5958185860164961,
      "learning_rate": 7.142219962301215e-06,
      "loss": 0.3132,
      "num_input_tokens_seen": 181951704,
      "step": 3155,
      "train_runtime": 17959.8965,
      "train_tokens_per_second": 10131.0
    },
    {
      "epoch": 0.7194899817850637,
      "grad_norm": 0.5680207938959156,
      "learning_rate": 7.134137523681491e-06,
      "loss": 0.3084,
      "num_input_tokens_seen": 182226488,
      "step": 3160,
      "train_runtime": 17990.9905,
      "train_tokens_per_second": 10128.764
    },
    {
      "epoch": 0.7206284153005464,
      "grad_norm": 0.6119495853380654,
      "learning_rate": 7.126048260458874e-06,
      "loss": 0.3092,
      "num_input_tokens_seen": 182517304,
      "step": 3165,
      "train_runtime": 18020.7488,
      "train_tokens_per_second": 10128.175
    },
    {
      "epoch": 0.7217668488160291,
      "grad_norm": 0.539866350643605,
      "learning_rate": 7.117952198501427e-06,
      "loss": 0.3168,
      "num_input_tokens_seen": 182814976,
      "step": 3170,
      "train_runtime": 18047.5385,
      "train_tokens_per_second": 10129.635
    },
    {
      "epoch": 0.7229052823315119,
      "grad_norm": 0.5935855127975856,
      "learning_rate": 7.109849363698958e-06,
      "loss": 0.3194,
      "num_input_tokens_seen": 183102048,
      "step": 3175,
      "train_runtime": 18072.9316,
      "train_tokens_per_second": 10131.287
    },
    {
      "epoch": 0.7240437158469946,
      "grad_norm": 0.619977553778781,
      "learning_rate": 7.101739781962929e-06,
      "loss": 0.3341,
      "num_input_tokens_seen": 183392632,
      "step": 3180,
      "train_runtime": 18099.6929,
      "train_tokens_per_second": 10132.362
    },
    {
      "epoch": 0.7251821493624773,
      "grad_norm": 0.642278578165567,
      "learning_rate": 7.093623479226386e-06,
      "loss": 0.3099,
      "num_input_tokens_seen": 183691936,
      "step": 3185,
      "train_runtime": 18127.0839,
      "train_tokens_per_second": 10133.562
    },
    {
      "epoch": 0.72632058287796,
      "grad_norm": 0.6373844440318492,
      "learning_rate": 7.085500481443858e-06,
      "loss": 0.3387,
      "num_input_tokens_seen": 183974520,
      "step": 3190,
      "train_runtime": 18154.5281,
      "train_tokens_per_second": 10133.809
    },
    {
      "epoch": 0.7274590163934426,
      "grad_norm": 0.6085862169297604,
      "learning_rate": 7.077370814591289e-06,
      "loss": 0.3198,
      "num_input_tokens_seen": 184239936,
      "step": 3195,
      "train_runtime": 18185.0038,
      "train_tokens_per_second": 10131.421
    },
    {
      "epoch": 0.7285974499089253,
      "grad_norm": 0.5993784981707653,
      "learning_rate": 7.06923450466595e-06,
      "loss": 0.2826,
      "num_input_tokens_seen": 184523968,
      "step": 3200,
      "train_runtime": 18204.291,
      "train_tokens_per_second": 10136.29
    },
    {
      "epoch": 0.729735883424408,
      "grad_norm": 0.6306610152659253,
      "learning_rate": 7.061091577686349e-06,
      "loss": 0.3265,
      "num_input_tokens_seen": 184795080,
      "step": 3205,
      "train_runtime": 18230.8421,
      "train_tokens_per_second": 10136.398
    },
    {
      "epoch": 0.7308743169398907,
      "grad_norm": 0.6206249818335954,
      "learning_rate": 7.052942059692164e-06,
      "loss": 0.3254,
      "num_input_tokens_seen": 185085536,
      "step": 3210,
      "train_runtime": 18262.0153,
      "train_tokens_per_second": 10135.001
    },
    {
      "epoch": 0.7320127504553734,
      "grad_norm": 0.6604986241098448,
      "learning_rate": 7.044785976744143e-06,
      "loss": 0.3063,
      "num_input_tokens_seen": 185345616,
      "step": 3215,
      "train_runtime": 18287.9229,
      "train_tokens_per_second": 10134.864
    },
    {
      "epoch": 0.7331511839708561,
      "grad_norm": 0.6436761349991478,
      "learning_rate": 7.036623354924029e-06,
      "loss": 0.3416,
      "num_input_tokens_seen": 185643376,
      "step": 3220,
      "train_runtime": 18311.1316,
      "train_tokens_per_second": 10138.28
    },
    {
      "epoch": 0.7342896174863388,
      "grad_norm": 0.6767319686935755,
      "learning_rate": 7.028454220334476e-06,
      "loss": 0.302,
      "num_input_tokens_seen": 185919736,
      "step": 3225,
      "train_runtime": 18342.8067,
      "train_tokens_per_second": 10135.839
    },
    {
      "epoch": 0.7354280510018215,
      "grad_norm": 0.6255735748116643,
      "learning_rate": 7.020278599098964e-06,
      "loss": 0.3126,
      "num_input_tokens_seen": 186236096,
      "step": 3230,
      "train_runtime": 18373.1854,
      "train_tokens_per_second": 10136.299
    },
    {
      "epoch": 0.7365664845173042,
      "grad_norm": 0.703856727210415,
      "learning_rate": 7.012096517361716e-06,
      "loss": 0.3117,
      "num_input_tokens_seen": 186501496,
      "step": 3235,
      "train_runtime": 18398.4768,
      "train_tokens_per_second": 10136.79
    },
    {
      "epoch": 0.7377049180327869,
      "grad_norm": 0.5787144156816512,
      "learning_rate": 7.003908001287617e-06,
      "loss": 0.3172,
      "num_input_tokens_seen": 186801376,
      "step": 3240,
      "train_runtime": 18423.6112,
      "train_tokens_per_second": 10139.238
    },
    {
      "epoch": 0.7388433515482696,
      "grad_norm": 0.6672222873760523,
      "learning_rate": 6.995713077062123e-06,
      "loss": 0.29,
      "num_input_tokens_seen": 187086544,
      "step": 3245,
      "train_runtime": 18445.9598,
      "train_tokens_per_second": 10142.413
    },
    {
      "epoch": 0.7399817850637522,
      "grad_norm": 0.6003508477342633,
      "learning_rate": 6.987511770891187e-06,
      "loss": 0.3222,
      "num_input_tokens_seen": 187379648,
      "step": 3250,
      "train_runtime": 18475.0804,
      "train_tokens_per_second": 10142.291
    },
    {
      "epoch": 0.7411202185792349,
      "grad_norm": 0.561299992973021,
      "learning_rate": 6.979304109001168e-06,
      "loss": 0.3104,
      "num_input_tokens_seen": 187664232,
      "step": 3255,
      "train_runtime": 18507.5013,
      "train_tokens_per_second": 10139.901
    },
    {
      "epoch": 0.7422586520947176,
      "grad_norm": 0.6733027873777016,
      "learning_rate": 6.9710901176387515e-06,
      "loss": 0.3252,
      "num_input_tokens_seen": 187927968,
      "step": 3260,
      "train_runtime": 18539.935,
      "train_tokens_per_second": 10136.388
    },
    {
      "epoch": 0.7433970856102003,
      "grad_norm": 0.6479550752232683,
      "learning_rate": 6.9628698230708604e-06,
      "loss": 0.2988,
      "num_input_tokens_seen": 188204056,
      "step": 3265,
      "train_runtime": 18572.2153,
      "train_tokens_per_second": 10133.635
    },
    {
      "epoch": 0.744535519125683,
      "grad_norm": 0.7354723096527379,
      "learning_rate": 6.954643251584575e-06,
      "loss": 0.2899,
      "num_input_tokens_seen": 188493480,
      "step": 3270,
      "train_runtime": 18605.3199,
      "train_tokens_per_second": 10131.16
    },
    {
      "epoch": 0.7456739526411658,
      "grad_norm": 0.6606689753287348,
      "learning_rate": 6.9464104294870504e-06,
      "loss": 0.3177,
      "num_input_tokens_seen": 188755368,
      "step": 3275,
      "train_runtime": 18635.0652,
      "train_tokens_per_second": 10129.043
    },
    {
      "epoch": 0.7468123861566485,
      "grad_norm": 0.5896315302542207,
      "learning_rate": 6.938171383105428e-06,
      "loss": 0.31,
      "num_input_tokens_seen": 189066560,
      "step": 3280,
      "train_runtime": 18666.6147,
      "train_tokens_per_second": 10128.594
    },
    {
      "epoch": 0.7479508196721312,
      "grad_norm": 0.6125748421465316,
      "learning_rate": 6.929926138786753e-06,
      "loss": 0.2963,
      "num_input_tokens_seen": 189331608,
      "step": 3285,
      "train_runtime": 18691.7969,
      "train_tokens_per_second": 10129.128
    },
    {
      "epoch": 0.7490892531876139,
      "grad_norm": 0.6309657124528562,
      "learning_rate": 6.921674722897893e-06,
      "loss": 0.312,
      "num_input_tokens_seen": 189634192,
      "step": 3290,
      "train_runtime": 18719.731,
      "train_tokens_per_second": 10130.177
    },
    {
      "epoch": 0.7502276867030966,
      "grad_norm": 0.5764601754252722,
      "learning_rate": 6.913417161825449e-06,
      "loss": 0.3232,
      "num_input_tokens_seen": 189939704,
      "step": 3295,
      "train_runtime": 18746.3468,
      "train_tokens_per_second": 10132.092
    },
    {
      "epoch": 0.7513661202185792,
      "grad_norm": 0.6089422562631612,
      "learning_rate": 6.905153481975675e-06,
      "loss": 0.3256,
      "num_input_tokens_seen": 190232728,
      "step": 3300,
      "train_runtime": 18771.5565,
      "train_tokens_per_second": 10134.095
    },
    {
      "epoch": 0.7525045537340619,
      "grad_norm": 0.6247708820003036,
      "learning_rate": 6.896883709774389e-06,
      "loss": 0.3403,
      "num_input_tokens_seen": 190527824,
      "step": 3305,
      "train_runtime": 18799.1766,
      "train_tokens_per_second": 10134.903
    },
    {
      "epoch": 0.7536429872495446,
      "grad_norm": 0.6543918787982306,
      "learning_rate": 6.888607871666893e-06,
      "loss": 0.3208,
      "num_input_tokens_seen": 190807248,
      "step": 3310,
      "train_runtime": 18825.0542,
      "train_tokens_per_second": 10135.814
    },
    {
      "epoch": 0.7547814207650273,
      "grad_norm": 0.6106047304215955,
      "learning_rate": 6.8803259941178876e-06,
      "loss": 0.3186,
      "num_input_tokens_seen": 191100736,
      "step": 3315,
      "train_runtime": 18856.299,
      "train_tokens_per_second": 10134.583
    },
    {
      "epoch": 0.75591985428051,
      "grad_norm": 0.6163795947708982,
      "learning_rate": 6.872038103611386e-06,
      "loss": 0.2977,
      "num_input_tokens_seen": 191375976,
      "step": 3320,
      "train_runtime": 18884.7615,
      "train_tokens_per_second": 10133.884
    },
    {
      "epoch": 0.7570582877959927,
      "grad_norm": 0.5868523153478404,
      "learning_rate": 6.863744226650628e-06,
      "loss": 0.313,
      "num_input_tokens_seen": 191659760,
      "step": 3325,
      "train_runtime": 18917.7254,
      "train_tokens_per_second": 10131.226
    },
    {
      "epoch": 0.7581967213114754,
      "grad_norm": 0.6087690546439798,
      "learning_rate": 6.855444389758002e-06,
      "loss": 0.3172,
      "num_input_tokens_seen": 191938680,
      "step": 3330,
      "train_runtime": 18939.7451,
      "train_tokens_per_second": 10134.174
    },
    {
      "epoch": 0.7593351548269581,
      "grad_norm": 0.6351000691527142,
      "learning_rate": 6.847138619474945e-06,
      "loss": 0.3287,
      "num_input_tokens_seen": 192237112,
      "step": 3335,
      "train_runtime": 18972.4528,
      "train_tokens_per_second": 10132.433
    },
    {
      "epoch": 0.7604735883424408,
      "grad_norm": 0.5880905729036928,
      "learning_rate": 6.838826942361881e-06,
      "loss": 0.3067,
      "num_input_tokens_seen": 192539528,
      "step": 3340,
      "train_runtime": 18995.7371,
      "train_tokens_per_second": 10135.934
    },
    {
      "epoch": 0.7616120218579235,
      "grad_norm": 0.5719064470759437,
      "learning_rate": 6.830509384998114e-06,
      "loss": 0.3004,
      "num_input_tokens_seen": 192858208,
      "step": 3345,
      "train_runtime": 19029.8574,
      "train_tokens_per_second": 10134.506
    },
    {
      "epoch": 0.7627504553734062,
      "grad_norm": 0.591321974913851,
      "learning_rate": 6.822185973981753e-06,
      "loss": 0.3575,
      "num_input_tokens_seen": 193150976,
      "step": 3350,
      "train_runtime": 19050.3744,
      "train_tokens_per_second": 10138.96
    },
    {
      "epoch": 0.7638888888888888,
      "grad_norm": 0.6698915773525833,
      "learning_rate": 6.813856735929632e-06,
      "loss": 0.3079,
      "num_input_tokens_seen": 193440296,
      "step": 3355,
      "train_runtime": 19079.3796,
      "train_tokens_per_second": 10138.71
    },
    {
      "epoch": 0.7650273224043715,
      "grad_norm": 0.5774799512479107,
      "learning_rate": 6.80552169747721e-06,
      "loss": 0.3015,
      "num_input_tokens_seen": 193732104,
      "step": 3360,
      "train_runtime": 19102.2449,
      "train_tokens_per_second": 10141.85
    },
    {
      "epoch": 0.7661657559198543,
      "grad_norm": 0.634923067937733,
      "learning_rate": 6.797180885278503e-06,
      "loss": 0.3388,
      "num_input_tokens_seen": 194023352,
      "step": 3365,
      "train_runtime": 19125.9009,
      "train_tokens_per_second": 10144.534
    },
    {
      "epoch": 0.767304189435337,
      "grad_norm": 0.6295769593425998,
      "learning_rate": 6.788834326005983e-06,
      "loss": 0.3354,
      "num_input_tokens_seen": 194323480,
      "step": 3370,
      "train_runtime": 19152.1376,
      "train_tokens_per_second": 10146.308
    },
    {
      "epoch": 0.7684426229508197,
      "grad_norm": 0.6165319598088169,
      "learning_rate": 6.780482046350507e-06,
      "loss": 0.3082,
      "num_input_tokens_seen": 194593816,
      "step": 3375,
      "train_runtime": 19181.4984,
      "train_tokens_per_second": 10144.87
    },
    {
      "epoch": 0.7695810564663024,
      "grad_norm": 0.603901208297879,
      "learning_rate": 6.772124073021219e-06,
      "loss": 0.3075,
      "num_input_tokens_seen": 194881504,
      "step": 3380,
      "train_runtime": 19214.6215,
      "train_tokens_per_second": 10142.355
    },
    {
      "epoch": 0.7707194899817851,
      "grad_norm": 0.5671182987288537,
      "learning_rate": 6.763760432745475e-06,
      "loss": 0.3459,
      "num_input_tokens_seen": 195149560,
      "step": 3385,
      "train_runtime": 19243.696,
      "train_tokens_per_second": 10140.96
    },
    {
      "epoch": 0.7718579234972678,
      "grad_norm": 0.5775191757227901,
      "learning_rate": 6.755391152268749e-06,
      "loss": 0.3101,
      "num_input_tokens_seen": 195443008,
      "step": 3390,
      "train_runtime": 19278.7399,
      "train_tokens_per_second": 10137.748
    },
    {
      "epoch": 0.7729963570127505,
      "grad_norm": 0.6872273465675454,
      "learning_rate": 6.747016258354555e-06,
      "loss": 0.3118,
      "num_input_tokens_seen": 195681120,
      "step": 3395,
      "train_runtime": 19308.685,
      "train_tokens_per_second": 10134.358
    },
    {
      "epoch": 0.7741347905282332,
      "grad_norm": 0.6982759231570427,
      "learning_rate": 6.738635777784356e-06,
      "loss": 0.3143,
      "num_input_tokens_seen": 195979720,
      "step": 3400,
      "train_runtime": 19332.9667,
      "train_tokens_per_second": 10137.074
    },
    {
      "epoch": 0.7752732240437158,
      "grad_norm": 0.5994476716457433,
      "learning_rate": 6.730249737357478e-06,
      "loss": 0.3078,
      "num_input_tokens_seen": 196254536,
      "step": 3405,
      "train_runtime": 19359.2365,
      "train_tokens_per_second": 10137.514
    },
    {
      "epoch": 0.7764116575591985,
      "grad_norm": 1.3156891135032425,
      "learning_rate": 6.721858163891032e-06,
      "loss": 0.2971,
      "num_input_tokens_seen": 196536912,
      "step": 3410,
      "train_runtime": 19387.5799,
      "train_tokens_per_second": 10137.259
    },
    {
      "epoch": 0.7775500910746812,
      "grad_norm": 0.597259134881506,
      "learning_rate": 6.713461084219814e-06,
      "loss": 0.3326,
      "num_input_tokens_seen": 196831232,
      "step": 3415,
      "train_runtime": 19419.5848,
      "train_tokens_per_second": 10135.708
    },
    {
      "epoch": 0.7786885245901639,
      "grad_norm": 0.6497059458619701,
      "learning_rate": 6.705058525196241e-06,
      "loss": 0.2981,
      "num_input_tokens_seen": 197107040,
      "step": 3420,
      "train_runtime": 19451.2249,
      "train_tokens_per_second": 10133.4
    },
    {
      "epoch": 0.7798269581056466,
      "grad_norm": 0.6233006071916825,
      "learning_rate": 6.696650513690238e-06,
      "loss": 0.3191,
      "num_input_tokens_seen": 197381208,
      "step": 3425,
      "train_runtime": 19478.7952,
      "train_tokens_per_second": 10133.132
    },
    {
      "epoch": 0.7809653916211293,
      "grad_norm": 0.6249138695976334,
      "learning_rate": 6.688237076589177e-06,
      "loss": 0.2829,
      "num_input_tokens_seen": 197665936,
      "step": 3430,
      "train_runtime": 19508.6688,
      "train_tokens_per_second": 10132.21
    },
    {
      "epoch": 0.782103825136612,
      "grad_norm": 0.5483449841011208,
      "learning_rate": 6.67981824079777e-06,
      "loss": 0.2846,
      "num_input_tokens_seen": 197971360,
      "step": 3435,
      "train_runtime": 19528.7848,
      "train_tokens_per_second": 10137.413
    },
    {
      "epoch": 0.7832422586520947,
      "grad_norm": 0.697344772932538,
      "learning_rate": 6.671394033238002e-06,
      "loss": 0.314,
      "num_input_tokens_seen": 198265768,
      "step": 3440,
      "train_runtime": 19561.1538,
      "train_tokens_per_second": 10135.689
    },
    {
      "epoch": 0.7843806921675774,
      "grad_norm": 0.4962987830587941,
      "learning_rate": 6.662964480849033e-06,
      "loss": 0.3082,
      "num_input_tokens_seen": 198545296,
      "step": 3445,
      "train_runtime": 19588.1595,
      "train_tokens_per_second": 10135.985
    },
    {
      "epoch": 0.7855191256830601,
      "grad_norm": 0.6453385251464692,
      "learning_rate": 6.654529610587114e-06,
      "loss": 0.316,
      "num_input_tokens_seen": 198831664,
      "step": 3450,
      "train_runtime": 19617.1074,
      "train_tokens_per_second": 10135.626
    },
    {
      "epoch": 0.7866575591985429,
      "grad_norm": 0.6125067726396957,
      "learning_rate": 6.6460894494255e-06,
      "loss": 0.3017,
      "num_input_tokens_seen": 199117744,
      "step": 3455,
      "train_runtime": 19650.2368,
      "train_tokens_per_second": 10133.096
    },
    {
      "epoch": 0.7877959927140255,
      "grad_norm": 0.5749132699608686,
      "learning_rate": 6.637644024354371e-06,
      "loss": 0.3089,
      "num_input_tokens_seen": 199405976,
      "step": 3460,
      "train_runtime": 19680.4099,
      "train_tokens_per_second": 10132.206
    },
    {
      "epoch": 0.7889344262295082,
      "grad_norm": 0.5518876826538658,
      "learning_rate": 6.629193362380734e-06,
      "loss": 0.2826,
      "num_input_tokens_seen": 199710424,
      "step": 3465,
      "train_runtime": 19707.2372,
      "train_tokens_per_second": 10133.862
    },
    {
      "epoch": 0.7900728597449909,
      "grad_norm": 0.6425007454695029,
      "learning_rate": 6.620737490528345e-06,
      "loss": 0.3211,
      "num_input_tokens_seen": 199993240,
      "step": 3470,
      "train_runtime": 19735.786,
      "train_tokens_per_second": 10133.533
    },
    {
      "epoch": 0.7912112932604736,
      "grad_norm": 0.6851715559385028,
      "learning_rate": 6.612276435837622e-06,
      "loss": 0.2973,
      "num_input_tokens_seen": 200291688,
      "step": 3475,
      "train_runtime": 19764.3699,
      "train_tokens_per_second": 10133.978
    },
    {
      "epoch": 0.7923497267759563,
      "grad_norm": 0.5827171430037213,
      "learning_rate": 6.603810225365554e-06,
      "loss": 0.3001,
      "num_input_tokens_seen": 200609608,
      "step": 3480,
      "train_runtime": 19785.63,
      "train_tokens_per_second": 10139.157
    },
    {
      "epoch": 0.793488160291439,
      "grad_norm": 0.7051053179423449,
      "learning_rate": 6.5953388861856205e-06,
      "loss": 0.3318,
      "num_input_tokens_seen": 200898168,
      "step": 3485,
      "train_runtime": 19815.4885,
      "train_tokens_per_second": 10138.441
    },
    {
      "epoch": 0.7946265938069217,
      "grad_norm": 0.5452812860945512,
      "learning_rate": 6.5868624453876984e-06,
      "loss": 0.291,
      "num_input_tokens_seen": 201198360,
      "step": 3490,
      "train_runtime": 19840.0582,
      "train_tokens_per_second": 10141.017
    },
    {
      "epoch": 0.7957650273224044,
      "grad_norm": 0.6906402783297478,
      "learning_rate": 6.57838093007798e-06,
      "loss": 0.3096,
      "num_input_tokens_seen": 201473728,
      "step": 3495,
      "train_runtime": 19866.4396,
      "train_tokens_per_second": 10141.411
    },
    {
      "epoch": 0.7969034608378871,
      "grad_norm": 0.527867443795774,
      "learning_rate": 6.569894367378884e-06,
      "loss": 0.3558,
      "num_input_tokens_seen": 201757776,
      "step": 3500,
      "train_runtime": 19897.9569,
      "train_tokens_per_second": 10139.623
    },
    {
      "epoch": 0.7980418943533698,
      "grad_norm": 0.5322932106947776,
      "learning_rate": 6.561402784428974e-06,
      "loss": 0.2753,
      "num_input_tokens_seen": 202053840,
      "step": 3505,
      "train_runtime": 19918.4554,
      "train_tokens_per_second": 10144.052
    },
    {
      "epoch": 0.7991803278688525,
      "grad_norm": 0.7295012594718194,
      "learning_rate": 6.552906208382863e-06,
      "loss": 0.3079,
      "num_input_tokens_seen": 202334368,
      "step": 3510,
      "train_runtime": 19946.8397,
      "train_tokens_per_second": 10143.68
    },
    {
      "epoch": 0.8003187613843351,
      "grad_norm": 0.6369088221664108,
      "learning_rate": 6.544404666411131e-06,
      "loss": 0.3187,
      "num_input_tokens_seen": 202625664,
      "step": 3515,
      "train_runtime": 19971.5706,
      "train_tokens_per_second": 10145.705
    },
    {
      "epoch": 0.8014571948998178,
      "grad_norm": 0.634061206650241,
      "learning_rate": 6.535898185700241e-06,
      "loss": 0.3318,
      "num_input_tokens_seen": 202922640,
      "step": 3520,
      "train_runtime": 20002.9946,
      "train_tokens_per_second": 10144.613
    },
    {
      "epoch": 0.8025956284153005,
      "grad_norm": 0.6648999303921186,
      "learning_rate": 6.52738679345245e-06,
      "loss": 0.2928,
      "num_input_tokens_seen": 203210376,
      "step": 3525,
      "train_runtime": 20029.6239,
      "train_tokens_per_second": 10145.491
    },
    {
      "epoch": 0.8037340619307832,
      "grad_norm": 0.7629510187380146,
      "learning_rate": 6.518870516885716e-06,
      "loss": 0.32,
      "num_input_tokens_seen": 203505112,
      "step": 3530,
      "train_runtime": 20059.9215,
      "train_tokens_per_second": 10144.861
    },
    {
      "epoch": 0.8048724954462659,
      "grad_norm": 0.5971016018213878,
      "learning_rate": 6.51034938323362e-06,
      "loss": 0.288,
      "num_input_tokens_seen": 203774552,
      "step": 3535,
      "train_runtime": 20083.8856,
      "train_tokens_per_second": 10146.172
    },
    {
      "epoch": 0.8060109289617486,
      "grad_norm": 0.6246756574678475,
      "learning_rate": 6.501823419745274e-06,
      "loss": 0.3217,
      "num_input_tokens_seen": 204066832,
      "step": 3540,
      "train_runtime": 20111.8675,
      "train_tokens_per_second": 10146.588
    },
    {
      "epoch": 0.8071493624772313,
      "grad_norm": 0.6551784257599976,
      "learning_rate": 6.493292653685237e-06,
      "loss": 0.2835,
      "num_input_tokens_seen": 204350680,
      "step": 3545,
      "train_runtime": 20140.7352,
      "train_tokens_per_second": 10146.138
    },
    {
      "epoch": 0.808287795992714,
      "grad_norm": 0.5708450987809898,
      "learning_rate": 6.484757112333424e-06,
      "loss": 0.3048,
      "num_input_tokens_seen": 204644904,
      "step": 3550,
      "train_runtime": 20168.9477,
      "train_tokens_per_second": 10146.534
    },
    {
      "epoch": 0.8094262295081968,
      "grad_norm": 0.5729087228183274,
      "learning_rate": 6.47621682298502e-06,
      "loss": 0.3141,
      "num_input_tokens_seen": 204949688,
      "step": 3555,
      "train_runtime": 20201.5033,
      "train_tokens_per_second": 10145.269
    },
    {
      "epoch": 0.8105646630236795,
      "grad_norm": 0.6577677953687233,
      "learning_rate": 6.4676718129503944e-06,
      "loss": 0.3072,
      "num_input_tokens_seen": 205236120,
      "step": 3560,
      "train_runtime": 20230.1245,
      "train_tokens_per_second": 10145.075
    },
    {
      "epoch": 0.8117030965391621,
      "grad_norm": 0.9185127521356087,
      "learning_rate": 6.459122109555011e-06,
      "loss": 0.312,
      "num_input_tokens_seen": 205502448,
      "step": 3565,
      "train_runtime": 20256.7267,
      "train_tokens_per_second": 10144.899
    },
    {
      "epoch": 0.8128415300546448,
      "grad_norm": 1.0491153560583755,
      "learning_rate": 6.450567740139345e-06,
      "loss": 0.2952,
      "num_input_tokens_seen": 205790168,
      "step": 3570,
      "train_runtime": 20286.8772,
      "train_tokens_per_second": 10144.004
    },
    {
      "epoch": 0.8139799635701275,
      "grad_norm": 0.8031178035634609,
      "learning_rate": 6.442008732058791e-06,
      "loss": 0.2838,
      "num_input_tokens_seen": 206085072,
      "step": 3575,
      "train_runtime": 20314.8285,
      "train_tokens_per_second": 10144.564
    },
    {
      "epoch": 0.8151183970856102,
      "grad_norm": 0.5732277552254513,
      "learning_rate": 6.433445112683577e-06,
      "loss": 0.3323,
      "num_input_tokens_seen": 206389104,
      "step": 3580,
      "train_runtime": 20343.4959,
      "train_tokens_per_second": 10145.213
    },
    {
      "epoch": 0.8162568306010929,
      "grad_norm": 0.641417793276911,
      "learning_rate": 6.4248769093986765e-06,
      "loss": 0.2928,
      "num_input_tokens_seen": 206673784,
      "step": 3585,
      "train_runtime": 20375.5,
      "train_tokens_per_second": 10143.25
    },
    {
      "epoch": 0.8173952641165756,
      "grad_norm": 0.632917164953292,
      "learning_rate": 6.416304149603725e-06,
      "loss": 0.3031,
      "num_input_tokens_seen": 206972456,
      "step": 3590,
      "train_runtime": 20398.5356,
      "train_tokens_per_second": 10146.437
    },
    {
      "epoch": 0.8185336976320583,
      "grad_norm": 0.7124934759744965,
      "learning_rate": 6.407726860712924e-06,
      "loss": 0.304,
      "num_input_tokens_seen": 207281136,
      "step": 3595,
      "train_runtime": 20428.8932,
      "train_tokens_per_second": 10146.469
    },
    {
      "epoch": 0.819672131147541,
      "grad_norm": 0.5096593171979907,
      "learning_rate": 6.399145070154962e-06,
      "loss": 0.3118,
      "num_input_tokens_seen": 207575912,
      "step": 3600,
      "train_runtime": 20457.0977,
      "train_tokens_per_second": 10146.89
    },
    {
      "epoch": 0.8208105646630237,
      "grad_norm": 0.6086952182719532,
      "learning_rate": 6.39055880537292e-06,
      "loss": 0.2925,
      "num_input_tokens_seen": 207863384,
      "step": 3605,
      "train_runtime": 20487.8137,
      "train_tokens_per_second": 10145.708
    },
    {
      "epoch": 0.8219489981785064,
      "grad_norm": 0.6489470284059234,
      "learning_rate": 6.38196809382419e-06,
      "loss": 0.321,
      "num_input_tokens_seen": 208154752,
      "step": 3610,
      "train_runtime": 20516.7821,
      "train_tokens_per_second": 10145.585
    },
    {
      "epoch": 0.8230874316939891,
      "grad_norm": 0.573869183016445,
      "learning_rate": 6.373372962980383e-06,
      "loss": 0.3195,
      "num_input_tokens_seen": 208457800,
      "step": 3615,
      "train_runtime": 20546.2501,
      "train_tokens_per_second": 10145.783
    },
    {
      "epoch": 0.8242258652094717,
      "grad_norm": 0.6202373881112844,
      "learning_rate": 6.364773440327241e-06,
      "loss": 0.3185,
      "num_input_tokens_seen": 208751472,
      "step": 3620,
      "train_runtime": 20573.0079,
      "train_tokens_per_second": 10146.862
    },
    {
      "epoch": 0.8253642987249544,
      "grad_norm": 0.7412512151950714,
      "learning_rate": 6.3561695533645505e-06,
      "loss": 0.3014,
      "num_input_tokens_seen": 209058112,
      "step": 3625,
      "train_runtime": 20606.7375,
      "train_tokens_per_second": 10145.134
    },
    {
      "epoch": 0.8265027322404371,
      "grad_norm": 0.6095612393082727,
      "learning_rate": 6.347561329606055e-06,
      "loss": 0.3104,
      "num_input_tokens_seen": 209319064,
      "step": 3630,
      "train_runtime": 20634.8439,
      "train_tokens_per_second": 10143.962
    },
    {
      "epoch": 0.8276411657559198,
      "grad_norm": 0.5842155964482182,
      "learning_rate": 6.338948796579367e-06,
      "loss": 0.2745,
      "num_input_tokens_seen": 209596736,
      "step": 3635,
      "train_runtime": 20659.4691,
      "train_tokens_per_second": 10145.311
    },
    {
      "epoch": 0.8287795992714025,
      "grad_norm": 0.557860572185387,
      "learning_rate": 6.330331981825879e-06,
      "loss": 0.2955,
      "num_input_tokens_seen": 209879928,
      "step": 3640,
      "train_runtime": 20687.3109,
      "train_tokens_per_second": 10145.346
    },
    {
      "epoch": 0.8299180327868853,
      "grad_norm": 0.6087154723017472,
      "learning_rate": 6.321710912900672e-06,
      "loss": 0.2859,
      "num_input_tokens_seen": 210171904,
      "step": 3645,
      "train_runtime": 20715.5566,
      "train_tokens_per_second": 10145.607
    },
    {
      "epoch": 0.831056466302368,
      "grad_norm": 0.9714824661017997,
      "learning_rate": 6.3130856173724366e-06,
      "loss": 0.2993,
      "num_input_tokens_seen": 210436400,
      "step": 3650,
      "train_runtime": 20743.7368,
      "train_tokens_per_second": 10144.575
    },
    {
      "epoch": 0.8321948998178507,
      "grad_norm": 0.6531795720312444,
      "learning_rate": 6.304456122823377e-06,
      "loss": 0.2724,
      "num_input_tokens_seen": 210711304,
      "step": 3655,
      "train_runtime": 20769.5888,
      "train_tokens_per_second": 10145.184
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.6077380661717592,
      "learning_rate": 6.2958224568491235e-06,
      "loss": 0.3393,
      "num_input_tokens_seen": 210999912,
      "step": 3660,
      "train_runtime": 20795.3375,
      "train_tokens_per_second": 10146.501
    },
    {
      "epoch": 0.8344717668488161,
      "grad_norm": 0.5167720289715317,
      "learning_rate": 6.287184647058648e-06,
      "loss": 0.2844,
      "num_input_tokens_seen": 211302432,
      "step": 3665,
      "train_runtime": 20818.6741,
      "train_tokens_per_second": 10149.658
    },
    {
      "epoch": 0.8356102003642987,
      "grad_norm": 0.6517712308628251,
      "learning_rate": 6.2785427210741725e-06,
      "loss": 0.3419,
      "num_input_tokens_seen": 211610304,
      "step": 3670,
      "train_runtime": 20845.3536,
      "train_tokens_per_second": 10151.437
    },
    {
      "epoch": 0.8367486338797814,
      "grad_norm": 0.6671693839719672,
      "learning_rate": 6.269896706531084e-06,
      "loss": 0.292,
      "num_input_tokens_seen": 211902920,
      "step": 3675,
      "train_runtime": 20870.256,
      "train_tokens_per_second": 10153.346
    },
    {
      "epoch": 0.8378870673952641,
      "grad_norm": 0.5096001579696128,
      "learning_rate": 6.261246631077844e-06,
      "loss": 0.3191,
      "num_input_tokens_seen": 212205512,
      "step": 3680,
      "train_runtime": 20906.5691,
      "train_tokens_per_second": 10150.183
    },
    {
      "epoch": 0.8390255009107468,
      "grad_norm": 0.5740901010634543,
      "learning_rate": 6.2525925223758944e-06,
      "loss": 0.3012,
      "num_input_tokens_seen": 212496120,
      "step": 3685,
      "train_runtime": 20931.9309,
      "train_tokens_per_second": 10151.769
    },
    {
      "epoch": 0.8401639344262295,
      "grad_norm": 0.6185473930680014,
      "learning_rate": 6.243934408099581e-06,
      "loss": 0.3342,
      "num_input_tokens_seen": 212797720,
      "step": 3690,
      "train_runtime": 20963.8494,
      "train_tokens_per_second": 10150.699
    },
    {
      "epoch": 0.8413023679417122,
      "grad_norm": 0.4930120040623356,
      "learning_rate": 6.235272315936058e-06,
      "loss": 0.293,
      "num_input_tokens_seen": 213085432,
      "step": 3695,
      "train_runtime": 20996.4029,
      "train_tokens_per_second": 10148.664
    },
    {
      "epoch": 0.8424408014571949,
      "grad_norm": 0.5735025190990949,
      "learning_rate": 6.226606273585195e-06,
      "loss": 0.3073,
      "num_input_tokens_seen": 213371176,
      "step": 3700,
      "train_runtime": 21022.6543,
      "train_tokens_per_second": 10149.583
    },
    {
      "epoch": 0.8435792349726776,
      "grad_norm": 0.6301099055266363,
      "learning_rate": 6.217936308759502e-06,
      "loss": 0.3096,
      "num_input_tokens_seen": 213649560,
      "step": 3705,
      "train_runtime": 21055.049,
      "train_tokens_per_second": 10147.189
    },
    {
      "epoch": 0.8447176684881603,
      "grad_norm": 0.6319391179503848,
      "learning_rate": 6.209262449184023e-06,
      "loss": 0.2714,
      "num_input_tokens_seen": 213959520,
      "step": 3710,
      "train_runtime": 21085.4985,
      "train_tokens_per_second": 10147.236
    },
    {
      "epoch": 0.845856102003643,
      "grad_norm": 0.6226703564288502,
      "learning_rate": 6.200584722596265e-06,
      "loss": 0.2738,
      "num_input_tokens_seen": 214242224,
      "step": 3715,
      "train_runtime": 21114.2408,
      "train_tokens_per_second": 10146.812
    },
    {
      "epoch": 0.8469945355191257,
      "grad_norm": 0.6295232515181237,
      "learning_rate": 6.191903156746097e-06,
      "loss": 0.2989,
      "num_input_tokens_seen": 214553464,
      "step": 3720,
      "train_runtime": 21148.0741,
      "train_tokens_per_second": 10145.296
    },
    {
      "epoch": 0.8481329690346083,
      "grad_norm": 0.5999295975149794,
      "learning_rate": 6.1832177793956675e-06,
      "loss": 0.2925,
      "num_input_tokens_seen": 214830520,
      "step": 3725,
      "train_runtime": 21175.9472,
      "train_tokens_per_second": 10145.025
    },
    {
      "epoch": 0.849271402550091,
      "grad_norm": 0.7084304098088177,
      "learning_rate": 6.174528618319308e-06,
      "loss": 0.3314,
      "num_input_tokens_seen": 215094128,
      "step": 3730,
      "train_runtime": 21204.5433,
      "train_tokens_per_second": 10143.776
    },
    {
      "epoch": 0.8504098360655737,
      "grad_norm": 0.5589012180496722,
      "learning_rate": 6.165835701303454e-06,
      "loss": 0.31,
      "num_input_tokens_seen": 215395784,
      "step": 3735,
      "train_runtime": 21235.7082,
      "train_tokens_per_second": 10143.094
    },
    {
      "epoch": 0.8515482695810564,
      "grad_norm": 0.6071823451198255,
      "learning_rate": 6.157139056146556e-06,
      "loss": 0.3303,
      "num_input_tokens_seen": 215678160,
      "step": 3740,
      "train_runtime": 21268.8938,
      "train_tokens_per_second": 10140.544
    },
    {
      "epoch": 0.8526867030965392,
      "grad_norm": 0.6018561376712896,
      "learning_rate": 6.148438710658979e-06,
      "loss": 0.3113,
      "num_input_tokens_seen": 215982128,
      "step": 3745,
      "train_runtime": 21297.195,
      "train_tokens_per_second": 10141.342
    },
    {
      "epoch": 0.8538251366120219,
      "grad_norm": 0.5574187178685455,
      "learning_rate": 6.1397346926629235e-06,
      "loss": 0.2716,
      "num_input_tokens_seen": 216291704,
      "step": 3750,
      "train_runtime": 21330.7471,
      "train_tokens_per_second": 10139.903
    },
    {
      "epoch": 0.8549635701275046,
      "grad_norm": 0.6498628681907623,
      "learning_rate": 6.131027029992334e-06,
      "loss": 0.3123,
      "num_input_tokens_seen": 216600736,
      "step": 3755,
      "train_runtime": 21357.8135,
      "train_tokens_per_second": 10141.522
    },
    {
      "epoch": 0.8561020036429873,
      "grad_norm": 0.6642144406583773,
      "learning_rate": 6.122315750492811e-06,
      "loss": 0.3077,
      "num_input_tokens_seen": 216901096,
      "step": 3760,
      "train_runtime": 21390.4354,
      "train_tokens_per_second": 10140.097
    },
    {
      "epoch": 0.85724043715847,
      "grad_norm": 0.6208156507953592,
      "learning_rate": 6.113600882021522e-06,
      "loss": 0.2984,
      "num_input_tokens_seen": 217192536,
      "step": 3765,
      "train_runtime": 21413.1267,
      "train_tokens_per_second": 10142.962
    },
    {
      "epoch": 0.8583788706739527,
      "grad_norm": 0.6347435793965125,
      "learning_rate": 6.10488245244711e-06,
      "loss": 0.3031,
      "num_input_tokens_seen": 217490328,
      "step": 3770,
      "train_runtime": 21434.6705,
      "train_tokens_per_second": 10146.661
    },
    {
      "epoch": 0.8595173041894353,
      "grad_norm": 0.5124905547867328,
      "learning_rate": 6.096160489649602e-06,
      "loss": 0.3089,
      "num_input_tokens_seen": 217773032,
      "step": 3775,
      "train_runtime": 21466.8247,
      "train_tokens_per_second": 10144.632
    },
    {
      "epoch": 0.860655737704918,
      "grad_norm": 0.6585484225428198,
      "learning_rate": 6.087435021520332e-06,
      "loss": 0.3162,
      "num_input_tokens_seen": 218058952,
      "step": 3780,
      "train_runtime": 21492.6438,
      "train_tokens_per_second": 10145.748
    },
    {
      "epoch": 0.8617941712204007,
      "grad_norm": 1.1583327569624233,
      "learning_rate": 6.078706075961836e-06,
      "loss": 0.3049,
      "num_input_tokens_seen": 218344504,
      "step": 3785,
      "train_runtime": 21516.3883,
      "train_tokens_per_second": 10147.823
    },
    {
      "epoch": 0.8629326047358834,
      "grad_norm": 0.6840484103933203,
      "learning_rate": 6.069973680887774e-06,
      "loss": 0.3198,
      "num_input_tokens_seen": 218626464,
      "step": 3790,
      "train_runtime": 21546.7047,
      "train_tokens_per_second": 10146.631
    },
    {
      "epoch": 0.8640710382513661,
      "grad_norm": 0.6053725034812181,
      "learning_rate": 6.0612378642228355e-06,
      "loss": 0.3004,
      "num_input_tokens_seen": 218908224,
      "step": 3795,
      "train_runtime": 21573.8055,
      "train_tokens_per_second": 10146.945
    },
    {
      "epoch": 0.8652094717668488,
      "grad_norm": 0.6616063378943885,
      "learning_rate": 6.052498653902652e-06,
      "loss": 0.2977,
      "num_input_tokens_seen": 219201080,
      "step": 3800,
      "train_runtime": 21601.7861,
      "train_tokens_per_second": 10147.359
    },
    {
      "epoch": 0.8663479052823315,
      "grad_norm": 0.5563246846575644,
      "learning_rate": 6.043756077873709e-06,
      "loss": 0.3083,
      "num_input_tokens_seen": 219487376,
      "step": 3805,
      "train_runtime": 21631.2003,
      "train_tokens_per_second": 10146.796
    },
    {
      "epoch": 0.8674863387978142,
      "grad_norm": 0.7354874509287656,
      "learning_rate": 6.035010164093251e-06,
      "loss": 0.297,
      "num_input_tokens_seen": 219760336,
      "step": 3810,
      "train_runtime": 21658.8458,
      "train_tokens_per_second": 10146.447
    },
    {
      "epoch": 0.8686247723132969,
      "grad_norm": 0.6169479664359581,
      "learning_rate": 6.0262609405292e-06,
      "loss": 0.3016,
      "num_input_tokens_seen": 220056368,
      "step": 3815,
      "train_runtime": 21691.5163,
      "train_tokens_per_second": 10144.813
    },
    {
      "epoch": 0.8697632058287796,
      "grad_norm": 0.5811133492333321,
      "learning_rate": 6.01750843516006e-06,
      "loss": 0.2973,
      "num_input_tokens_seen": 220336560,
      "step": 3820,
      "train_runtime": 21721.3016,
      "train_tokens_per_second": 10143.801
    },
    {
      "epoch": 0.8709016393442623,
      "grad_norm": 0.7442252733571959,
      "learning_rate": 6.00875267597483e-06,
      "loss": 0.3363,
      "num_input_tokens_seen": 220619672,
      "step": 3825,
      "train_runtime": 21749.0393,
      "train_tokens_per_second": 10143.881
    },
    {
      "epoch": 0.8720400728597449,
      "grad_norm": 0.732957945376876,
      "learning_rate": 5.999993690972916e-06,
      "loss": 0.3405,
      "num_input_tokens_seen": 220878024,
      "step": 3830,
      "train_runtime": 21777.5134,
      "train_tokens_per_second": 10142.481
    },
    {
      "epoch": 0.8731785063752276,
      "grad_norm": 0.6494304428124353,
      "learning_rate": 5.991231508164037e-06,
      "loss": 0.3076,
      "num_input_tokens_seen": 221191240,
      "step": 3835,
      "train_runtime": 21811.0069,
      "train_tokens_per_second": 10141.267
    },
    {
      "epoch": 0.8743169398907104,
      "grad_norm": 0.6213470112173113,
      "learning_rate": 5.982466155568137e-06,
      "loss": 0.2921,
      "num_input_tokens_seen": 221486600,
      "step": 3840,
      "train_runtime": 21847.1267,
      "train_tokens_per_second": 10138.02
    },
    {
      "epoch": 0.8754553734061931,
      "grad_norm": 0.5893778247660696,
      "learning_rate": 5.973697661215302e-06,
      "loss": 0.3094,
      "num_input_tokens_seen": 221778560,
      "step": 3845,
      "train_runtime": 21875.4442,
      "train_tokens_per_second": 10138.243
    },
    {
      "epoch": 0.8765938069216758,
      "grad_norm": 0.7313914934674289,
      "learning_rate": 5.96492605314566e-06,
      "loss": 0.2955,
      "num_input_tokens_seen": 222062808,
      "step": 3850,
      "train_runtime": 21909.13,
      "train_tokens_per_second": 10135.629
    },
    {
      "epoch": 0.8777322404371585,
      "grad_norm": 0.6891814139990615,
      "learning_rate": 5.956151359409295e-06,
      "loss": 0.3352,
      "num_input_tokens_seen": 222358208,
      "step": 3855,
      "train_runtime": 21941.2595,
      "train_tokens_per_second": 10134.25
    },
    {
      "epoch": 0.8788706739526412,
      "grad_norm": 0.6028976658482357,
      "learning_rate": 5.947373608066163e-06,
      "loss": 0.3223,
      "num_input_tokens_seen": 222652440,
      "step": 3860,
      "train_runtime": 21969.266,
      "train_tokens_per_second": 10134.724
    },
    {
      "epoch": 0.8800091074681239,
      "grad_norm": 0.7694900151597006,
      "learning_rate": 5.938592827185994e-06,
      "loss": 0.3146,
      "num_input_tokens_seen": 222910768,
      "step": 3865,
      "train_runtime": 21997.3146,
      "train_tokens_per_second": 10133.545
    },
    {
      "epoch": 0.8811475409836066,
      "grad_norm": 0.632078365843284,
      "learning_rate": 5.92980904484821e-06,
      "loss": 0.2924,
      "num_input_tokens_seen": 223182144,
      "step": 3870,
      "train_runtime": 22023.3799,
      "train_tokens_per_second": 10133.873
    },
    {
      "epoch": 0.8822859744990893,
      "grad_norm": 0.5547328055685299,
      "learning_rate": 5.921022289141827e-06,
      "loss": 0.2942,
      "num_input_tokens_seen": 223459784,
      "step": 3875,
      "train_runtime": 22054.9322,
      "train_tokens_per_second": 10131.964
    },
    {
      "epoch": 0.8834244080145719,
      "grad_norm": 0.6237155076559947,
      "learning_rate": 5.91223258816537e-06,
      "loss": 0.3144,
      "num_input_tokens_seen": 223759472,
      "step": 3880,
      "train_runtime": 22079.7444,
      "train_tokens_per_second": 10134.151
    },
    {
      "epoch": 0.8845628415300546,
      "grad_norm": 0.5599186910308167,
      "learning_rate": 5.903439970026783e-06,
      "loss": 0.2919,
      "num_input_tokens_seen": 224041000,
      "step": 3885,
      "train_runtime": 22106.9841,
      "train_tokens_per_second": 10134.399
    },
    {
      "epoch": 0.8857012750455373,
      "grad_norm": 0.6892471762693694,
      "learning_rate": 5.8946444628433416e-06,
      "loss": 0.3415,
      "num_input_tokens_seen": 224307352,
      "step": 3890,
      "train_runtime": 22129.2028,
      "train_tokens_per_second": 10136.26
    },
    {
      "epoch": 0.88683970856102,
      "grad_norm": 0.5690627563457296,
      "learning_rate": 5.885846094741556e-06,
      "loss": 0.3023,
      "num_input_tokens_seen": 224612520,
      "step": 3895,
      "train_runtime": 22161.3114,
      "train_tokens_per_second": 10135.344
    },
    {
      "epoch": 0.8879781420765027,
      "grad_norm": 0.7058109538602005,
      "learning_rate": 5.877044893857087e-06,
      "loss": 0.347,
      "num_input_tokens_seen": 224877696,
      "step": 3900,
      "train_runtime": 22189.412,
      "train_tokens_per_second": 10134.459
    },
    {
      "epoch": 0.8891165755919854,
      "grad_norm": 0.590642643325689,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.3262,
      "num_input_tokens_seen": 225146208,
      "step": 3905,
      "train_runtime": 22219.3641,
      "train_tokens_per_second": 10132.883
    },
    {
      "epoch": 0.8902550091074681,
      "grad_norm": 0.6726448210246072,
      "learning_rate": 5.8594341063279415e-06,
      "loss": 0.3392,
      "num_input_tokens_seen": 225417816,
      "step": 3910,
      "train_runtime": 22247.349,
      "train_tokens_per_second": 10132.345
    },
    {
      "epoch": 0.8913934426229508,
      "grad_norm": 0.6554648454329909,
      "learning_rate": 5.850624575999519e-06,
      "loss": 0.3339,
      "num_input_tokens_seen": 225708200,
      "step": 3915,
      "train_runtime": 22275.9344,
      "train_tokens_per_second": 10132.379
    },
    {
      "epoch": 0.8925318761384335,
      "grad_norm": 0.5856785559030583,
      "learning_rate": 5.841812325520742e-06,
      "loss": 0.2929,
      "num_input_tokens_seen": 226000416,
      "step": 3920,
      "train_runtime": 22305.4705,
      "train_tokens_per_second": 10132.062
    },
    {
      "epoch": 0.8936703096539163,
      "grad_norm": 0.5722633528041303,
      "learning_rate": 5.83299738307166e-06,
      "loss": 0.3168,
      "num_input_tokens_seen": 226295904,
      "step": 3925,
      "train_runtime": 22334.7274,
      "train_tokens_per_second": 10132.02
    },
    {
      "epoch": 0.894808743169399,
      "grad_norm": 0.5838155144375481,
      "learning_rate": 5.8241797768409415e-06,
      "loss": 0.343,
      "num_input_tokens_seen": 226605240,
      "step": 3930,
      "train_runtime": 22365.3916,
      "train_tokens_per_second": 10131.959
    },
    {
      "epoch": 0.8959471766848816,
      "grad_norm": 0.6241348608497758,
      "learning_rate": 5.815359535025764e-06,
      "loss": 0.2966,
      "num_input_tokens_seen": 226878784,
      "step": 3935,
      "train_runtime": 22392.1865,
      "train_tokens_per_second": 10132.051
    },
    {
      "epoch": 0.8970856102003643,
      "grad_norm": 0.4903857799922837,
      "learning_rate": 5.806536685831737e-06,
      "loss": 0.2702,
      "num_input_tokens_seen": 227179672,
      "step": 3940,
      "train_runtime": 22420.0027,
      "train_tokens_per_second": 10132.901
    },
    {
      "epoch": 0.898224043715847,
      "grad_norm": 0.6042096430386047,
      "learning_rate": 5.7977112574728045e-06,
      "loss": 0.2934,
      "num_input_tokens_seen": 227461016,
      "step": 3945,
      "train_runtime": 22445.2622,
      "train_tokens_per_second": 10134.032
    },
    {
      "epoch": 0.8993624772313297,
      "grad_norm": 0.6736228784269789,
      "learning_rate": 5.788883278171163e-06,
      "loss": 0.2844,
      "num_input_tokens_seen": 227750608,
      "step": 3950,
      "train_runtime": 22476.8641,
      "train_tokens_per_second": 10132.668
    },
    {
      "epoch": 0.9005009107468124,
      "grad_norm": 0.5435168196108667,
      "learning_rate": 5.780052776157164e-06,
      "loss": 0.2694,
      "num_input_tokens_seen": 228058776,
      "step": 3955,
      "train_runtime": 22506.867,
      "train_tokens_per_second": 10132.853
    },
    {
      "epoch": 0.9016393442622951,
      "grad_norm": 0.5755412730815386,
      "learning_rate": 5.771219779669227e-06,
      "loss": 0.2964,
      "num_input_tokens_seen": 228355016,
      "step": 3960,
      "train_runtime": 22532.9693,
      "train_tokens_per_second": 10134.262
    },
    {
      "epoch": 0.9027777777777778,
      "grad_norm": 0.6165353097621734,
      "learning_rate": 5.762384316953746e-06,
      "loss": 0.3104,
      "num_input_tokens_seen": 228631008,
      "step": 3965,
      "train_runtime": 22560.3876,
      "train_tokens_per_second": 10134.179
    },
    {
      "epoch": 0.9039162112932605,
      "grad_norm": 0.5414157935717591,
      "learning_rate": 5.753546416265006e-06,
      "loss": 0.3011,
      "num_input_tokens_seen": 228937712,
      "step": 3970,
      "train_runtime": 22585.3935,
      "train_tokens_per_second": 10136.539
    },
    {
      "epoch": 0.9050546448087432,
      "grad_norm": 0.6502303450033341,
      "learning_rate": 5.744706105865085e-06,
      "loss": 0.3099,
      "num_input_tokens_seen": 229224584,
      "step": 3975,
      "train_runtime": 22613.9562,
      "train_tokens_per_second": 10136.421
    },
    {
      "epoch": 0.9061930783242259,
      "grad_norm": 0.6078984330014551,
      "learning_rate": 5.735863414023765e-06,
      "loss": 0.3121,
      "num_input_tokens_seen": 229502848,
      "step": 3980,
      "train_runtime": 22642.6323,
      "train_tokens_per_second": 10135.873
    },
    {
      "epoch": 0.9073315118397086,
      "grad_norm": 0.7170846029043959,
      "learning_rate": 5.72701836901845e-06,
      "loss": 0.3212,
      "num_input_tokens_seen": 229782744,
      "step": 3985,
      "train_runtime": 22673.7117,
      "train_tokens_per_second": 10134.324
    },
    {
      "epoch": 0.9084699453551912,
      "grad_norm": 0.689391430587664,
      "learning_rate": 5.718170999134062e-06,
      "loss": 0.316,
      "num_input_tokens_seen": 230075656,
      "step": 3990,
      "train_runtime": 22703.02,
      "train_tokens_per_second": 10134.143
    },
    {
      "epoch": 0.9096083788706739,
      "grad_norm": 0.5233773736605434,
      "learning_rate": 5.709321332662961e-06,
      "loss": 0.2908,
      "num_input_tokens_seen": 230363648,
      "step": 3995,
      "train_runtime": 22723.9433,
      "train_tokens_per_second": 10137.486
    },
    {
      "epoch": 0.9107468123861566,
      "grad_norm": 0.6453932482617858,
      "learning_rate": 5.700469397904853e-06,
      "loss": 0.2942,
      "num_input_tokens_seen": 230631840,
      "step": 4000,
      "train_runtime": 22752.5946,
      "train_tokens_per_second": 10136.507
    },
    {
      "epoch": 0.9118852459016393,
      "grad_norm": 0.6361957103336958,
      "learning_rate": 5.691615223166692e-06,
      "loss": 0.2984,
      "num_input_tokens_seen": 230912176,
      "step": 4005,
      "train_runtime": 22782.7698,
      "train_tokens_per_second": 10135.386
    },
    {
      "epoch": 0.913023679417122,
      "grad_norm": 0.6901345004202163,
      "learning_rate": 5.682758836762601e-06,
      "loss": 0.3041,
      "num_input_tokens_seen": 231209920,
      "step": 4010,
      "train_runtime": 22806.7011,
      "train_tokens_per_second": 10137.806
    },
    {
      "epoch": 0.9141621129326047,
      "grad_norm": 0.6585587586370052,
      "learning_rate": 5.67390026701377e-06,
      "loss": 0.3337,
      "num_input_tokens_seen": 231460952,
      "step": 4015,
      "train_runtime": 22831.168,
      "train_tokens_per_second": 10137.937
    },
    {
      "epoch": 0.9153005464480874,
      "grad_norm": 0.5790706663803125,
      "learning_rate": 5.665039542248376e-06,
      "loss": 0.2804,
      "num_input_tokens_seen": 231752080,
      "step": 4020,
      "train_runtime": 22863.2851,
      "train_tokens_per_second": 10136.43
    },
    {
      "epoch": 0.9164389799635702,
      "grad_norm": 0.6095486281562563,
      "learning_rate": 5.656176690801484e-06,
      "loss": 0.3105,
      "num_input_tokens_seen": 232034856,
      "step": 4025,
      "train_runtime": 22892.4033,
      "train_tokens_per_second": 10135.889
    },
    {
      "epoch": 0.9175774134790529,
      "grad_norm": 0.7294379700499342,
      "learning_rate": 5.6473117410149606e-06,
      "loss": 0.3247,
      "num_input_tokens_seen": 232332552,
      "step": 4030,
      "train_runtime": 22920.297,
      "train_tokens_per_second": 10136.542
    },
    {
      "epoch": 0.9187158469945356,
      "grad_norm": 0.5740005690602671,
      "learning_rate": 5.6384447212373835e-06,
      "loss": 0.3011,
      "num_input_tokens_seen": 232644192,
      "step": 4035,
      "train_runtime": 22947.1638,
      "train_tokens_per_second": 10138.255
    },
    {
      "epoch": 0.9198542805100182,
      "grad_norm": 0.5936769905552767,
      "learning_rate": 5.629575659823949e-06,
      "loss": 0.2882,
      "num_input_tokens_seen": 232925728,
      "step": 4040,
      "train_runtime": 22974.1189,
      "train_tokens_per_second": 10138.614
    },
    {
      "epoch": 0.9209927140255009,
      "grad_norm": 0.5382563985556068,
      "learning_rate": 5.6207045851363836e-06,
      "loss": 0.3213,
      "num_input_tokens_seen": 233215224,
      "step": 4045,
      "train_runtime": 23004.5404,
      "train_tokens_per_second": 10137.791
    },
    {
      "epoch": 0.9221311475409836,
      "grad_norm": 0.6024884804416843,
      "learning_rate": 5.61183152554285e-06,
      "loss": 0.2867,
      "num_input_tokens_seen": 233492416,
      "step": 4050,
      "train_runtime": 23041.6092,
      "train_tokens_per_second": 10133.512
    },
    {
      "epoch": 0.9232695810564663,
      "grad_norm": 0.672743152662441,
      "learning_rate": 5.6029565094178575e-06,
      "loss": 0.3032,
      "num_input_tokens_seen": 233784424,
      "step": 4055,
      "train_runtime": 23068.6089,
      "train_tokens_per_second": 10134.31
    },
    {
      "epoch": 0.924408014571949,
      "grad_norm": 0.5848109205418683,
      "learning_rate": 5.594079565142175e-06,
      "loss": 0.3053,
      "num_input_tokens_seen": 234094560,
      "step": 4060,
      "train_runtime": 23097.3656,
      "train_tokens_per_second": 10135.119
    },
    {
      "epoch": 0.9255464480874317,
      "grad_norm": 0.5734104098004567,
      "learning_rate": 5.585200721102737e-06,
      "loss": 0.286,
      "num_input_tokens_seen": 234398576,
      "step": 4065,
      "train_runtime": 23128.2433,
      "train_tokens_per_second": 10134.733
    },
    {
      "epoch": 0.9266848816029144,
      "grad_norm": 0.6387749922813392,
      "learning_rate": 5.576320005692549e-06,
      "loss": 0.2911,
      "num_input_tokens_seen": 234677344,
      "step": 4070,
      "train_runtime": 23153.7651,
      "train_tokens_per_second": 10135.602
    },
    {
      "epoch": 0.9278233151183971,
      "grad_norm": 0.6510523837173143,
      "learning_rate": 5.5674374473106055e-06,
      "loss": 0.3093,
      "num_input_tokens_seen": 234963664,
      "step": 4075,
      "train_runtime": 23179.9424,
      "train_tokens_per_second": 10136.508
    },
    {
      "epoch": 0.9289617486338798,
      "grad_norm": 0.6041828682575267,
      "learning_rate": 5.558553074361792e-06,
      "loss": 0.2926,
      "num_input_tokens_seen": 235213864,
      "step": 4080,
      "train_runtime": 23207.1869,
      "train_tokens_per_second": 10135.389
    },
    {
      "epoch": 0.9301001821493625,
      "grad_norm": 0.6582341424544447,
      "learning_rate": 5.549666915256798e-06,
      "loss": 0.3165,
      "num_input_tokens_seen": 235492688,
      "step": 4085,
      "train_runtime": 23232.5431,
      "train_tokens_per_second": 10136.328
    },
    {
      "epoch": 0.9312386156648452,
      "grad_norm": 0.5378488752367951,
      "learning_rate": 5.540778998412025e-06,
      "loss": 0.2676,
      "num_input_tokens_seen": 235787816,
      "step": 4090,
      "train_runtime": 23254.7541,
      "train_tokens_per_second": 10139.338
    },
    {
      "epoch": 0.9323770491803278,
      "grad_norm": 0.5720512577238369,
      "learning_rate": 5.531889352249493e-06,
      "loss": 0.3045,
      "num_input_tokens_seen": 236068720,
      "step": 4095,
      "train_runtime": 23288.3307,
      "train_tokens_per_second": 10136.781
    },
    {
      "epoch": 0.9335154826958105,
      "grad_norm": 0.6129205421944063,
      "learning_rate": 5.522998005196756e-06,
      "loss": 0.3066,
      "num_input_tokens_seen": 236359784,
      "step": 4100,
      "train_runtime": 23316.5919,
      "train_tokens_per_second": 10136.978
    },
    {
      "epoch": 0.9346539162112932,
      "grad_norm": 0.5007282233362479,
      "learning_rate": 5.514104985686802e-06,
      "loss": 0.2676,
      "num_input_tokens_seen": 236641328,
      "step": 4105,
      "train_runtime": 23352.1008,
      "train_tokens_per_second": 10133.621
    },
    {
      "epoch": 0.9357923497267759,
      "grad_norm": 0.8204203038124033,
      "learning_rate": 5.5052103221579735e-06,
      "loss": 0.3375,
      "num_input_tokens_seen": 236937464,
      "step": 4110,
      "train_runtime": 23376.6191,
      "train_tokens_per_second": 10135.66
    },
    {
      "epoch": 0.9369307832422586,
      "grad_norm": 0.6403709211592323,
      "learning_rate": 5.496314043053864e-06,
      "loss": 0.3347,
      "num_input_tokens_seen": 237222128,
      "step": 4115,
      "train_runtime": 23410.5255,
      "train_tokens_per_second": 10133.14
    },
    {
      "epoch": 0.9380692167577414,
      "grad_norm": 0.5016978734073679,
      "learning_rate": 5.4874161768232385e-06,
      "loss": 0.2903,
      "num_input_tokens_seen": 237513728,
      "step": 4120,
      "train_runtime": 23432.9808,
      "train_tokens_per_second": 10135.873
    },
    {
      "epoch": 0.9392076502732241,
      "grad_norm": 0.6706903888902487,
      "learning_rate": 5.478516751919933e-06,
      "loss": 0.3252,
      "num_input_tokens_seen": 237809240,
      "step": 4125,
      "train_runtime": 23462.0373,
      "train_tokens_per_second": 10135.916
    },
    {
      "epoch": 0.9403460837887068,
      "grad_norm": 0.8875463951669881,
      "learning_rate": 5.4696157968027715e-06,
      "loss": 0.3046,
      "num_input_tokens_seen": 238103608,
      "step": 4130,
      "train_runtime": 23492.0475,
      "train_tokens_per_second": 10135.498
    },
    {
      "epoch": 0.9414845173041895,
      "grad_norm": 0.8283970133947485,
      "learning_rate": 5.460713339935469e-06,
      "loss": 0.3034,
      "num_input_tokens_seen": 238403736,
      "step": 4135,
      "train_runtime": 23522.5582,
      "train_tokens_per_second": 10135.111
    },
    {
      "epoch": 0.9426229508196722,
      "grad_norm": 0.6073565107493597,
      "learning_rate": 5.4518094097865436e-06,
      "loss": 0.3378,
      "num_input_tokens_seen": 238678504,
      "step": 4140,
      "train_runtime": 23550.9923,
      "train_tokens_per_second": 10134.541
    },
    {
      "epoch": 0.9437613843351548,
      "grad_norm": 0.5156914512983283,
      "learning_rate": 5.442904034829226e-06,
      "loss": 0.3106,
      "num_input_tokens_seen": 238978384,
      "step": 4145,
      "train_runtime": 23583.7597,
      "train_tokens_per_second": 10133.176
    },
    {
      "epoch": 0.9448998178506375,
      "grad_norm": 0.5870147590153068,
      "learning_rate": 5.433997243541366e-06,
      "loss": 0.2931,
      "num_input_tokens_seen": 239256912,
      "step": 4150,
      "train_runtime": 23608.2467,
      "train_tokens_per_second": 10134.463
    },
    {
      "epoch": 0.9460382513661202,
      "grad_norm": 0.5973207101319311,
      "learning_rate": 5.425089064405341e-06,
      "loss": 0.2775,
      "num_input_tokens_seen": 239550200,
      "step": 4155,
      "train_runtime": 23639.3854,
      "train_tokens_per_second": 10133.521
    },
    {
      "epoch": 0.9471766848816029,
      "grad_norm": 0.562960288174972,
      "learning_rate": 5.416179525907968e-06,
      "loss": 0.3293,
      "num_input_tokens_seen": 239847136,
      "step": 4160,
      "train_runtime": 23666.5953,
      "train_tokens_per_second": 10134.417
    },
    {
      "epoch": 0.9483151183970856,
      "grad_norm": 0.5922278965473649,
      "learning_rate": 5.407268656540412e-06,
      "loss": 0.2851,
      "num_input_tokens_seen": 240140504,
      "step": 4165,
      "train_runtime": 23697.8025,
      "train_tokens_per_second": 10133.45
    },
    {
      "epoch": 0.9494535519125683,
      "grad_norm": 0.5543411384046067,
      "learning_rate": 5.398356484798095e-06,
      "loss": 0.278,
      "num_input_tokens_seen": 240431184,
      "step": 4170,
      "train_runtime": 23726.4596,
      "train_tokens_per_second": 10133.462
    },
    {
      "epoch": 0.950591985428051,
      "grad_norm": 0.8436705288213979,
      "learning_rate": 5.389443039180597e-06,
      "loss": 0.2907,
      "num_input_tokens_seen": 240700192,
      "step": 4175,
      "train_runtime": 23754.3506,
      "train_tokens_per_second": 10132.889
    },
    {
      "epoch": 0.9517304189435337,
      "grad_norm": 0.7217947615554676,
      "learning_rate": 5.38052834819158e-06,
      "loss": 0.3035,
      "num_input_tokens_seen": 241000968,
      "step": 4180,
      "train_runtime": 23782.1366,
      "train_tokens_per_second": 10133.697
    },
    {
      "epoch": 0.9528688524590164,
      "grad_norm": 0.6067910792358878,
      "learning_rate": 5.371612440338682e-06,
      "loss": 0.3081,
      "num_input_tokens_seen": 241285560,
      "step": 4185,
      "train_runtime": 23812.3372,
      "train_tokens_per_second": 10132.796
    },
    {
      "epoch": 0.9540072859744991,
      "grad_norm": 0.5377891123981703,
      "learning_rate": 5.362695344133437e-06,
      "loss": 0.2975,
      "num_input_tokens_seen": 241592600,
      "step": 4190,
      "train_runtime": 23834.2127,
      "train_tokens_per_second": 10136.378
    },
    {
      "epoch": 0.9551457194899818,
      "grad_norm": 0.5965663308334085,
      "learning_rate": 5.353777088091177e-06,
      "loss": 0.3304,
      "num_input_tokens_seen": 241871224,
      "step": 4195,
      "train_runtime": 23865.5563,
      "train_tokens_per_second": 10134.741
    },
    {
      "epoch": 0.9562841530054644,
      "grad_norm": 0.6168870016045134,
      "learning_rate": 5.344857700730943e-06,
      "loss": 0.2991,
      "num_input_tokens_seen": 242165912,
      "step": 4200,
      "train_runtime": 23886.9025,
      "train_tokens_per_second": 10138.021
    },
    {
      "epoch": 0.9574225865209471,
      "grad_norm": 0.6529966592043543,
      "learning_rate": 5.335937210575394e-06,
      "loss": 0.2952,
      "num_input_tokens_seen": 242447776,
      "step": 4205,
      "train_runtime": 23918.7362,
      "train_tokens_per_second": 10136.312
    },
    {
      "epoch": 0.9585610200364298,
      "grad_norm": 0.5286083906071438,
      "learning_rate": 5.327015646150716e-06,
      "loss": 0.2815,
      "num_input_tokens_seen": 242745960,
      "step": 4210,
      "train_runtime": 23944.1197,
      "train_tokens_per_second": 10138.02
    },
    {
      "epoch": 0.9596994535519126,
      "grad_norm": 0.6288299053939014,
      "learning_rate": 5.318093035986529e-06,
      "loss": 0.2968,
      "num_input_tokens_seen": 243044384,
      "step": 4215,
      "train_runtime": 23972.8783,
      "train_tokens_per_second": 10138.306
    },
    {
      "epoch": 0.9608378870673953,
      "grad_norm": 0.7333094914148242,
      "learning_rate": 5.309169408615799e-06,
      "loss": 0.2879,
      "num_input_tokens_seen": 243320576,
      "step": 4220,
      "train_runtime": 24004.6491,
      "train_tokens_per_second": 10136.394
    },
    {
      "epoch": 0.961976320582878,
      "grad_norm": 0.661231701192262,
      "learning_rate": 5.300244792574743e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 243588960,
      "step": 4225,
      "train_runtime": 24026.716,
      "train_tokens_per_second": 10138.254
    },
    {
      "epoch": 0.9631147540983607,
      "grad_norm": 0.790872117133461,
      "learning_rate": 5.2913192164027395e-06,
      "loss": 0.3193,
      "num_input_tokens_seen": 243877144,
      "step": 4230,
      "train_runtime": 24057.7304,
      "train_tokens_per_second": 10137.163
    },
    {
      "epoch": 0.9642531876138434,
      "grad_norm": 0.6306935676526666,
      "learning_rate": 5.28239270864224e-06,
      "loss": 0.2838,
      "num_input_tokens_seen": 244167536,
      "step": 4235,
      "train_runtime": 24091.8593,
      "train_tokens_per_second": 10134.856
    },
    {
      "epoch": 0.9653916211293261,
      "grad_norm": 0.5982771636615039,
      "learning_rate": 5.273465297838672e-06,
      "loss": 0.2882,
      "num_input_tokens_seen": 244452296,
      "step": 4240,
      "train_runtime": 24115.2605,
      "train_tokens_per_second": 10136.83
    },
    {
      "epoch": 0.9665300546448088,
      "grad_norm": 0.5909295460241148,
      "learning_rate": 5.264537012540353e-06,
      "loss": 0.2992,
      "num_input_tokens_seen": 244755992,
      "step": 4245,
      "train_runtime": 24140.3517,
      "train_tokens_per_second": 10138.874
    },
    {
      "epoch": 0.9676684881602914,
      "grad_norm": 0.6664489652990311,
      "learning_rate": 5.2556078812983955e-06,
      "loss": 0.2992,
      "num_input_tokens_seen": 245027704,
      "step": 4250,
      "train_runtime": 24172.2768,
      "train_tokens_per_second": 10136.724
    },
    {
      "epoch": 0.9688069216757741,
      "grad_norm": 0.6436186328709055,
      "learning_rate": 5.246677932666618e-06,
      "loss": 0.3052,
      "num_input_tokens_seen": 245305824,
      "step": 4255,
      "train_runtime": 24199.5135,
      "train_tokens_per_second": 10136.808
    },
    {
      "epoch": 0.9699453551912568,
      "grad_norm": 0.595992053007179,
      "learning_rate": 5.237747195201453e-06,
      "loss": 0.2922,
      "num_input_tokens_seen": 245606296,
      "step": 4260,
      "train_runtime": 24229.8986,
      "train_tokens_per_second": 10136.497
    },
    {
      "epoch": 0.9710837887067395,
      "grad_norm": 0.6583973669447014,
      "learning_rate": 5.2288156974618535e-06,
      "loss": 0.2783,
      "num_input_tokens_seen": 245906384,
      "step": 4265,
      "train_runtime": 24261.78,
      "train_tokens_per_second": 10135.546
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 0.6286468392718337,
      "learning_rate": 5.219883468009207e-06,
      "loss": 0.305,
      "num_input_tokens_seen": 246203336,
      "step": 4270,
      "train_runtime": 24289.7134,
      "train_tokens_per_second": 10136.115
    },
    {
      "epoch": 0.9733606557377049,
      "grad_norm": 0.6302575313817507,
      "learning_rate": 5.210950535407241e-06,
      "loss": 0.2979,
      "num_input_tokens_seen": 246497176,
      "step": 4275,
      "train_runtime": 24317.1006,
      "train_tokens_per_second": 10136.783
    },
    {
      "epoch": 0.9744990892531876,
      "grad_norm": 0.755853576781382,
      "learning_rate": 5.202016928221926e-06,
      "loss": 0.3157,
      "num_input_tokens_seen": 246781656,
      "step": 4280,
      "train_runtime": 24342.7988,
      "train_tokens_per_second": 10137.768
    },
    {
      "epoch": 0.9756375227686703,
      "grad_norm": 0.5406508820570908,
      "learning_rate": 5.193082675021393e-06,
      "loss": 0.28,
      "num_input_tokens_seen": 247062952,
      "step": 4285,
      "train_runtime": 24365.8775,
      "train_tokens_per_second": 10139.711
    },
    {
      "epoch": 0.976775956284153,
      "grad_norm": 0.6192006283491058,
      "learning_rate": 5.184147804375843e-06,
      "loss": 0.3085,
      "num_input_tokens_seen": 247340320,
      "step": 4290,
      "train_runtime": 24395.7117,
      "train_tokens_per_second": 10138.68
    },
    {
      "epoch": 0.9779143897996357,
      "grad_norm": 0.5549776194653263,
      "learning_rate": 5.175212344857446e-06,
      "loss": 0.291,
      "num_input_tokens_seen": 247626000,
      "step": 4295,
      "train_runtime": 24424.1372,
      "train_tokens_per_second": 10138.577
    },
    {
      "epoch": 0.9790528233151184,
      "grad_norm": 0.6655992114461238,
      "learning_rate": 5.166276325040259e-06,
      "loss": 0.3019,
      "num_input_tokens_seen": 247922096,
      "step": 4300,
      "train_runtime": 24454.1016,
      "train_tokens_per_second": 10138.262
    },
    {
      "epoch": 0.980191256830601,
      "grad_norm": 0.655848575603903,
      "learning_rate": 5.1573397735001254e-06,
      "loss": 0.3106,
      "num_input_tokens_seen": 248201232,
      "step": 4305,
      "train_runtime": 24481.7739,
      "train_tokens_per_second": 10138.205
    },
    {
      "epoch": 0.9813296903460837,
      "grad_norm": 0.6365791334514828,
      "learning_rate": 5.148402718814594e-06,
      "loss": 0.2959,
      "num_input_tokens_seen": 248494776,
      "step": 4310,
      "train_runtime": 24514.1435,
      "train_tokens_per_second": 10136.792
    },
    {
      "epoch": 0.9824681238615665,
      "grad_norm": 0.6202643127651225,
      "learning_rate": 5.13946518956282e-06,
      "loss": 0.3203,
      "num_input_tokens_seen": 248792432,
      "step": 4315,
      "train_runtime": 24550.477,
      "train_tokens_per_second": 10133.914
    },
    {
      "epoch": 0.9836065573770492,
      "grad_norm": 0.4974361179820846,
      "learning_rate": 5.130527214325479e-06,
      "loss": 0.324,
      "num_input_tokens_seen": 249115440,
      "step": 4320,
      "train_runtime": 24573.8759,
      "train_tokens_per_second": 10137.409
    },
    {
      "epoch": 0.9847449908925319,
      "grad_norm": 0.531287490887208,
      "learning_rate": 5.121588821684669e-06,
      "loss": 0.3012,
      "num_input_tokens_seen": 249412856,
      "step": 4325,
      "train_runtime": 24600.5333,
      "train_tokens_per_second": 10138.514
    },
    {
      "epoch": 0.9858834244080146,
      "grad_norm": 0.5715030767670292,
      "learning_rate": 5.112650040223825e-06,
      "loss": 0.307,
      "num_input_tokens_seen": 249698456,
      "step": 4330,
      "train_runtime": 24631.951,
      "train_tokens_per_second": 10137.177
    },
    {
      "epoch": 0.9870218579234973,
      "grad_norm": 0.6138186295084531,
      "learning_rate": 5.103710898527625e-06,
      "loss": 0.3457,
      "num_input_tokens_seen": 250006320,
      "step": 4335,
      "train_runtime": 24659.403,
      "train_tokens_per_second": 10138.377
    },
    {
      "epoch": 0.98816029143898,
      "grad_norm": 0.6641737839046079,
      "learning_rate": 5.094771425181899e-06,
      "loss": 0.3148,
      "num_input_tokens_seen": 250274256,
      "step": 4340,
      "train_runtime": 24685.466,
      "train_tokens_per_second": 10138.527
    },
    {
      "epoch": 0.9892987249544627,
      "grad_norm": 0.6034603741558324,
      "learning_rate": 5.0858316487735385e-06,
      "loss": 0.3073,
      "num_input_tokens_seen": 250552272,
      "step": 4345,
      "train_runtime": 24714.7765,
      "train_tokens_per_second": 10137.752
    },
    {
      "epoch": 0.9904371584699454,
      "grad_norm": 0.6796088029250308,
      "learning_rate": 5.0768915978904e-06,
      "loss": 0.2967,
      "num_input_tokens_seen": 250830336,
      "step": 4350,
      "train_runtime": 24746.0616,
      "train_tokens_per_second": 10136.172
    },
    {
      "epoch": 0.9915755919854281,
      "grad_norm": 0.622064366074152,
      "learning_rate": 5.067951301121223e-06,
      "loss": 0.2876,
      "num_input_tokens_seen": 251124856,
      "step": 4355,
      "train_runtime": 24773.6472,
      "train_tokens_per_second": 10136.774
    },
    {
      "epoch": 0.9927140255009107,
      "grad_norm": 0.5848021308434899,
      "learning_rate": 5.059010787055533e-06,
      "loss": 0.3096,
      "num_input_tokens_seen": 251398600,
      "step": 4360,
      "train_runtime": 24810.5042,
      "train_tokens_per_second": 10132.749
    },
    {
      "epoch": 0.9938524590163934,
      "grad_norm": 0.6230437686280557,
      "learning_rate": 5.050070084283545e-06,
      "loss": 0.3013,
      "num_input_tokens_seen": 251694944,
      "step": 4365,
      "train_runtime": 24833.3736,
      "train_tokens_per_second": 10135.35
    },
    {
      "epoch": 0.9949908925318761,
      "grad_norm": 0.577102850933345,
      "learning_rate": 5.041129221396083e-06,
      "loss": 0.2615,
      "num_input_tokens_seen": 252003816,
      "step": 4370,
      "train_runtime": 24857.0805,
      "train_tokens_per_second": 10138.11
    },
    {
      "epoch": 0.9961293260473588,
      "grad_norm": 0.5510232493810872,
      "learning_rate": 5.03218822698448e-06,
      "loss": 0.3042,
      "num_input_tokens_seen": 252315008,
      "step": 4375,
      "train_runtime": 24879.7593,
      "train_tokens_per_second": 10141.377
    },
    {
      "epoch": 0.9972677595628415,
      "grad_norm": 0.8201626123249361,
      "learning_rate": 5.023247129640493e-06,
      "loss": 0.2986,
      "num_input_tokens_seen": 252596264,
      "step": 4380,
      "train_runtime": 24911.5711,
      "train_tokens_per_second": 10139.716
    },
    {
      "epoch": 0.9984061930783242,
      "grad_norm": 0.6268777908170947,
      "learning_rate": 5.014305957956203e-06,
      "loss": 0.2998,
      "num_input_tokens_seen": 252895840,
      "step": 4385,
      "train_runtime": 24935.3445,
      "train_tokens_per_second": 10142.063
    },
    {
      "epoch": 0.9995446265938069,
      "grad_norm": 0.5104399599239527,
      "learning_rate": 5.0053647405239305e-06,
      "loss": 0.303,
      "num_input_tokens_seen": 253203704,
      "step": 4390,
      "train_runtime": 24966.4849,
      "train_tokens_per_second": 10141.744
    },
    {
      "epoch": 1.0006830601092895,
      "grad_norm": 0.6523801989507008,
      "learning_rate": 4.996423505936145e-06,
      "loss": 0.2766,
      "num_input_tokens_seen": 253493392,
      "step": 4395,
      "train_runtime": 24996.3353,
      "train_tokens_per_second": 10141.222
    },
    {
      "epoch": 1.0018214936247722,
      "grad_norm": 0.7862826238404068,
      "learning_rate": 4.987482282785372e-06,
      "loss": 0.2833,
      "num_input_tokens_seen": 253779456,
      "step": 4400,
      "train_runtime": 25026.1197,
      "train_tokens_per_second": 10140.583
    },
    {
      "epoch": 1.002959927140255,
      "grad_norm": 0.6568237078309661,
      "learning_rate": 4.978541099664096e-06,
      "loss": 0.2771,
      "num_input_tokens_seen": 254062280,
      "step": 4405,
      "train_runtime": 25057.6814,
      "train_tokens_per_second": 10139.098
    },
    {
      "epoch": 1.0040983606557377,
      "grad_norm": 0.5606720933931256,
      "learning_rate": 4.96959998516467e-06,
      "loss": 0.2752,
      "num_input_tokens_seen": 254360680,
      "step": 4410,
      "train_runtime": 25081.421,
      "train_tokens_per_second": 10141.398
    },
    {
      "epoch": 1.0052367941712204,
      "grad_norm": 0.6865620841518127,
      "learning_rate": 4.9606589678792385e-06,
      "loss": 0.2584,
      "num_input_tokens_seen": 254632264,
      "step": 4415,
      "train_runtime": 25120.252,
      "train_tokens_per_second": 10136.533
    },
    {
      "epoch": 1.006375227686703,
      "grad_norm": 0.765152245094524,
      "learning_rate": 4.951718076399628e-06,
      "loss": 0.2568,
      "num_input_tokens_seen": 254913768,
      "step": 4420,
      "train_runtime": 25150.8102,
      "train_tokens_per_second": 10135.41
    },
    {
      "epoch": 1.0075136612021858,
      "grad_norm": 0.6172614914027398,
      "learning_rate": 4.942777339317262e-06,
      "loss": 0.2502,
      "num_input_tokens_seen": 255192296,
      "step": 4425,
      "train_runtime": 25174.3528,
      "train_tokens_per_second": 10136.995
    },
    {
      "epoch": 1.0086520947176685,
      "grad_norm": 0.5191277929847345,
      "learning_rate": 4.933836785223074e-06,
      "loss": 0.2581,
      "num_input_tokens_seen": 255460776,
      "step": 4430,
      "train_runtime": 25205.0211,
      "train_tokens_per_second": 10135.313
    },
    {
      "epoch": 1.0097905282331512,
      "grad_norm": 0.6636596892834323,
      "learning_rate": 4.924896442707408e-06,
      "loss": 0.2612,
      "num_input_tokens_seen": 255745008,
      "step": 4435,
      "train_runtime": 25232.9002,
      "train_tokens_per_second": 10135.379
    },
    {
      "epoch": 1.010928961748634,
      "grad_norm": 0.5635722786926061,
      "learning_rate": 4.915956340359937e-06,
      "loss": 0.2502,
      "num_input_tokens_seen": 256022944,
      "step": 4440,
      "train_runtime": 25257.0545,
      "train_tokens_per_second": 10136.69
    },
    {
      "epoch": 1.0120673952641166,
      "grad_norm": 0.5781615461872435,
      "learning_rate": 4.907016506769562e-06,
      "loss": 0.2632,
      "num_input_tokens_seen": 256328144,
      "step": 4445,
      "train_runtime": 25286.6569,
      "train_tokens_per_second": 10136.893
    },
    {
      "epoch": 1.0132058287795993,
      "grad_norm": 0.5643734646413108,
      "learning_rate": 4.898076970524325e-06,
      "loss": 0.2445,
      "num_input_tokens_seen": 256618984,
      "step": 4450,
      "train_runtime": 25315.6278,
      "train_tokens_per_second": 10136.781
    },
    {
      "epoch": 1.014344262295082,
      "grad_norm": 0.6023415098258968,
      "learning_rate": 4.889137760211318e-06,
      "loss": 0.2345,
      "num_input_tokens_seen": 256902720,
      "step": 4455,
      "train_runtime": 25341.943,
      "train_tokens_per_second": 10137.452
    },
    {
      "epoch": 1.0154826958105647,
      "grad_norm": 0.8737166062967157,
      "learning_rate": 4.88019890441659e-06,
      "loss": 0.2808,
      "num_input_tokens_seen": 257186976,
      "step": 4460,
      "train_runtime": 25368.6357,
      "train_tokens_per_second": 10137.99
    },
    {
      "epoch": 1.0166211293260474,
      "grad_norm": 0.7152958897506475,
      "learning_rate": 4.871260431725058e-06,
      "loss": 0.2494,
      "num_input_tokens_seen": 257460536,
      "step": 4465,
      "train_runtime": 25394.0863,
      "train_tokens_per_second": 10138.602
    },
    {
      "epoch": 1.0177595628415301,
      "grad_norm": 0.663277553792361,
      "learning_rate": 4.862322370720411e-06,
      "loss": 0.2373,
      "num_input_tokens_seen": 257751336,
      "step": 4470,
      "train_runtime": 25422.132,
      "train_tokens_per_second": 10138.856
    },
    {
      "epoch": 1.0188979963570128,
      "grad_norm": 0.6058829744363283,
      "learning_rate": 4.853384749985023e-06,
      "loss": 0.2557,
      "num_input_tokens_seen": 258033504,
      "step": 4475,
      "train_runtime": 25453.2526,
      "train_tokens_per_second": 10137.545
    },
    {
      "epoch": 1.0200364298724955,
      "grad_norm": 0.6982321170103606,
      "learning_rate": 4.84444759809986e-06,
      "loss": 0.2575,
      "num_input_tokens_seen": 258321664,
      "step": 4480,
      "train_runtime": 25480.2329,
      "train_tokens_per_second": 10138.12
    },
    {
      "epoch": 1.0211748633879782,
      "grad_norm": 0.5619035849833189,
      "learning_rate": 4.835510943644391e-06,
      "loss": 0.25,
      "num_input_tokens_seen": 258611128,
      "step": 4485,
      "train_runtime": 25505.7057,
      "train_tokens_per_second": 10139.344
    },
    {
      "epoch": 1.0223132969034607,
      "grad_norm": 0.6736968395572303,
      "learning_rate": 4.82657481519649e-06,
      "loss": 0.2459,
      "num_input_tokens_seen": 258902768,
      "step": 4490,
      "train_runtime": 25532.8732,
      "train_tokens_per_second": 10139.978
    },
    {
      "epoch": 1.0234517304189434,
      "grad_norm": 0.6900919722151104,
      "learning_rate": 4.817639241332353e-06,
      "loss": 0.2814,
      "num_input_tokens_seen": 259183392,
      "step": 4495,
      "train_runtime": 25562.68,
      "train_tokens_per_second": 10139.132
    },
    {
      "epoch": 1.0245901639344261,
      "grad_norm": 0.6587514913891024,
      "learning_rate": 4.808704250626398e-06,
      "loss": 0.2641,
      "num_input_tokens_seen": 259464536,
      "step": 4500,
      "train_runtime": 25593.6672,
      "train_tokens_per_second": 10137.841
    },
    {
      "epoch": 1.0257285974499089,
      "grad_norm": 0.5977126387924332,
      "learning_rate": 4.799769871651182e-06,
      "loss": 0.2434,
      "num_input_tokens_seen": 259734824,
      "step": 4505,
      "train_runtime": 25621.1364,
      "train_tokens_per_second": 10137.522
    },
    {
      "epoch": 1.0268670309653916,
      "grad_norm": 0.5838898215796945,
      "learning_rate": 4.790836132977308e-06,
      "loss": 0.2503,
      "num_input_tokens_seen": 260028960,
      "step": 4510,
      "train_runtime": 25651.639,
      "train_tokens_per_second": 10136.934
    },
    {
      "epoch": 1.0280054644808743,
      "grad_norm": 0.5328958568586313,
      "learning_rate": 4.781903063173321e-06,
      "loss": 0.2187,
      "num_input_tokens_seen": 260324632,
      "step": 4515,
      "train_runtime": 25672.7281,
      "train_tokens_per_second": 10140.123
    },
    {
      "epoch": 1.029143897996357,
      "grad_norm": 0.6095696531963996,
      "learning_rate": 4.7729706908056385e-06,
      "loss": 0.2736,
      "num_input_tokens_seen": 260595368,
      "step": 4520,
      "train_runtime": 25701.4069,
      "train_tokens_per_second": 10139.343
    },
    {
      "epoch": 1.0302823315118397,
      "grad_norm": 0.6509553515361335,
      "learning_rate": 4.764039044438444e-06,
      "loss": 0.2487,
      "num_input_tokens_seen": 260847952,
      "step": 4525,
      "train_runtime": 25725.3351,
      "train_tokens_per_second": 10139.73
    },
    {
      "epoch": 1.0314207650273224,
      "grad_norm": 0.5078382226276285,
      "learning_rate": 4.755108152633596e-06,
      "loss": 0.2589,
      "num_input_tokens_seen": 261122576,
      "step": 4530,
      "train_runtime": 25752.1915,
      "train_tokens_per_second": 10139.82
    },
    {
      "epoch": 1.032559198542805,
      "grad_norm": 0.6715210899617781,
      "learning_rate": 4.746178043950545e-06,
      "loss": 0.2449,
      "num_input_tokens_seen": 261406040,
      "step": 4535,
      "train_runtime": 25784.3522,
      "train_tokens_per_second": 10138.166
    },
    {
      "epoch": 1.0336976320582878,
      "grad_norm": 0.8163587052484466,
      "learning_rate": 4.737248746946232e-06,
      "loss": 0.277,
      "num_input_tokens_seen": 261661816,
      "step": 4540,
      "train_runtime": 25809.2199,
      "train_tokens_per_second": 10138.308
    },
    {
      "epoch": 1.0348360655737705,
      "grad_norm": 0.5141134813871132,
      "learning_rate": 4.7283202901750095e-06,
      "loss": 0.2466,
      "num_input_tokens_seen": 261948160,
      "step": 4545,
      "train_runtime": 25831.9059,
      "train_tokens_per_second": 10140.489
    },
    {
      "epoch": 1.0359744990892532,
      "grad_norm": 0.6387868629915058,
      "learning_rate": 4.719392702188535e-06,
      "loss": 0.2641,
      "num_input_tokens_seen": 262242408,
      "step": 4550,
      "train_runtime": 25864.6977,
      "train_tokens_per_second": 10139.009
    },
    {
      "epoch": 1.037112932604736,
      "grad_norm": 0.6934908881593881,
      "learning_rate": 4.710466011535695e-06,
      "loss": 0.2429,
      "num_input_tokens_seen": 262507096,
      "step": 4555,
      "train_runtime": 25897.6447,
      "train_tokens_per_second": 10136.331
    },
    {
      "epoch": 1.0382513661202186,
      "grad_norm": 0.6123615152240919,
      "learning_rate": 4.701540246762499e-06,
      "loss": 0.2403,
      "num_input_tokens_seen": 262769984,
      "step": 4560,
      "train_runtime": 25924.3687,
      "train_tokens_per_second": 10136.022
    },
    {
      "epoch": 1.0393897996357013,
      "grad_norm": 0.652643093325979,
      "learning_rate": 4.692615436412003e-06,
      "loss": 0.2466,
      "num_input_tokens_seen": 263066768,
      "step": 4565,
      "train_runtime": 25953.467,
      "train_tokens_per_second": 10136.094
    },
    {
      "epoch": 1.040528233151184,
      "grad_norm": 0.6263033322977528,
      "learning_rate": 4.6836916090242075e-06,
      "loss": 0.2536,
      "num_input_tokens_seen": 263351112,
      "step": 4570,
      "train_runtime": 25981.5641,
      "train_tokens_per_second": 10136.076
    },
    {
      "epoch": 1.0416666666666667,
      "grad_norm": 0.6006695177588677,
      "learning_rate": 4.6747687931359694e-06,
      "loss": 0.2532,
      "num_input_tokens_seen": 263652512,
      "step": 4575,
      "train_runtime": 26010.621,
      "train_tokens_per_second": 10136.341
    },
    {
      "epoch": 1.0428051001821494,
      "grad_norm": 0.625808667107273,
      "learning_rate": 4.665847017280911e-06,
      "loss": 0.2543,
      "num_input_tokens_seen": 263946760,
      "step": 4580,
      "train_runtime": 26036.9493,
      "train_tokens_per_second": 10137.392
    },
    {
      "epoch": 1.0439435336976322,
      "grad_norm": 0.6868025460627646,
      "learning_rate": 4.65692630998933e-06,
      "loss": 0.2574,
      "num_input_tokens_seen": 264263688,
      "step": 4585,
      "train_runtime": 26062.1942,
      "train_tokens_per_second": 10139.733
    },
    {
      "epoch": 1.0450819672131149,
      "grad_norm": 0.5836928840687117,
      "learning_rate": 4.648006699788107e-06,
      "loss": 0.2529,
      "num_input_tokens_seen": 264577552,
      "step": 4590,
      "train_runtime": 26093.6167,
      "train_tokens_per_second": 10139.551
    },
    {
      "epoch": 1.0462204007285973,
      "grad_norm": 0.6528612151261581,
      "learning_rate": 4.639088215200611e-06,
      "loss": 0.262,
      "num_input_tokens_seen": 264873112,
      "step": 4595,
      "train_runtime": 26121.2771,
      "train_tokens_per_second": 10140.129
    },
    {
      "epoch": 1.04735883424408,
      "grad_norm": 0.7344157857669642,
      "learning_rate": 4.630170884746617e-06,
      "loss": 0.2529,
      "num_input_tokens_seen": 265154840,
      "step": 4600,
      "train_runtime": 26145.4997,
      "train_tokens_per_second": 10141.51
    },
    {
      "epoch": 1.0484972677595628,
      "grad_norm": 0.696601305699015,
      "learning_rate": 4.621254736942204e-06,
      "loss": 0.2426,
      "num_input_tokens_seen": 265417792,
      "step": 4605,
      "train_runtime": 26170.8722,
      "train_tokens_per_second": 10141.725
    },
    {
      "epoch": 1.0496357012750455,
      "grad_norm": 0.6559395561926379,
      "learning_rate": 4.612339800299674e-06,
      "loss": 0.2689,
      "num_input_tokens_seen": 265702184,
      "step": 4610,
      "train_runtime": 26203.1112,
      "train_tokens_per_second": 10140.101
    },
    {
      "epoch": 1.0507741347905282,
      "grad_norm": 0.6163428425846708,
      "learning_rate": 4.603426103327452e-06,
      "loss": 0.2719,
      "num_input_tokens_seen": 265994376,
      "step": 4615,
      "train_runtime": 26233.6559,
      "train_tokens_per_second": 10139.432
    },
    {
      "epoch": 1.0519125683060109,
      "grad_norm": 0.7148282206605012,
      "learning_rate": 4.59451367453e-06,
      "loss": 0.2656,
      "num_input_tokens_seen": 266272648,
      "step": 4620,
      "train_runtime": 26262.351,
      "train_tokens_per_second": 10138.949
    },
    {
      "epoch": 1.0530510018214936,
      "grad_norm": 0.7195553655863822,
      "learning_rate": 4.585602542407723e-06,
      "loss": 0.2739,
      "num_input_tokens_seen": 266544832,
      "step": 4625,
      "train_runtime": 26287.3035,
      "train_tokens_per_second": 10139.679
    },
    {
      "epoch": 1.0541894353369763,
      "grad_norm": 0.7051001561821287,
      "learning_rate": 4.576692735456883e-06,
      "loss": 0.256,
      "num_input_tokens_seen": 266841720,
      "step": 4630,
      "train_runtime": 26313.2097,
      "train_tokens_per_second": 10140.98
    },
    {
      "epoch": 1.055327868852459,
      "grad_norm": 0.6028442124417148,
      "learning_rate": 4.567784282169503e-06,
      "loss": 0.2593,
      "num_input_tokens_seen": 267117248,
      "step": 4635,
      "train_runtime": 26340.4619,
      "train_tokens_per_second": 10140.948
    },
    {
      "epoch": 1.0564663023679417,
      "grad_norm": 0.5602066545212969,
      "learning_rate": 4.558877211033276e-06,
      "loss": 0.252,
      "num_input_tokens_seen": 267400168,
      "step": 4640,
      "train_runtime": 26373.0972,
      "train_tokens_per_second": 10139.126
    },
    {
      "epoch": 1.0576047358834244,
      "grad_norm": 0.5938153376621137,
      "learning_rate": 4.549971550531474e-06,
      "loss": 0.2786,
      "num_input_tokens_seen": 267665400,
      "step": 4645,
      "train_runtime": 26396.8188,
      "train_tokens_per_second": 10140.063
    },
    {
      "epoch": 1.0587431693989071,
      "grad_norm": 0.5985287197986752,
      "learning_rate": 4.541067329142864e-06,
      "loss": 0.2453,
      "num_input_tokens_seen": 267956696,
      "step": 4650,
      "train_runtime": 26425.387,
      "train_tokens_per_second": 10140.124
    },
    {
      "epoch": 1.0598816029143898,
      "grad_norm": 0.5070948034931475,
      "learning_rate": 4.532164575341604e-06,
      "loss": 0.2284,
      "num_input_tokens_seen": 268269664,
      "step": 4655,
      "train_runtime": 26447.7617,
      "train_tokens_per_second": 10143.379
    },
    {
      "epoch": 1.0610200364298725,
      "grad_norm": 0.6397253430779991,
      "learning_rate": 4.523263317597162e-06,
      "loss": 0.2451,
      "num_input_tokens_seen": 268564264,
      "step": 4660,
      "train_runtime": 26477.2575,
      "train_tokens_per_second": 10143.206
    },
    {
      "epoch": 1.0621584699453552,
      "grad_norm": 0.7048735622381982,
      "learning_rate": 4.514363584374222e-06,
      "loss": 0.2784,
      "num_input_tokens_seen": 268855680,
      "step": 4665,
      "train_runtime": 26507.6749,
      "train_tokens_per_second": 10142.56
    },
    {
      "epoch": 1.063296903460838,
      "grad_norm": 0.5878763642565563,
      "learning_rate": 4.505465404132591e-06,
      "loss": 0.2886,
      "num_input_tokens_seen": 269160120,
      "step": 4670,
      "train_runtime": 26534.0663,
      "train_tokens_per_second": 10143.945
    },
    {
      "epoch": 1.0644353369763206,
      "grad_norm": 0.7575643380382592,
      "learning_rate": 4.496568805327114e-06,
      "loss": 0.2603,
      "num_input_tokens_seen": 269431680,
      "step": 4675,
      "train_runtime": 26564.9664,
      "train_tokens_per_second": 10142.369
    },
    {
      "epoch": 1.0655737704918034,
      "grad_norm": 0.6224375537930177,
      "learning_rate": 4.487673816407575e-06,
      "loss": 0.2801,
      "num_input_tokens_seen": 269687832,
      "step": 4680,
      "train_runtime": 26596.858,
      "train_tokens_per_second": 10139.838
    },
    {
      "epoch": 1.066712204007286,
      "grad_norm": 0.5898767209705471,
      "learning_rate": 4.47878046581861e-06,
      "loss": 0.2395,
      "num_input_tokens_seen": 270003288,
      "step": 4685,
      "train_runtime": 26629.8491,
      "train_tokens_per_second": 10139.122
    },
    {
      "epoch": 1.0678506375227688,
      "grad_norm": 0.6086383611079564,
      "learning_rate": 4.469888781999617e-06,
      "loss": 0.2648,
      "num_input_tokens_seen": 270308680,
      "step": 4690,
      "train_runtime": 26653.5179,
      "train_tokens_per_second": 10141.576
    },
    {
      "epoch": 1.0689890710382515,
      "grad_norm": 0.6407127204460754,
      "learning_rate": 4.460998793384663e-06,
      "loss": 0.2916,
      "num_input_tokens_seen": 270583640,
      "step": 4695,
      "train_runtime": 26687.0022,
      "train_tokens_per_second": 10139.155
    },
    {
      "epoch": 1.0701275045537342,
      "grad_norm": 0.6179644771644861,
      "learning_rate": 4.4521105284023984e-06,
      "loss": 0.2609,
      "num_input_tokens_seen": 270875360,
      "step": 4700,
      "train_runtime": 26718.0462,
      "train_tokens_per_second": 10138.292
    },
    {
      "epoch": 1.0712659380692167,
      "grad_norm": 0.5769285652726741,
      "learning_rate": 4.443224015475956e-06,
      "loss": 0.2481,
      "num_input_tokens_seen": 271165760,
      "step": 4705,
      "train_runtime": 26747.5582,
      "train_tokens_per_second": 10137.963
    },
    {
      "epoch": 1.0724043715846994,
      "grad_norm": 0.5433791861509789,
      "learning_rate": 4.4343392830228685e-06,
      "loss": 0.2498,
      "num_input_tokens_seen": 271468576,
      "step": 4710,
      "train_runtime": 26777.4457,
      "train_tokens_per_second": 10137.956
    },
    {
      "epoch": 1.073542805100182,
      "grad_norm": 0.5724350718637539,
      "learning_rate": 4.425456359454972e-06,
      "loss": 0.2514,
      "num_input_tokens_seen": 271765136,
      "step": 4715,
      "train_runtime": 26800.8842,
      "train_tokens_per_second": 10140.156
    },
    {
      "epoch": 1.0746812386156648,
      "grad_norm": 0.6615429122527031,
      "learning_rate": 4.416575273178323e-06,
      "loss": 0.2544,
      "num_input_tokens_seen": 272046872,
      "step": 4720,
      "train_runtime": 26828.9811,
      "train_tokens_per_second": 10140.037
    },
    {
      "epoch": 1.0758196721311475,
      "grad_norm": 0.641565005886214,
      "learning_rate": 4.407696052593103e-06,
      "loss": 0.2534,
      "num_input_tokens_seen": 272326056,
      "step": 4725,
      "train_runtime": 26854.5337,
      "train_tokens_per_second": 10140.785
    },
    {
      "epoch": 1.0769581056466302,
      "grad_norm": 1.344672616896549,
      "learning_rate": 4.398818726093518e-06,
      "loss": 0.2836,
      "num_input_tokens_seen": 272604856,
      "step": 4730,
      "train_runtime": 26887.2759,
      "train_tokens_per_second": 10138.805
    },
    {
      "epoch": 1.078096539162113,
      "grad_norm": 0.6051226259012819,
      "learning_rate": 4.389943322067728e-06,
      "loss": 0.2497,
      "num_input_tokens_seen": 272890424,
      "step": 4735,
      "train_runtime": 26914.7653,
      "train_tokens_per_second": 10139.06
    },
    {
      "epoch": 1.0792349726775956,
      "grad_norm": 0.6437893013894405,
      "learning_rate": 4.38106986889774e-06,
      "loss": 0.2528,
      "num_input_tokens_seen": 273193408,
      "step": 4740,
      "train_runtime": 26937.6159,
      "train_tokens_per_second": 10141.707
    },
    {
      "epoch": 1.0803734061930783,
      "grad_norm": 0.6461659808289557,
      "learning_rate": 4.372198394959324e-06,
      "loss": 0.2616,
      "num_input_tokens_seen": 273499248,
      "step": 4745,
      "train_runtime": 26959.6844,
      "train_tokens_per_second": 10144.75
    },
    {
      "epoch": 1.081511839708561,
      "grad_norm": 0.5510262352378575,
      "learning_rate": 4.363328928621918e-06,
      "loss": 0.2656,
      "num_input_tokens_seen": 273804608,
      "step": 4750,
      "train_runtime": 26981.4166,
      "train_tokens_per_second": 10147.896
    },
    {
      "epoch": 1.0826502732240437,
      "grad_norm": 0.7084700875237415,
      "learning_rate": 4.3544614982485435e-06,
      "loss": 0.2549,
      "num_input_tokens_seen": 274081736,
      "step": 4755,
      "train_runtime": 27006.5909,
      "train_tokens_per_second": 10148.698
    },
    {
      "epoch": 1.0837887067395264,
      "grad_norm": 0.550499163546633,
      "learning_rate": 4.345596132195711e-06,
      "loss": 0.2562,
      "num_input_tokens_seen": 274381864,
      "step": 4760,
      "train_runtime": 27037.7657,
      "train_tokens_per_second": 10148.097
    },
    {
      "epoch": 1.0849271402550091,
      "grad_norm": 0.6426479437598281,
      "learning_rate": 4.336732858813328e-06,
      "loss": 0.2766,
      "num_input_tokens_seen": 274656832,
      "step": 4765,
      "train_runtime": 27066.3355,
      "train_tokens_per_second": 10147.544
    },
    {
      "epoch": 1.0860655737704918,
      "grad_norm": 0.5858536238831275,
      "learning_rate": 4.3278717064446095e-06,
      "loss": 0.2619,
      "num_input_tokens_seen": 274922928,
      "step": 4770,
      "train_runtime": 27092.9995,
      "train_tokens_per_second": 10147.379
    },
    {
      "epoch": 1.0872040072859745,
      "grad_norm": 0.7354526340696356,
      "learning_rate": 4.3190127034259895e-06,
      "loss": 0.2619,
      "num_input_tokens_seen": 275212080,
      "step": 4775,
      "train_runtime": 27117.9969,
      "train_tokens_per_second": 10148.688
    },
    {
      "epoch": 1.0883424408014573,
      "grad_norm": 0.6138121263568656,
      "learning_rate": 4.310155878087029e-06,
      "loss": 0.2565,
      "num_input_tokens_seen": 275491832,
      "step": 4780,
      "train_runtime": 27146.8658,
      "train_tokens_per_second": 10148.2
    },
    {
      "epoch": 1.08948087431694,
      "grad_norm": 0.5482601446693817,
      "learning_rate": 4.301301258750325e-06,
      "loss": 0.2525,
      "num_input_tokens_seen": 275787760,
      "step": 4785,
      "train_runtime": 27176.988,
      "train_tokens_per_second": 10147.841
    },
    {
      "epoch": 1.0906193078324227,
      "grad_norm": 0.6288651256167124,
      "learning_rate": 4.292448873731419e-06,
      "loss": 0.2628,
      "num_input_tokens_seen": 276087904,
      "step": 4790,
      "train_runtime": 27205.774,
      "train_tokens_per_second": 10148.136
    },
    {
      "epoch": 1.0917577413479054,
      "grad_norm": 0.5615444789209119,
      "learning_rate": 4.2835987513387075e-06,
      "loss": 0.2389,
      "num_input_tokens_seen": 276396680,
      "step": 4795,
      "train_runtime": 27231.9887,
      "train_tokens_per_second": 10149.706
    },
    {
      "epoch": 1.092896174863388,
      "grad_norm": 0.6487555371629741,
      "learning_rate": 4.274750919873352e-06,
      "loss": 0.2585,
      "num_input_tokens_seen": 276680480,
      "step": 4800,
      "train_runtime": 27265.915,
      "train_tokens_per_second": 10147.486
    },
    {
      "epoch": 1.0940346083788706,
      "grad_norm": 0.5611518325660149,
      "learning_rate": 4.265905407629189e-06,
      "loss": 0.2532,
      "num_input_tokens_seen": 276967544,
      "step": 4805,
      "train_runtime": 27292.1744,
      "train_tokens_per_second": 10148.24
    },
    {
      "epoch": 1.0951730418943533,
      "grad_norm": 0.5831320822324298,
      "learning_rate": 4.257062242892636e-06,
      "loss": 0.2592,
      "num_input_tokens_seen": 277259920,
      "step": 4810,
      "train_runtime": 27323.3578,
      "train_tokens_per_second": 10147.359
    },
    {
      "epoch": 1.096311475409836,
      "grad_norm": 0.6329040589204561,
      "learning_rate": 4.248221453942608e-06,
      "loss": 0.2407,
      "num_input_tokens_seen": 277528792,
      "step": 4815,
      "train_runtime": 27348.246,
      "train_tokens_per_second": 10147.956
    },
    {
      "epoch": 1.0974499089253187,
      "grad_norm": 0.7266711526886316,
      "learning_rate": 4.239383069050417e-06,
      "loss": 0.2631,
      "num_input_tokens_seen": 277801088,
      "step": 4820,
      "train_runtime": 27373.0195,
      "train_tokens_per_second": 10148.719
    },
    {
      "epoch": 1.0985883424408014,
      "grad_norm": 0.7160784034508294,
      "learning_rate": 4.230547116479691e-06,
      "loss": 0.2651,
      "num_input_tokens_seen": 278101024,
      "step": 4825,
      "train_runtime": 27400.0975,
      "train_tokens_per_second": 10149.636
    },
    {
      "epoch": 1.099726775956284,
      "grad_norm": 0.536631804921634,
      "learning_rate": 4.221713624486282e-06,
      "loss": 0.2392,
      "num_input_tokens_seen": 278384856,
      "step": 4830,
      "train_runtime": 27429.274,
      "train_tokens_per_second": 10149.188
    },
    {
      "epoch": 1.1008652094717668,
      "grad_norm": 0.6276228220602398,
      "learning_rate": 4.212882621318164e-06,
      "loss": 0.2465,
      "num_input_tokens_seen": 278671944,
      "step": 4835,
      "train_runtime": 27459.7474,
      "train_tokens_per_second": 10148.38
    },
    {
      "epoch": 1.1020036429872495,
      "grad_norm": 0.5882928252390793,
      "learning_rate": 4.204054135215364e-06,
      "loss": 0.2569,
      "num_input_tokens_seen": 278944352,
      "step": 4840,
      "train_runtime": 27486.1256,
      "train_tokens_per_second": 10148.551
    },
    {
      "epoch": 1.1031420765027322,
      "grad_norm": 0.8082805255131839,
      "learning_rate": 4.195228194409852e-06,
      "loss": 0.2642,
      "num_input_tokens_seen": 279217840,
      "step": 4845,
      "train_runtime": 27516.9567,
      "train_tokens_per_second": 10147.119
    },
    {
      "epoch": 1.104280510018215,
      "grad_norm": 0.646809219909014,
      "learning_rate": 4.186404827125463e-06,
      "loss": 0.247,
      "num_input_tokens_seen": 279491184,
      "step": 4850,
      "train_runtime": 27548.9086,
      "train_tokens_per_second": 10145.272
    },
    {
      "epoch": 1.1054189435336976,
      "grad_norm": 0.5908698746461865,
      "learning_rate": 4.177584061577798e-06,
      "loss": 0.2449,
      "num_input_tokens_seen": 279762592,
      "step": 4855,
      "train_runtime": 27581.5292,
      "train_tokens_per_second": 10143.114
    },
    {
      "epoch": 1.1065573770491803,
      "grad_norm": 0.5831316169976283,
      "learning_rate": 4.168765925974141e-06,
      "loss": 0.2331,
      "num_input_tokens_seen": 280058400,
      "step": 4860,
      "train_runtime": 27610.5362,
      "train_tokens_per_second": 10143.171
    },
    {
      "epoch": 1.107695810564663,
      "grad_norm": 0.6962478079063866,
      "learning_rate": 4.159950448513367e-06,
      "loss": 0.2624,
      "num_input_tokens_seen": 280336088,
      "step": 4865,
      "train_runtime": 27643.2895,
      "train_tokens_per_second": 10141.199
    },
    {
      "epoch": 1.1088342440801457,
      "grad_norm": 0.6061737832006396,
      "learning_rate": 4.151137657385848e-06,
      "loss": 0.2308,
      "num_input_tokens_seen": 280618512,
      "step": 4870,
      "train_runtime": 27669.8974,
      "train_tokens_per_second": 10141.654
    },
    {
      "epoch": 1.1099726775956285,
      "grad_norm": 0.5843561309943963,
      "learning_rate": 4.142327580773366e-06,
      "loss": 0.2565,
      "num_input_tokens_seen": 280902168,
      "step": 4875,
      "train_runtime": 27695.0213,
      "train_tokens_per_second": 10142.696
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.6044100866039253,
      "learning_rate": 4.133520246849025e-06,
      "loss": 0.2515,
      "num_input_tokens_seen": 281196784,
      "step": 4880,
      "train_runtime": 27720.743,
      "train_tokens_per_second": 10143.912
    },
    {
      "epoch": 1.1122495446265939,
      "grad_norm": 0.7194662643233706,
      "learning_rate": 4.124715683777153e-06,
      "loss": 0.2699,
      "num_input_tokens_seen": 281489224,
      "step": 4885,
      "train_runtime": 27751.7189,
      "train_tokens_per_second": 10143.128
    },
    {
      "epoch": 1.1133879781420766,
      "grad_norm": 0.5987595016125647,
      "learning_rate": 4.115913919713225e-06,
      "loss": 0.2434,
      "num_input_tokens_seen": 281775144,
      "step": 4890,
      "train_runtime": 27780.6249,
      "train_tokens_per_second": 10142.866
    },
    {
      "epoch": 1.1145264116575593,
      "grad_norm": 0.6759014939642373,
      "learning_rate": 4.107114982803758e-06,
      "loss": 0.2497,
      "num_input_tokens_seen": 282067720,
      "step": 4895,
      "train_runtime": 27803.3701,
      "train_tokens_per_second": 10145.091
    },
    {
      "epoch": 1.115664845173042,
      "grad_norm": 0.6161627756781307,
      "learning_rate": 4.098318901186234e-06,
      "loss": 0.2547,
      "num_input_tokens_seen": 282344520,
      "step": 4900,
      "train_runtime": 27826.8778,
      "train_tokens_per_second": 10146.468
    },
    {
      "epoch": 1.1168032786885247,
      "grad_norm": 0.6137385417119158,
      "learning_rate": 4.089525702988996e-06,
      "loss": 0.2701,
      "num_input_tokens_seen": 282644024,
      "step": 4905,
      "train_runtime": 27852.7271,
      "train_tokens_per_second": 10147.804
    },
    {
      "epoch": 1.1179417122040074,
      "grad_norm": 0.6399849481776242,
      "learning_rate": 4.080735416331179e-06,
      "loss": 0.2788,
      "num_input_tokens_seen": 282915920,
      "step": 4910,
      "train_runtime": 27881.7054,
      "train_tokens_per_second": 10147.009
    },
    {
      "epoch": 1.1190801457194899,
      "grad_norm": 0.6220220218410837,
      "learning_rate": 4.0719480693225964e-06,
      "loss": 0.2369,
      "num_input_tokens_seen": 283208608,
      "step": 4915,
      "train_runtime": 27911.2656,
      "train_tokens_per_second": 10146.749
    },
    {
      "epoch": 1.1202185792349726,
      "grad_norm": 0.6528651300370565,
      "learning_rate": 4.063163690063665e-06,
      "loss": 0.2527,
      "num_input_tokens_seen": 283490880,
      "step": 4920,
      "train_runtime": 27945.5429,
      "train_tokens_per_second": 10144.404
    },
    {
      "epoch": 1.1213570127504553,
      "grad_norm": 0.5869998796748384,
      "learning_rate": 4.054382306645312e-06,
      "loss": 0.2596,
      "num_input_tokens_seen": 283782256,
      "step": 4925,
      "train_runtime": 27974.868,
      "train_tokens_per_second": 10144.186
    },
    {
      "epoch": 1.122495446265938,
      "grad_norm": 0.6829144495607188,
      "learning_rate": 4.045603947148882e-06,
      "loss": 0.2617,
      "num_input_tokens_seen": 284051512,
      "step": 4930,
      "train_runtime": 28005.7942,
      "train_tokens_per_second": 10142.598
    },
    {
      "epoch": 1.1236338797814207,
      "grad_norm": 0.6419294453101737,
      "learning_rate": 4.036828639646056e-06,
      "loss": 0.2559,
      "num_input_tokens_seen": 284341256,
      "step": 4935,
      "train_runtime": 28032.9186,
      "train_tokens_per_second": 10143.12
    },
    {
      "epoch": 1.1247723132969034,
      "grad_norm": 0.5876319566725319,
      "learning_rate": 4.028056412198744e-06,
      "loss": 0.2395,
      "num_input_tokens_seen": 284638384,
      "step": 4940,
      "train_runtime": 28064.2898,
      "train_tokens_per_second": 10142.369
    },
    {
      "epoch": 1.1259107468123861,
      "grad_norm": 0.746745278766013,
      "learning_rate": 4.019287292859016e-06,
      "loss": 0.2564,
      "num_input_tokens_seen": 284937312,
      "step": 4945,
      "train_runtime": 28094.1093,
      "train_tokens_per_second": 10142.244
    },
    {
      "epoch": 1.1270491803278688,
      "grad_norm": 0.6670569081959491,
      "learning_rate": 4.010521309669e-06,
      "loss": 0.2324,
      "num_input_tokens_seen": 285227864,
      "step": 4950,
      "train_runtime": 28125.4408,
      "train_tokens_per_second": 10141.276
    },
    {
      "epoch": 1.1281876138433515,
      "grad_norm": 0.6256747347473257,
      "learning_rate": 4.001758490660796e-06,
      "loss": 0.253,
      "num_input_tokens_seen": 285524376,
      "step": 4955,
      "train_runtime": 28150.4477,
      "train_tokens_per_second": 10142.801
    },
    {
      "epoch": 1.1293260473588342,
      "grad_norm": 0.5895864175736671,
      "learning_rate": 3.992998863856382e-06,
      "loss": 0.2403,
      "num_input_tokens_seen": 285829328,
      "step": 4960,
      "train_runtime": 28177.6228,
      "train_tokens_per_second": 10143.841
    },
    {
      "epoch": 1.130464480874317,
      "grad_norm": 0.6472153809846105,
      "learning_rate": 3.984242457267531e-06,
      "loss": 0.2545,
      "num_input_tokens_seen": 286132776,
      "step": 4965,
      "train_runtime": 28205.8786,
      "train_tokens_per_second": 10144.438
    },
    {
      "epoch": 1.1316029143897997,
      "grad_norm": 0.6104572235322193,
      "learning_rate": 3.97548929889572e-06,
      "loss": 0.2799,
      "num_input_tokens_seen": 286426584,
      "step": 4970,
      "train_runtime": 28239.7089,
      "train_tokens_per_second": 10142.689
    },
    {
      "epoch": 1.1327413479052824,
      "grad_norm": 0.6992868065716459,
      "learning_rate": 3.966739416732034e-06,
      "loss": 0.2401,
      "num_input_tokens_seen": 286708712,
      "step": 4975,
      "train_runtime": 28267.3121,
      "train_tokens_per_second": 10142.765
    },
    {
      "epoch": 1.133879781420765,
      "grad_norm": 0.6755655642789162,
      "learning_rate": 3.9579928387570845e-06,
      "loss": 0.2382,
      "num_input_tokens_seen": 286997144,
      "step": 4980,
      "train_runtime": 28296.3166,
      "train_tokens_per_second": 10142.562
    },
    {
      "epoch": 1.1350182149362478,
      "grad_norm": 0.5564331228467091,
      "learning_rate": 3.949249592940915e-06,
      "loss": 0.226,
      "num_input_tokens_seen": 287284464,
      "step": 4985,
      "train_runtime": 28323.8195,
      "train_tokens_per_second": 10142.857
    },
    {
      "epoch": 1.1361566484517305,
      "grad_norm": 0.8635864196075829,
      "learning_rate": 3.940509707242914e-06,
      "loss": 0.2752,
      "num_input_tokens_seen": 287559392,
      "step": 4990,
      "train_runtime": 28351.2903,
      "train_tokens_per_second": 10142.727
    },
    {
      "epoch": 1.1372950819672132,
      "grad_norm": 0.5562961876084285,
      "learning_rate": 3.931773209611729e-06,
      "loss": 0.2592,
      "num_input_tokens_seen": 287840376,
      "step": 4995,
      "train_runtime": 28376.8317,
      "train_tokens_per_second": 10143.499
    },
    {
      "epoch": 1.138433515482696,
      "grad_norm": 0.7155541905534916,
      "learning_rate": 3.9230401279851645e-06,
      "loss": 0.2771,
      "num_input_tokens_seen": 288119824,
      "step": 5000,
      "train_runtime": 28408.6334,
      "train_tokens_per_second": 10141.981
    },
    {
      "epoch": 1.1395719489981786,
      "grad_norm": 0.635771721611938,
      "learning_rate": 3.9143104902901085e-06,
      "loss": 0.2694,
      "num_input_tokens_seen": 288415664,
      "step": 5005,
      "train_runtime": 28434.0503,
      "train_tokens_per_second": 10143.32
    },
    {
      "epoch": 1.140710382513661,
      "grad_norm": 0.6146029748685439,
      "learning_rate": 3.905584324442432e-06,
      "loss": 0.2602,
      "num_input_tokens_seen": 288710048,
      "step": 5010,
      "train_runtime": 28459.124,
      "train_tokens_per_second": 10144.727
    },
    {
      "epoch": 1.1418488160291438,
      "grad_norm": 0.6154714416324568,
      "learning_rate": 3.896861658346905e-06,
      "loss": 0.2754,
      "num_input_tokens_seen": 289013288,
      "step": 5015,
      "train_runtime": 28487.1943,
      "train_tokens_per_second": 10145.376
    },
    {
      "epoch": 1.1429872495446265,
      "grad_norm": 0.7285030102884341,
      "learning_rate": 3.888142519897109e-06,
      "loss": 0.2444,
      "num_input_tokens_seen": 289299624,
      "step": 5020,
      "train_runtime": 28515.2139,
      "train_tokens_per_second": 10145.448
    },
    {
      "epoch": 1.1441256830601092,
      "grad_norm": 0.6368888432740089,
      "learning_rate": 3.879426936975338e-06,
      "loss": 0.2706,
      "num_input_tokens_seen": 289603624,
      "step": 5025,
      "train_runtime": 28544.7196,
      "train_tokens_per_second": 10145.611
    },
    {
      "epoch": 1.145264116575592,
      "grad_norm": 0.6324443868356356,
      "learning_rate": 3.87071493745252e-06,
      "loss": 0.2527,
      "num_input_tokens_seen": 289889624,
      "step": 5030,
      "train_runtime": 28574.6125,
      "train_tokens_per_second": 10145.006
    },
    {
      "epoch": 1.1464025500910746,
      "grad_norm": 0.6615688017558103,
      "learning_rate": 3.862006549188125e-06,
      "loss": 0.2687,
      "num_input_tokens_seen": 290166448,
      "step": 5035,
      "train_runtime": 28604.559,
      "train_tokens_per_second": 10144.063
    },
    {
      "epoch": 1.1475409836065573,
      "grad_norm": 0.696920680432317,
      "learning_rate": 3.8533018000300745e-06,
      "loss": 0.2462,
      "num_input_tokens_seen": 290450528,
      "step": 5040,
      "train_runtime": 28630.6544,
      "train_tokens_per_second": 10144.739
    },
    {
      "epoch": 1.14867941712204,
      "grad_norm": 0.7353431420165627,
      "learning_rate": 3.844600717814648e-06,
      "loss": 0.2656,
      "num_input_tokens_seen": 290743280,
      "step": 5045,
      "train_runtime": 28659.2836,
      "train_tokens_per_second": 10144.82
    },
    {
      "epoch": 1.1498178506375227,
      "grad_norm": 0.7575085150913131,
      "learning_rate": 3.835903330366403e-06,
      "loss": 0.2628,
      "num_input_tokens_seen": 291018224,
      "step": 5050,
      "train_runtime": 28683.9036,
      "train_tokens_per_second": 10145.698
    },
    {
      "epoch": 1.1509562841530054,
      "grad_norm": 0.7186963257415414,
      "learning_rate": 3.8272096654980835e-06,
      "loss": 0.2507,
      "num_input_tokens_seen": 291307352,
      "step": 5055,
      "train_runtime": 28711.8576,
      "train_tokens_per_second": 10145.89
    },
    {
      "epoch": 1.1520947176684881,
      "grad_norm": 0.7226762880541142,
      "learning_rate": 3.818519751010525e-06,
      "loss": 0.259,
      "num_input_tokens_seen": 291581112,
      "step": 5060,
      "train_runtime": 28737.7793,
      "train_tokens_per_second": 10146.265
    },
    {
      "epoch": 1.1532331511839709,
      "grad_norm": 0.6101773294329443,
      "learning_rate": 3.8098336146925728e-06,
      "loss": 0.2725,
      "num_input_tokens_seen": 291866360,
      "step": 5065,
      "train_runtime": 28762.3411,
      "train_tokens_per_second": 10147.518
    },
    {
      "epoch": 1.1543715846994536,
      "grad_norm": 0.6031776308013538,
      "learning_rate": 3.8011512843209895e-06,
      "loss": 0.2373,
      "num_input_tokens_seen": 292163736,
      "step": 5070,
      "train_runtime": 28790.5792,
      "train_tokens_per_second": 10147.894
    },
    {
      "epoch": 1.1555100182149363,
      "grad_norm": 0.5926759620668206,
      "learning_rate": 3.7924727876603657e-06,
      "loss": 0.2498,
      "num_input_tokens_seen": 292462368,
      "step": 5075,
      "train_runtime": 28822.2367,
      "train_tokens_per_second": 10147.109
    },
    {
      "epoch": 1.156648451730419,
      "grad_norm": 0.599687265649078,
      "learning_rate": 3.7837981524630345e-06,
      "loss": 0.2724,
      "num_input_tokens_seen": 292761616,
      "step": 5080,
      "train_runtime": 28851.5476,
      "train_tokens_per_second": 10147.172
    },
    {
      "epoch": 1.1577868852459017,
      "grad_norm": 0.6078064609539696,
      "learning_rate": 3.7751274064689803e-06,
      "loss": 0.255,
      "num_input_tokens_seen": 293072008,
      "step": 5085,
      "train_runtime": 28876.6974,
      "train_tokens_per_second": 10149.083
    },
    {
      "epoch": 1.1589253187613844,
      "grad_norm": 0.7264576135499071,
      "learning_rate": 3.766460577405749e-06,
      "loss": 0.2645,
      "num_input_tokens_seen": 293362056,
      "step": 5090,
      "train_runtime": 28908.8213,
      "train_tokens_per_second": 10147.839
    },
    {
      "epoch": 1.160063752276867,
      "grad_norm": 0.6435070461682606,
      "learning_rate": 3.7577976929883608e-06,
      "loss": 0.2691,
      "num_input_tokens_seen": 293628040,
      "step": 5095,
      "train_runtime": 28938.3098,
      "train_tokens_per_second": 10146.689
    },
    {
      "epoch": 1.1612021857923498,
      "grad_norm": 0.6878630130244197,
      "learning_rate": 3.7491387809192246e-06,
      "loss": 0.265,
      "num_input_tokens_seen": 293878440,
      "step": 5100,
      "train_runtime": 28969.532,
      "train_tokens_per_second": 10144.397
    },
    {
      "epoch": 1.1623406193078325,
      "grad_norm": 0.7043608156843258,
      "learning_rate": 3.7404838688880445e-06,
      "loss": 0.2624,
      "num_input_tokens_seen": 294148576,
      "step": 5105,
      "train_runtime": 28993.6915,
      "train_tokens_per_second": 10145.261
    },
    {
      "epoch": 1.1634790528233152,
      "grad_norm": 0.6763161646245802,
      "learning_rate": 3.7318329845717318e-06,
      "loss": 0.2661,
      "num_input_tokens_seen": 294452432,
      "step": 5110,
      "train_runtime": 29022.1131,
      "train_tokens_per_second": 10145.796
    },
    {
      "epoch": 1.164617486338798,
      "grad_norm": 0.543048178847618,
      "learning_rate": 3.7231861556343197e-06,
      "loss": 0.2636,
      "num_input_tokens_seen": 294739672,
      "step": 5115,
      "train_runtime": 29056.5206,
      "train_tokens_per_second": 10143.667
    },
    {
      "epoch": 1.1657559198542806,
      "grad_norm": 0.6312614261630505,
      "learning_rate": 3.714543409726871e-06,
      "loss": 0.2472,
      "num_input_tokens_seen": 295026680,
      "step": 5120,
      "train_runtime": 29082.5544,
      "train_tokens_per_second": 10144.456
    },
    {
      "epoch": 1.1668943533697633,
      "grad_norm": 0.9750267296749378,
      "learning_rate": 3.705904774487396e-06,
      "loss": 0.266,
      "num_input_tokens_seen": 295304024,
      "step": 5125,
      "train_runtime": 29110.8567,
      "train_tokens_per_second": 10144.12
    },
    {
      "epoch": 1.1680327868852458,
      "grad_norm": 0.6888640692754979,
      "learning_rate": 3.6972702775407554e-06,
      "loss": 0.2655,
      "num_input_tokens_seen": 295576040,
      "step": 5130,
      "train_runtime": 29142.6938,
      "train_tokens_per_second": 10142.372
    },
    {
      "epoch": 1.1691712204007285,
      "grad_norm": 0.7277945638047794,
      "learning_rate": 3.688639946498578e-06,
      "loss": 0.2512,
      "num_input_tokens_seen": 295871088,
      "step": 5135,
      "train_runtime": 29167.7345,
      "train_tokens_per_second": 10143.78
    },
    {
      "epoch": 1.1703096539162112,
      "grad_norm": 0.5569129701340273,
      "learning_rate": 3.6800138089591687e-06,
      "loss": 0.249,
      "num_input_tokens_seen": 296163848,
      "step": 5140,
      "train_runtime": 29196.7618,
      "train_tokens_per_second": 10143.722
    },
    {
      "epoch": 1.171448087431694,
      "grad_norm": 0.5345678245292997,
      "learning_rate": 3.67139189250743e-06,
      "loss": 0.2276,
      "num_input_tokens_seen": 296448744,
      "step": 5145,
      "train_runtime": 29221.2424,
      "train_tokens_per_second": 10144.974
    },
    {
      "epoch": 1.1725865209471766,
      "grad_norm": 0.6197139263046183,
      "learning_rate": 3.662774224714754e-06,
      "loss": 0.2538,
      "num_input_tokens_seen": 296732480,
      "step": 5150,
      "train_runtime": 29249.4529,
      "train_tokens_per_second": 10144.89
    },
    {
      "epoch": 1.1737249544626593,
      "grad_norm": 0.6434161653336743,
      "learning_rate": 3.6541608331389543e-06,
      "loss": 0.2582,
      "num_input_tokens_seen": 297020904,
      "step": 5155,
      "train_runtime": 29279.3241,
      "train_tokens_per_second": 10144.391
    },
    {
      "epoch": 1.174863387978142,
      "grad_norm": 0.6258290041234156,
      "learning_rate": 3.6455517453241697e-06,
      "loss": 0.2459,
      "num_input_tokens_seen": 297322792,
      "step": 5160,
      "train_runtime": 29311.5218,
      "train_tokens_per_second": 10143.547
    },
    {
      "epoch": 1.1760018214936248,
      "grad_norm": 0.840727068967799,
      "learning_rate": 3.636946988800775e-06,
      "loss": 0.2616,
      "num_input_tokens_seen": 297597792,
      "step": 5165,
      "train_runtime": 29339.2985,
      "train_tokens_per_second": 10143.317
    },
    {
      "epoch": 1.1771402550091075,
      "grad_norm": 0.6737042435853557,
      "learning_rate": 3.628346591085292e-06,
      "loss": 0.2626,
      "num_input_tokens_seen": 297890600,
      "step": 5170,
      "train_runtime": 29365.04,
      "train_tokens_per_second": 10144.396
    },
    {
      "epoch": 1.1782786885245902,
      "grad_norm": 0.6037049811389706,
      "learning_rate": 3.6197505796803077e-06,
      "loss": 0.2666,
      "num_input_tokens_seen": 298178704,
      "step": 5175,
      "train_runtime": 29390.109,
      "train_tokens_per_second": 10145.546
    },
    {
      "epoch": 1.1794171220400729,
      "grad_norm": 0.5503685306200765,
      "learning_rate": 3.6111589820743796e-06,
      "loss": 0.2395,
      "num_input_tokens_seen": 298471984,
      "step": 5180,
      "train_runtime": 29417.8684,
      "train_tokens_per_second": 10145.942
    },
    {
      "epoch": 1.1805555555555556,
      "grad_norm": 0.6562967156291097,
      "learning_rate": 3.6025718257419532e-06,
      "loss": 0.2638,
      "num_input_tokens_seen": 298765904,
      "step": 5185,
      "train_runtime": 29446.4999,
      "train_tokens_per_second": 10146.058
    },
    {
      "epoch": 1.1816939890710383,
      "grad_norm": 0.6570925488290529,
      "learning_rate": 3.5939891381432697e-06,
      "loss": 0.2564,
      "num_input_tokens_seen": 299036704,
      "step": 5190,
      "train_runtime": 29475.2282,
      "train_tokens_per_second": 10145.357
    },
    {
      "epoch": 1.182832422586521,
      "grad_norm": 0.589780024594162,
      "learning_rate": 3.5854109467242794e-06,
      "loss": 0.2766,
      "num_input_tokens_seen": 299341880,
      "step": 5195,
      "train_runtime": 29501.6819,
      "train_tokens_per_second": 10146.604
    },
    {
      "epoch": 1.1839708561020037,
      "grad_norm": 0.6595939000131416,
      "learning_rate": 3.576837278916555e-06,
      "loss": 0.2508,
      "num_input_tokens_seen": 299619344,
      "step": 5200,
      "train_runtime": 29531.7771,
      "train_tokens_per_second": 10145.659
    },
    {
      "epoch": 1.1851092896174864,
      "grad_norm": 0.5786722598333565,
      "learning_rate": 3.5682681621372068e-06,
      "loss": 0.2627,
      "num_input_tokens_seen": 299926144,
      "step": 5205,
      "train_runtime": 29560.4399,
      "train_tokens_per_second": 10146.2
    },
    {
      "epoch": 1.1862477231329691,
      "grad_norm": 0.6166879425989061,
      "learning_rate": 3.5597036237887873e-06,
      "loss": 0.2672,
      "num_input_tokens_seen": 300216392,
      "step": 5210,
      "train_runtime": 29591.5398,
      "train_tokens_per_second": 10145.345
    },
    {
      "epoch": 1.1873861566484518,
      "grad_norm": 0.6116283473750022,
      "learning_rate": 3.5511436912592103e-06,
      "loss": 0.2826,
      "num_input_tokens_seen": 300517048,
      "step": 5215,
      "train_runtime": 29619.6989,
      "train_tokens_per_second": 10145.851
    },
    {
      "epoch": 1.1885245901639343,
      "grad_norm": 0.5874505457373426,
      "learning_rate": 3.5425883919216588e-06,
      "loss": 0.2531,
      "num_input_tokens_seen": 300808696,
      "step": 5220,
      "train_runtime": 29646.8413,
      "train_tokens_per_second": 10146.4
    },
    {
      "epoch": 1.189663023679417,
      "grad_norm": 0.7308211019946039,
      "learning_rate": 3.534037753134502e-06,
      "loss": 0.2462,
      "num_input_tokens_seen": 301085872,
      "step": 5225,
      "train_runtime": 29674.4248,
      "train_tokens_per_second": 10146.309
    },
    {
      "epoch": 1.1908014571948997,
      "grad_norm": 0.7467963106829029,
      "learning_rate": 3.5254918022412067e-06,
      "loss": 0.2731,
      "num_input_tokens_seen": 301372784,
      "step": 5230,
      "train_runtime": 29704.3396,
      "train_tokens_per_second": 10145.749
    },
    {
      "epoch": 1.1919398907103824,
      "grad_norm": 0.6307206886996573,
      "learning_rate": 3.516950566570244e-06,
      "loss": 0.2435,
      "num_input_tokens_seen": 301661808,
      "step": 5235,
      "train_runtime": 29727.1468,
      "train_tokens_per_second": 10147.688
    },
    {
      "epoch": 1.1930783242258651,
      "grad_norm": 0.620267536753748,
      "learning_rate": 3.50841407343501e-06,
      "loss": 0.2744,
      "num_input_tokens_seen": 301952344,
      "step": 5240,
      "train_runtime": 29755.5991,
      "train_tokens_per_second": 10147.749
    },
    {
      "epoch": 1.1942167577413478,
      "grad_norm": 0.6226548195062804,
      "learning_rate": 3.4998823501337354e-06,
      "loss": 0.2468,
      "num_input_tokens_seen": 302251960,
      "step": 5245,
      "train_runtime": 29789.314,
      "train_tokens_per_second": 10146.322
    },
    {
      "epoch": 1.1953551912568305,
      "grad_norm": 0.6001006574693274,
      "learning_rate": 3.491355423949396e-06,
      "loss": 0.2569,
      "num_input_tokens_seen": 302556112,
      "step": 5250,
      "train_runtime": 29818.0044,
      "train_tokens_per_second": 10146.759
    },
    {
      "epoch": 1.1964936247723132,
      "grad_norm": 0.5568045919272258,
      "learning_rate": 3.4828333221496257e-06,
      "loss": 0.2563,
      "num_input_tokens_seen": 302832824,
      "step": 5255,
      "train_runtime": 29846.5757,
      "train_tokens_per_second": 10146.317
    },
    {
      "epoch": 1.197632058287796,
      "grad_norm": 0.6397287680273025,
      "learning_rate": 3.474316071986633e-06,
      "loss": 0.2481,
      "num_input_tokens_seen": 303127520,
      "step": 5260,
      "train_runtime": 29869.7356,
      "train_tokens_per_second": 10148.316
    },
    {
      "epoch": 1.1987704918032787,
      "grad_norm": 0.6942487004353572,
      "learning_rate": 3.4658037006971147e-06,
      "loss": 0.2695,
      "num_input_tokens_seen": 303411408,
      "step": 5265,
      "train_runtime": 29904.6911,
      "train_tokens_per_second": 10145.947
    },
    {
      "epoch": 1.1999089253187614,
      "grad_norm": 0.5857159736017197,
      "learning_rate": 3.4572962355021613e-06,
      "loss": 0.2642,
      "num_input_tokens_seen": 303715112,
      "step": 5270,
      "train_runtime": 29932.561,
      "train_tokens_per_second": 10146.646
    },
    {
      "epoch": 1.201047358834244,
      "grad_norm": 0.5791393018071086,
      "learning_rate": 3.448793703607175e-06,
      "loss": 0.2531,
      "num_input_tokens_seen": 304003960,
      "step": 5275,
      "train_runtime": 29964.1702,
      "train_tokens_per_second": 10145.582
    },
    {
      "epoch": 1.2021857923497268,
      "grad_norm": 0.673715331648879,
      "learning_rate": 3.4402961322017834e-06,
      "loss": 0.2531,
      "num_input_tokens_seen": 304284664,
      "step": 5280,
      "train_runtime": 29991.8325,
      "train_tokens_per_second": 10145.584
    },
    {
      "epoch": 1.2033242258652095,
      "grad_norm": 0.4625179184672645,
      "learning_rate": 3.4318035484597505e-06,
      "loss": 0.2578,
      "num_input_tokens_seen": 304569408,
      "step": 5285,
      "train_runtime": 30024.7131,
      "train_tokens_per_second": 10143.957
    },
    {
      "epoch": 1.2044626593806922,
      "grad_norm": 0.5984363866942993,
      "learning_rate": 3.4233159795388933e-06,
      "loss": 0.2668,
      "num_input_tokens_seen": 304833232,
      "step": 5290,
      "train_runtime": 30052.5183,
      "train_tokens_per_second": 10143.351
    },
    {
      "epoch": 1.205601092896175,
      "grad_norm": 0.6240063197966348,
      "learning_rate": 3.414833452580988e-06,
      "loss": 0.2473,
      "num_input_tokens_seen": 305125264,
      "step": 5295,
      "train_runtime": 30078.7555,
      "train_tokens_per_second": 10144.212
    },
    {
      "epoch": 1.2067395264116576,
      "grad_norm": 0.7039975850954285,
      "learning_rate": 3.406355994711691e-06,
      "loss": 0.2494,
      "num_input_tokens_seen": 305408384,
      "step": 5300,
      "train_runtime": 30107.215,
      "train_tokens_per_second": 10144.026
    },
    {
      "epoch": 1.2078779599271403,
      "grad_norm": 0.7571738199742777,
      "learning_rate": 3.3978836330404453e-06,
      "loss": 0.2606,
      "num_input_tokens_seen": 305663632,
      "step": 5305,
      "train_runtime": 30133.4348,
      "train_tokens_per_second": 10143.67
    },
    {
      "epoch": 1.209016393442623,
      "grad_norm": 0.575767295581172,
      "learning_rate": 3.389416394660399e-06,
      "loss": 0.257,
      "num_input_tokens_seen": 305958688,
      "step": 5310,
      "train_runtime": 30160.2317,
      "train_tokens_per_second": 10144.441
    },
    {
      "epoch": 1.2101548269581057,
      "grad_norm": 0.8704579990401979,
      "learning_rate": 3.380954306648319e-06,
      "loss": 0.2552,
      "num_input_tokens_seen": 306234568,
      "step": 5315,
      "train_runtime": 30190.5338,
      "train_tokens_per_second": 10143.397
    },
    {
      "epoch": 1.2112932604735884,
      "grad_norm": 0.6619867078489228,
      "learning_rate": 3.372497396064499e-06,
      "loss": 0.2335,
      "num_input_tokens_seen": 306511344,
      "step": 5320,
      "train_runtime": 30221.5348,
      "train_tokens_per_second": 10142.15
    },
    {
      "epoch": 1.2124316939890711,
      "grad_norm": 0.5234302396144975,
      "learning_rate": 3.364045689952677e-06,
      "loss": 0.2549,
      "num_input_tokens_seen": 306817744,
      "step": 5325,
      "train_runtime": 30257.3119,
      "train_tokens_per_second": 10140.284
    },
    {
      "epoch": 1.2135701275045538,
      "grad_norm": 0.6632633140667966,
      "learning_rate": 3.3555992153399464e-06,
      "loss": 0.2468,
      "num_input_tokens_seen": 307104232,
      "step": 5330,
      "train_runtime": 30287.5081,
      "train_tokens_per_second": 10139.634
    },
    {
      "epoch": 1.2147085610200365,
      "grad_norm": 0.5150677537611633,
      "learning_rate": 3.3471579992366764e-06,
      "loss": 0.2259,
      "num_input_tokens_seen": 307399016,
      "step": 5335,
      "train_runtime": 30309.41,
      "train_tokens_per_second": 10142.032
    },
    {
      "epoch": 1.215846994535519,
      "grad_norm": 0.5364536094999306,
      "learning_rate": 3.338722068636415e-06,
      "loss": 0.255,
      "num_input_tokens_seen": 307681432,
      "step": 5340,
      "train_runtime": 30336.792,
      "train_tokens_per_second": 10142.187
    },
    {
      "epoch": 1.2169854280510017,
      "grad_norm": 0.6349338381893049,
      "learning_rate": 3.3302914505158107e-06,
      "loss": 0.2601,
      "num_input_tokens_seen": 307960232,
      "step": 5345,
      "train_runtime": 30363.7301,
      "train_tokens_per_second": 10142.372
    },
    {
      "epoch": 1.2181238615664844,
      "grad_norm": 0.5230358587029846,
      "learning_rate": 3.3218661718345224e-06,
      "loss": 0.2436,
      "num_input_tokens_seen": 308269784,
      "step": 5350,
      "train_runtime": 30386.3981,
      "train_tokens_per_second": 10144.993
    },
    {
      "epoch": 1.2192622950819672,
      "grad_norm": 0.5965307581700604,
      "learning_rate": 3.3134462595351356e-06,
      "loss": 0.2768,
      "num_input_tokens_seen": 308551856,
      "step": 5355,
      "train_runtime": 30414.1749,
      "train_tokens_per_second": 10145.002
    },
    {
      "epoch": 1.2204007285974499,
      "grad_norm": 0.7059047361759504,
      "learning_rate": 3.305031740543073e-06,
      "loss": 0.2702,
      "num_input_tokens_seen": 308836016,
      "step": 5360,
      "train_runtime": 30445.6865,
      "train_tokens_per_second": 10143.835
    },
    {
      "epoch": 1.2215391621129326,
      "grad_norm": 0.6020852432473894,
      "learning_rate": 3.2966226417665125e-06,
      "loss": 0.2523,
      "num_input_tokens_seen": 309110368,
      "step": 5365,
      "train_runtime": 30474.4762,
      "train_tokens_per_second": 10143.254
    },
    {
      "epoch": 1.2226775956284153,
      "grad_norm": 0.6108642616229132,
      "learning_rate": 3.288218990096298e-06,
      "loss": 0.2549,
      "num_input_tokens_seen": 309404416,
      "step": 5370,
      "train_runtime": 30507.0678,
      "train_tokens_per_second": 10142.057
    },
    {
      "epoch": 1.223816029143898,
      "grad_norm": 0.6617735364697512,
      "learning_rate": 3.2798208124058563e-06,
      "loss": 0.2624,
      "num_input_tokens_seen": 309689472,
      "step": 5375,
      "train_runtime": 30543.3075,
      "train_tokens_per_second": 10139.356
    },
    {
      "epoch": 1.2249544626593807,
      "grad_norm": 0.5908189962988264,
      "learning_rate": 3.271428135551107e-06,
      "loss": 0.2914,
      "num_input_tokens_seen": 309971496,
      "step": 5380,
      "train_runtime": 30573.1979,
      "train_tokens_per_second": 10138.668
    },
    {
      "epoch": 1.2260928961748634,
      "grad_norm": 0.7020279169485167,
      "learning_rate": 3.2630409863703793e-06,
      "loss": 0.2561,
      "num_input_tokens_seen": 310251336,
      "step": 5385,
      "train_runtime": 30596.7474,
      "train_tokens_per_second": 10140.01
    },
    {
      "epoch": 1.227231329690346,
      "grad_norm": 0.7545115210782255,
      "learning_rate": 3.2546593916843274e-06,
      "loss": 0.2658,
      "num_input_tokens_seen": 310535096,
      "step": 5390,
      "train_runtime": 30618.9552,
      "train_tokens_per_second": 10141.923
    },
    {
      "epoch": 1.2283697632058288,
      "grad_norm": 0.6201262609226462,
      "learning_rate": 3.246283378295842e-06,
      "loss": 0.2488,
      "num_input_tokens_seen": 310821264,
      "step": 5395,
      "train_runtime": 30642.0123,
      "train_tokens_per_second": 10143.631
    },
    {
      "epoch": 1.2295081967213115,
      "grad_norm": 0.6474101772897621,
      "learning_rate": 3.2379129729899668e-06,
      "loss": 0.2466,
      "num_input_tokens_seen": 311124888,
      "step": 5400,
      "train_runtime": 30674.0102,
      "train_tokens_per_second": 10142.948
    },
    {
      "epoch": 1.2306466302367942,
      "grad_norm": 0.6031254810007524,
      "learning_rate": 3.22954820253381e-06,
      "loss": 0.2477,
      "num_input_tokens_seen": 311437672,
      "step": 5405,
      "train_runtime": 30704.8682,
      "train_tokens_per_second": 10142.941
    },
    {
      "epoch": 1.231785063752277,
      "grad_norm": 0.5472118546933852,
      "learning_rate": 3.2211890936764616e-06,
      "loss": 0.2253,
      "num_input_tokens_seen": 311737808,
      "step": 5410,
      "train_runtime": 30738.9034,
      "train_tokens_per_second": 10141.475
    },
    {
      "epoch": 1.2329234972677596,
      "grad_norm": 0.5662590876663182,
      "learning_rate": 3.2128356731489064e-06,
      "loss": 0.2245,
      "num_input_tokens_seen": 312037264,
      "step": 5415,
      "train_runtime": 30770.1565,
      "train_tokens_per_second": 10140.906
    },
    {
      "epoch": 1.2340619307832423,
      "grad_norm": 0.5324664792326701,
      "learning_rate": 3.2044879676639413e-06,
      "loss": 0.2449,
      "num_input_tokens_seen": 312351208,
      "step": 5420,
      "train_runtime": 30797.3919,
      "train_tokens_per_second": 10142.132
    },
    {
      "epoch": 1.235200364298725,
      "grad_norm": 0.6631323852611096,
      "learning_rate": 3.1961460039160842e-06,
      "loss": 0.2919,
      "num_input_tokens_seen": 312614016,
      "step": 5425,
      "train_runtime": 30823.3054,
      "train_tokens_per_second": 10142.131
    },
    {
      "epoch": 1.2363387978142077,
      "grad_norm": 0.6913037602241877,
      "learning_rate": 3.1878098085814926e-06,
      "loss": 0.2515,
      "num_input_tokens_seen": 312891352,
      "step": 5430,
      "train_runtime": 30850.0367,
      "train_tokens_per_second": 10142.333
    },
    {
      "epoch": 1.2374772313296902,
      "grad_norm": 0.599740092837195,
      "learning_rate": 3.179479408317878e-06,
      "loss": 0.2346,
      "num_input_tokens_seen": 313186168,
      "step": 5435,
      "train_runtime": 30884.799,
      "train_tokens_per_second": 10140.463
    },
    {
      "epoch": 1.238615664845173,
      "grad_norm": 0.6742242658283907,
      "learning_rate": 3.171154829764421e-06,
      "loss": 0.2498,
      "num_input_tokens_seen": 313448072,
      "step": 5440,
      "train_runtime": 30909.8005,
      "train_tokens_per_second": 10140.734
    },
    {
      "epoch": 1.2397540983606556,
      "grad_norm": 0.6980203319054186,
      "learning_rate": 3.1628360995416863e-06,
      "loss": 0.2624,
      "num_input_tokens_seen": 313717880,
      "step": 5445,
      "train_runtime": 30940.8352,
      "train_tokens_per_second": 10139.283
    },
    {
      "epoch": 1.2408925318761383,
      "grad_norm": 0.6553639963769873,
      "learning_rate": 3.1545232442515337e-06,
      "loss": 0.2804,
      "num_input_tokens_seen": 314007344,
      "step": 5450,
      "train_runtime": 30963.226,
      "train_tokens_per_second": 10141.299
    },
    {
      "epoch": 1.242030965391621,
      "grad_norm": 0.638196688482732,
      "learning_rate": 3.1462162904770376e-06,
      "loss": 0.2498,
      "num_input_tokens_seen": 314301360,
      "step": 5455,
      "train_runtime": 30986.1481,
      "train_tokens_per_second": 10143.286
    },
    {
      "epoch": 1.2431693989071038,
      "grad_norm": 0.6516347068563381,
      "learning_rate": 3.1379152647824017e-06,
      "loss": 0.2625,
      "num_input_tokens_seen": 314598968,
      "step": 5460,
      "train_runtime": 31007.5861,
      "train_tokens_per_second": 10145.871
    },
    {
      "epoch": 1.2443078324225865,
      "grad_norm": 0.5353819700827648,
      "learning_rate": 3.129620193712868e-06,
      "loss": 0.2345,
      "num_input_tokens_seen": 314886392,
      "step": 5465,
      "train_runtime": 31036.7286,
      "train_tokens_per_second": 10145.605
    },
    {
      "epoch": 1.2454462659380692,
      "grad_norm": 0.6794908760959474,
      "learning_rate": 3.121331103794642e-06,
      "loss": 0.2879,
      "num_input_tokens_seen": 315157792,
      "step": 5470,
      "train_runtime": 31062.0148,
      "train_tokens_per_second": 10146.083
    },
    {
      "epoch": 1.2465846994535519,
      "grad_norm": 0.6072378951538004,
      "learning_rate": 3.113048021534799e-06,
      "loss": 0.2436,
      "num_input_tokens_seen": 315444392,
      "step": 5475,
      "train_runtime": 31096.9695,
      "train_tokens_per_second": 10143.895
    },
    {
      "epoch": 1.2477231329690346,
      "grad_norm": 0.6106580285436416,
      "learning_rate": 3.1047709734212074e-06,
      "loss": 0.2456,
      "num_input_tokens_seen": 315737288,
      "step": 5480,
      "train_runtime": 31131.705,
      "train_tokens_per_second": 10141.985
    },
    {
      "epoch": 1.2488615664845173,
      "grad_norm": 0.5629404589152497,
      "learning_rate": 3.096499985922433e-06,
      "loss": 0.258,
      "num_input_tokens_seen": 316023576,
      "step": 5485,
      "train_runtime": 31157.4068,
      "train_tokens_per_second": 10142.807
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.7092844783938951,
      "learning_rate": 3.0882350854876657e-06,
      "loss": 0.256,
      "num_input_tokens_seen": 316311048,
      "step": 5490,
      "train_runtime": 31179.2379,
      "train_tokens_per_second": 10144.926
    },
    {
      "epoch": 1.2511384335154827,
      "grad_norm": 0.603244260223401,
      "learning_rate": 3.0799762985466284e-06,
      "loss": 0.2399,
      "num_input_tokens_seen": 316600352,
      "step": 5495,
      "train_runtime": 31208.8505,
      "train_tokens_per_second": 10144.569
    },
    {
      "epoch": 1.2522768670309654,
      "grad_norm": 0.5978951127695098,
      "learning_rate": 3.071723651509493e-06,
      "loss": 0.241,
      "num_input_tokens_seen": 316885280,
      "step": 5500,
      "train_runtime": 31240.5097,
      "train_tokens_per_second": 10143.409
    },
    {
      "epoch": 1.2534153005464481,
      "grad_norm": 0.7111466785614845,
      "learning_rate": 3.0634771707668e-06,
      "loss": 0.2652,
      "num_input_tokens_seen": 317191672,
      "step": 5505,
      "train_runtime": 31269.7541,
      "train_tokens_per_second": 10143.721
    },
    {
      "epoch": 1.2545537340619308,
      "grad_norm": 0.543064864478753,
      "learning_rate": 3.0552368826893677e-06,
      "loss": 0.2277,
      "num_input_tokens_seen": 317492416,
      "step": 5510,
      "train_runtime": 31294.7214,
      "train_tokens_per_second": 10145.239
    },
    {
      "epoch": 1.2556921675774135,
      "grad_norm": 0.6600870987168199,
      "learning_rate": 3.0470028136282136e-06,
      "loss": 0.239,
      "num_input_tokens_seen": 317754904,
      "step": 5515,
      "train_runtime": 31325.0529,
      "train_tokens_per_second": 10143.795
    },
    {
      "epoch": 1.2568306010928962,
      "grad_norm": 0.5489882725413457,
      "learning_rate": 3.0387749899144668e-06,
      "loss": 0.2535,
      "num_input_tokens_seen": 318027688,
      "step": 5520,
      "train_runtime": 31356.923,
      "train_tokens_per_second": 10142.184
    },
    {
      "epoch": 1.257969034608379,
      "grad_norm": 0.5400317068880017,
      "learning_rate": 3.0305534378592868e-06,
      "loss": 0.2601,
      "num_input_tokens_seen": 318318792,
      "step": 5525,
      "train_runtime": 31386.6382,
      "train_tokens_per_second": 10141.857
    },
    {
      "epoch": 1.2591074681238617,
      "grad_norm": 0.5911901538294119,
      "learning_rate": 3.0223381837537746e-06,
      "loss": 0.2468,
      "num_input_tokens_seen": 318598080,
      "step": 5530,
      "train_runtime": 31422.2137,
      "train_tokens_per_second": 10139.263
    },
    {
      "epoch": 1.2602459016393444,
      "grad_norm": 0.6105068988460803,
      "learning_rate": 3.0141292538688925e-06,
      "loss": 0.228,
      "num_input_tokens_seen": 318899376,
      "step": 5535,
      "train_runtime": 31449.5622,
      "train_tokens_per_second": 10140.026
    },
    {
      "epoch": 1.261384335154827,
      "grad_norm": 0.6439167134247245,
      "learning_rate": 3.0059266744553805e-06,
      "loss": 0.2288,
      "num_input_tokens_seen": 319201080,
      "step": 5540,
      "train_runtime": 31480.8917,
      "train_tokens_per_second": 10139.518
    },
    {
      "epoch": 1.2625227686703098,
      "grad_norm": 0.5358817450751882,
      "learning_rate": 2.997730471743667e-06,
      "loss": 0.2443,
      "num_input_tokens_seen": 319478328,
      "step": 5545,
      "train_runtime": 31508.5044,
      "train_tokens_per_second": 10139.432
    },
    {
      "epoch": 1.2636612021857925,
      "grad_norm": 0.5568235863048291,
      "learning_rate": 2.989540671943795e-06,
      "loss": 0.2606,
      "num_input_tokens_seen": 319752992,
      "step": 5550,
      "train_runtime": 31533.9542,
      "train_tokens_per_second": 10139.959
    },
    {
      "epoch": 1.2647996357012752,
      "grad_norm": 0.5681118216109174,
      "learning_rate": 2.9813573012453265e-06,
      "loss": 0.2707,
      "num_input_tokens_seen": 320027784,
      "step": 5555,
      "train_runtime": 31567.2712,
      "train_tokens_per_second": 10137.962
    },
    {
      "epoch": 1.2659380692167577,
      "grad_norm": 0.5692831894239662,
      "learning_rate": 2.9731803858172683e-06,
      "loss": 0.2431,
      "num_input_tokens_seen": 320316800,
      "step": 5560,
      "train_runtime": 31600.0145,
      "train_tokens_per_second": 10136.603
    },
    {
      "epoch": 1.2670765027322404,
      "grad_norm": 0.6349708756611019,
      "learning_rate": 2.9650099518079807e-06,
      "loss": 0.254,
      "num_input_tokens_seen": 320601752,
      "step": 5565,
      "train_runtime": 31628.7567,
      "train_tokens_per_second": 10136.401
    },
    {
      "epoch": 1.268214936247723,
      "grad_norm": 0.7154468647359359,
      "learning_rate": 2.956846025345099e-06,
      "loss": 0.3064,
      "num_input_tokens_seen": 320897080,
      "step": 5570,
      "train_runtime": 31653.8536,
      "train_tokens_per_second": 10137.694
    },
    {
      "epoch": 1.2693533697632058,
      "grad_norm": 0.7136490961146807,
      "learning_rate": 2.948688632535449e-06,
      "loss": 0.2605,
      "num_input_tokens_seen": 321185312,
      "step": 5575,
      "train_runtime": 31680.4396,
      "train_tokens_per_second": 10138.285
    },
    {
      "epoch": 1.2704918032786885,
      "grad_norm": 0.5847093746927353,
      "learning_rate": 2.940537799464963e-06,
      "loss": 0.2324,
      "num_input_tokens_seen": 321497560,
      "step": 5580,
      "train_runtime": 31706.4314,
      "train_tokens_per_second": 10139.822
    },
    {
      "epoch": 1.2716302367941712,
      "grad_norm": 0.6887719984960409,
      "learning_rate": 2.932393552198597e-06,
      "loss": 0.2529,
      "num_input_tokens_seen": 321801208,
      "step": 5585,
      "train_runtime": 31739.4937,
      "train_tokens_per_second": 10138.826
    },
    {
      "epoch": 1.272768670309654,
      "grad_norm": 0.6015937433951571,
      "learning_rate": 2.924255916780246e-06,
      "loss": 0.2395,
      "num_input_tokens_seen": 322093072,
      "step": 5590,
      "train_runtime": 31769.422,
      "train_tokens_per_second": 10138.462
    },
    {
      "epoch": 1.2739071038251366,
      "grad_norm": 0.6305797177365842,
      "learning_rate": 2.9161249192326602e-06,
      "loss": 0.2466,
      "num_input_tokens_seen": 322392264,
      "step": 5595,
      "train_runtime": 31803.288,
      "train_tokens_per_second": 10137.073
    },
    {
      "epoch": 1.2750455373406193,
      "grad_norm": 0.7107567615674324,
      "learning_rate": 2.9080005855573657e-06,
      "loss": 0.2759,
      "num_input_tokens_seen": 322677704,
      "step": 5600,
      "train_runtime": 31829.7132,
      "train_tokens_per_second": 10137.625
    },
    {
      "epoch": 1.276183970856102,
      "grad_norm": 0.6090934966910185,
      "learning_rate": 2.8998829417345764e-06,
      "loss": 0.2571,
      "num_input_tokens_seen": 322965416,
      "step": 5605,
      "train_runtime": 31859.1235,
      "train_tokens_per_second": 10137.298
    },
    {
      "epoch": 1.2773224043715847,
      "grad_norm": 0.5376320632742689,
      "learning_rate": 2.8917720137231133e-06,
      "loss": 0.249,
      "num_input_tokens_seen": 323251048,
      "step": 5610,
      "train_runtime": 31894.1237,
      "train_tokens_per_second": 10135.129
    },
    {
      "epoch": 1.2784608378870674,
      "grad_norm": 0.6480975904814603,
      "learning_rate": 2.8836678274603212e-06,
      "loss": 0.2515,
      "num_input_tokens_seen": 323529832,
      "step": 5615,
      "train_runtime": 31922.6193,
      "train_tokens_per_second": 10134.815
    },
    {
      "epoch": 1.2795992714025501,
      "grad_norm": 0.6216875725776184,
      "learning_rate": 2.8755704088619906e-06,
      "loss": 0.2639,
      "num_input_tokens_seen": 323824352,
      "step": 5620,
      "train_runtime": 31952.9839,
      "train_tokens_per_second": 10134.401
    },
    {
      "epoch": 1.2807377049180328,
      "grad_norm": 0.8043987419969916,
      "learning_rate": 2.8674797838222645e-06,
      "loss": 0.2649,
      "num_input_tokens_seen": 324127776,
      "step": 5625,
      "train_runtime": 31977.8004,
      "train_tokens_per_second": 10136.025
    },
    {
      "epoch": 1.2818761384335156,
      "grad_norm": 0.8200659150178067,
      "learning_rate": 2.859395978213564e-06,
      "loss": 0.2484,
      "num_input_tokens_seen": 324438520,
      "step": 5630,
      "train_runtime": 32000.5755,
      "train_tokens_per_second": 10138.521
    },
    {
      "epoch": 1.2830145719489983,
      "grad_norm": 0.7087999059022124,
      "learning_rate": 2.8513190178865004e-06,
      "loss": 0.2372,
      "num_input_tokens_seen": 324735864,
      "step": 5635,
      "train_runtime": 32028.0868,
      "train_tokens_per_second": 10139.097
    },
    {
      "epoch": 1.2841530054644807,
      "grad_norm": 0.541162902528671,
      "learning_rate": 2.8432489286697974e-06,
      "loss": 0.2454,
      "num_input_tokens_seen": 325027096,
      "step": 5640,
      "train_runtime": 32057.63,
      "train_tokens_per_second": 10138.837
    },
    {
      "epoch": 1.2852914389799635,
      "grad_norm": 0.6110954582414083,
      "learning_rate": 2.835185736370205e-06,
      "loss": 0.2492,
      "num_input_tokens_seen": 325326600,
      "step": 5645,
      "train_runtime": 32088.9719,
      "train_tokens_per_second": 10138.268
    },
    {
      "epoch": 1.2864298724954462,
      "grad_norm": 0.6550110469846098,
      "learning_rate": 2.82712946677242e-06,
      "loss": 0.25,
      "num_input_tokens_seen": 325627312,
      "step": 5650,
      "train_runtime": 32111.9069,
      "train_tokens_per_second": 10140.392
    },
    {
      "epoch": 1.2875683060109289,
      "grad_norm": 0.5884163928362575,
      "learning_rate": 2.8190801456389975e-06,
      "loss": 0.2626,
      "num_input_tokens_seen": 325911416,
      "step": 5655,
      "train_runtime": 32141.533,
      "train_tokens_per_second": 10139.884
    },
    {
      "epoch": 1.2887067395264116,
      "grad_norm": 0.5794511087727917,
      "learning_rate": 2.8110377987102773e-06,
      "loss": 0.264,
      "num_input_tokens_seen": 326197456,
      "step": 5660,
      "train_runtime": 32167.0751,
      "train_tokens_per_second": 10140.725
    },
    {
      "epoch": 1.2898451730418943,
      "grad_norm": 0.6548311437540509,
      "learning_rate": 2.8030024517042907e-06,
      "loss": 0.282,
      "num_input_tokens_seen": 326482080,
      "step": 5665,
      "train_runtime": 32196.7669,
      "train_tokens_per_second": 10140.213
    },
    {
      "epoch": 1.290983606557377,
      "grad_norm": 0.6544789591897539,
      "learning_rate": 2.7949741303166945e-06,
      "loss": 0.2337,
      "num_input_tokens_seen": 326771792,
      "step": 5670,
      "train_runtime": 32222.7817,
      "train_tokens_per_second": 10141.017
    },
    {
      "epoch": 1.2921220400728597,
      "grad_norm": 0.640884668980645,
      "learning_rate": 2.7869528602206714e-06,
      "loss": 0.2509,
      "num_input_tokens_seen": 327061776,
      "step": 5675,
      "train_runtime": 32249.3734,
      "train_tokens_per_second": 10141.647
    },
    {
      "epoch": 1.2932604735883424,
      "grad_norm": 0.5212365661146506,
      "learning_rate": 2.778938667066853e-06,
      "loss": 0.2585,
      "num_input_tokens_seen": 327360872,
      "step": 5680,
      "train_runtime": 32280.0389,
      "train_tokens_per_second": 10141.279
    },
    {
      "epoch": 1.294398907103825,
      "grad_norm": 0.584055546207017,
      "learning_rate": 2.770931576483245e-06,
      "loss": 0.242,
      "num_input_tokens_seen": 327656024,
      "step": 5685,
      "train_runtime": 32313.3461,
      "train_tokens_per_second": 10139.96
    },
    {
      "epoch": 1.2955373406193078,
      "grad_norm": 1.2316559240812461,
      "learning_rate": 2.7629316140751407e-06,
      "loss": 0.2466,
      "num_input_tokens_seen": 327942568,
      "step": 5690,
      "train_runtime": 32345.5408,
      "train_tokens_per_second": 10138.726
    },
    {
      "epoch": 1.2966757741347905,
      "grad_norm": 0.6694655515888192,
      "learning_rate": 2.754938805425035e-06,
      "loss": 0.2648,
      "num_input_tokens_seen": 328225416,
      "step": 5695,
      "train_runtime": 32372.0607,
      "train_tokens_per_second": 10139.157
    },
    {
      "epoch": 1.2978142076502732,
      "grad_norm": 0.67945510182789,
      "learning_rate": 2.7469531760925473e-06,
      "loss": 0.2661,
      "num_input_tokens_seen": 328518384,
      "step": 5700,
      "train_runtime": 32403.3386,
      "train_tokens_per_second": 10138.412
    },
    {
      "epoch": 1.298952641165756,
      "grad_norm": 0.7497583209526116,
      "learning_rate": 2.738974751614343e-06,
      "loss": 0.2438,
      "num_input_tokens_seen": 328826552,
      "step": 5705,
      "train_runtime": 32432.479,
      "train_tokens_per_second": 10138.804
    },
    {
      "epoch": 1.3000910746812386,
      "grad_norm": 0.6300060229079795,
      "learning_rate": 2.7310035575040438e-06,
      "loss": 0.2577,
      "num_input_tokens_seen": 329127704,
      "step": 5710,
      "train_runtime": 32460.9266,
      "train_tokens_per_second": 10139.196
    },
    {
      "epoch": 1.3012295081967213,
      "grad_norm": 0.6788952366302726,
      "learning_rate": 2.72303961925215e-06,
      "loss": 0.2536,
      "num_input_tokens_seen": 329429424,
      "step": 5715,
      "train_runtime": 32484.5234,
      "train_tokens_per_second": 10141.119
    },
    {
      "epoch": 1.302367941712204,
      "grad_norm": 0.6723040833868934,
      "learning_rate": 2.71508296232596e-06,
      "loss": 0.2686,
      "num_input_tokens_seen": 329714016,
      "step": 5720,
      "train_runtime": 32519.7866,
      "train_tokens_per_second": 10138.874
    },
    {
      "epoch": 1.3035063752276868,
      "grad_norm": 0.6497549675946064,
      "learning_rate": 2.7071336121694856e-06,
      "loss": 0.237,
      "num_input_tokens_seen": 329976080,
      "step": 5725,
      "train_runtime": 32546.7544,
      "train_tokens_per_second": 10138.525
    },
    {
      "epoch": 1.3046448087431695,
      "grad_norm": 0.7481364340414876,
      "learning_rate": 2.6991915942033766e-06,
      "loss": 0.2449,
      "num_input_tokens_seen": 330258912,
      "step": 5730,
      "train_runtime": 32575.4347,
      "train_tokens_per_second": 10138.281
    },
    {
      "epoch": 1.3057832422586522,
      "grad_norm": 0.5610910613750112,
      "learning_rate": 2.6912569338248317e-06,
      "loss": 0.2577,
      "num_input_tokens_seen": 330544336,
      "step": 5735,
      "train_runtime": 32605.3292,
      "train_tokens_per_second": 10137.74
    },
    {
      "epoch": 1.3069216757741349,
      "grad_norm": 0.7252810987696005,
      "learning_rate": 2.683329656407524e-06,
      "loss": 0.2518,
      "num_input_tokens_seen": 330822880,
      "step": 5740,
      "train_runtime": 32632.5709,
      "train_tokens_per_second": 10137.812
    },
    {
      "epoch": 1.3080601092896176,
      "grad_norm": 0.6583911589645147,
      "learning_rate": 2.6754097873015152e-06,
      "loss": 0.2462,
      "num_input_tokens_seen": 331109672,
      "step": 5745,
      "train_runtime": 32661.156,
      "train_tokens_per_second": 10137.721
    },
    {
      "epoch": 1.3091985428051003,
      "grad_norm": 0.6927158643726832,
      "learning_rate": 2.66749735183318e-06,
      "loss": 0.2673,
      "num_input_tokens_seen": 331418536,
      "step": 5750,
      "train_runtime": 32694.054,
      "train_tokens_per_second": 10136.967
    },
    {
      "epoch": 1.310336976320583,
      "grad_norm": 0.6901129997356825,
      "learning_rate": 2.6595923753051185e-06,
      "loss": 0.2591,
      "num_input_tokens_seen": 331713240,
      "step": 5755,
      "train_runtime": 32721.062,
      "train_tokens_per_second": 10137.606
    },
    {
      "epoch": 1.3114754098360657,
      "grad_norm": 0.6683885784135482,
      "learning_rate": 2.651694882996079e-06,
      "loss": 0.2695,
      "num_input_tokens_seen": 331992232,
      "step": 5760,
      "train_runtime": 32746.5114,
      "train_tokens_per_second": 10138.247
    },
    {
      "epoch": 1.3126138433515484,
      "grad_norm": 0.6298581460490142,
      "learning_rate": 2.643804900160877e-06,
      "loss": 0.2441,
      "num_input_tokens_seen": 332255432,
      "step": 5765,
      "train_runtime": 32773.7412,
      "train_tokens_per_second": 10137.855
    },
    {
      "epoch": 1.3137522768670309,
      "grad_norm": 0.6496588895358292,
      "learning_rate": 2.635922452030313e-06,
      "loss": 0.2419,
      "num_input_tokens_seen": 332556864,
      "step": 5770,
      "train_runtime": 32800.1514,
      "train_tokens_per_second": 10138.882
    },
    {
      "epoch": 1.3148907103825136,
      "grad_norm": 0.7240151205139206,
      "learning_rate": 2.6280475638110937e-06,
      "loss": 0.2755,
      "num_input_tokens_seen": 332848304,
      "step": 5775,
      "train_runtime": 32828.7271,
      "train_tokens_per_second": 10138.934
    },
    {
      "epoch": 1.3160291438979963,
      "grad_norm": 0.5968393954396605,
      "learning_rate": 2.6201802606857523e-06,
      "loss": 0.279,
      "num_input_tokens_seen": 333145512,
      "step": 5780,
      "train_runtime": 32862.0292,
      "train_tokens_per_second": 10137.704
    },
    {
      "epoch": 1.317167577413479,
      "grad_norm": 0.6437532357745029,
      "learning_rate": 2.6123205678125597e-06,
      "loss": 0.2355,
      "num_input_tokens_seen": 333440416,
      "step": 5785,
      "train_runtime": 32884.5756,
      "train_tokens_per_second": 10139.721
    },
    {
      "epoch": 1.3183060109289617,
      "grad_norm": 0.6086923091422616,
      "learning_rate": 2.6044685103254586e-06,
      "loss": 0.2637,
      "num_input_tokens_seen": 333704024,
      "step": 5790,
      "train_runtime": 32916.5805,
      "train_tokens_per_second": 10137.87
    },
    {
      "epoch": 1.3194444444444444,
      "grad_norm": 0.6143244359717525,
      "learning_rate": 2.596624113333972e-06,
      "loss": 0.27,
      "num_input_tokens_seen": 333972696,
      "step": 5795,
      "train_runtime": 32947.103,
      "train_tokens_per_second": 10136.633
    },
    {
      "epoch": 1.3205828779599271,
      "grad_norm": 0.5500933857521771,
      "learning_rate": 2.588787401923124e-06,
      "loss": 0.2422,
      "num_input_tokens_seen": 334262848,
      "step": 5800,
      "train_runtime": 32966.4306,
      "train_tokens_per_second": 10139.492
    },
    {
      "epoch": 1.3217213114754098,
      "grad_norm": 0.707889895434306,
      "learning_rate": 2.5809584011533635e-06,
      "loss": 0.2797,
      "num_input_tokens_seen": 334564784,
      "step": 5805,
      "train_runtime": 32996.7825,
      "train_tokens_per_second": 10139.315
    },
    {
      "epoch": 1.3228597449908925,
      "grad_norm": 0.5360287553963669,
      "learning_rate": 2.573137136060482e-06,
      "loss": 0.2736,
      "num_input_tokens_seen": 334861912,
      "step": 5810,
      "train_runtime": 33019.6396,
      "train_tokens_per_second": 10141.295
    },
    {
      "epoch": 1.3239981785063752,
      "grad_norm": 0.6674896192068114,
      "learning_rate": 2.565323631655532e-06,
      "loss": 0.2477,
      "num_input_tokens_seen": 335148112,
      "step": 5815,
      "train_runtime": 33050.1083,
      "train_tokens_per_second": 10140.606
    },
    {
      "epoch": 1.325136612021858,
      "grad_norm": 0.6232748549565426,
      "learning_rate": 2.5575179129247506e-06,
      "loss": 0.2578,
      "num_input_tokens_seen": 335435832,
      "step": 5820,
      "train_runtime": 33076.9653,
      "train_tokens_per_second": 10141.07
    },
    {
      "epoch": 1.3262750455373407,
      "grad_norm": 0.5762312017401019,
      "learning_rate": 2.5497200048294782e-06,
      "loss": 0.2737,
      "num_input_tokens_seen": 335730104,
      "step": 5825,
      "train_runtime": 33107.3388,
      "train_tokens_per_second": 10140.655
    },
    {
      "epoch": 1.3274134790528234,
      "grad_norm": 0.6893495064192422,
      "learning_rate": 2.5419299323060744e-06,
      "loss": 0.2501,
      "num_input_tokens_seen": 336019312,
      "step": 5830,
      "train_runtime": 33133.9372,
      "train_tokens_per_second": 10141.243
    },
    {
      "epoch": 1.328551912568306,
      "grad_norm": 0.6023020684564443,
      "learning_rate": 2.534147720265847e-06,
      "loss": 0.2553,
      "num_input_tokens_seen": 336318104,
      "step": 5835,
      "train_runtime": 33157.8448,
      "train_tokens_per_second": 10142.942
    },
    {
      "epoch": 1.3296903460837888,
      "grad_norm": 0.6747623176591974,
      "learning_rate": 2.526373393594964e-06,
      "loss": 0.2515,
      "num_input_tokens_seen": 336591672,
      "step": 5840,
      "train_runtime": 33186.5121,
      "train_tokens_per_second": 10142.424
    },
    {
      "epoch": 1.3308287795992715,
      "grad_norm": 0.6467051614666991,
      "learning_rate": 2.5186069771543787e-06,
      "loss": 0.2742,
      "num_input_tokens_seen": 336873984,
      "step": 5845,
      "train_runtime": 33214.2568,
      "train_tokens_per_second": 10142.451
    },
    {
      "epoch": 1.331967213114754,
      "grad_norm": 0.6386747854181041,
      "learning_rate": 2.5108484957797476e-06,
      "loss": 0.27,
      "num_input_tokens_seen": 337172328,
      "step": 5850,
      "train_runtime": 33242.5408,
      "train_tokens_per_second": 10142.797
    },
    {
      "epoch": 1.3331056466302367,
      "grad_norm": 0.5942077801740038,
      "learning_rate": 2.5030979742813533e-06,
      "loss": 0.2356,
      "num_input_tokens_seen": 337467080,
      "step": 5855,
      "train_runtime": 33271.091,
      "train_tokens_per_second": 10142.952
    },
    {
      "epoch": 1.3342440801457194,
      "grad_norm": 0.6746455667463985,
      "learning_rate": 2.4953554374440234e-06,
      "loss": 0.2551,
      "num_input_tokens_seen": 337739232,
      "step": 5860,
      "train_runtime": 33302.1491,
      "train_tokens_per_second": 10141.665
    },
    {
      "epoch": 1.335382513661202,
      "grad_norm": 0.5738909712239345,
      "learning_rate": 2.4876209100270538e-06,
      "loss": 0.2361,
      "num_input_tokens_seen": 338035424,
      "step": 5865,
      "train_runtime": 33326.4795,
      "train_tokens_per_second": 10143.148
    },
    {
      "epoch": 1.3365209471766848,
      "grad_norm": 0.5820312730074446,
      "learning_rate": 2.4798944167641243e-06,
      "loss": 0.251,
      "num_input_tokens_seen": 338317384,
      "step": 5870,
      "train_runtime": 33354.2508,
      "train_tokens_per_second": 10143.156
    },
    {
      "epoch": 1.3376593806921675,
      "grad_norm": 0.5179428090017822,
      "learning_rate": 2.472175982363224e-06,
      "loss": 0.2277,
      "num_input_tokens_seen": 338606088,
      "step": 5875,
      "train_runtime": 33384.034,
      "train_tokens_per_second": 10142.755
    },
    {
      "epoch": 1.3387978142076502,
      "grad_norm": 0.6550236675610305,
      "learning_rate": 2.464465631506575e-06,
      "loss": 0.2472,
      "num_input_tokens_seen": 338904528,
      "step": 5880,
      "train_runtime": 33419.1433,
      "train_tokens_per_second": 10141.03
    },
    {
      "epoch": 1.339936247723133,
      "grad_norm": 0.6281847682867405,
      "learning_rate": 2.456763388850548e-06,
      "loss": 0.2518,
      "num_input_tokens_seen": 339185024,
      "step": 5885,
      "train_runtime": 33450.532,
      "train_tokens_per_second": 10139.899
    },
    {
      "epoch": 1.3410746812386156,
      "grad_norm": 0.570636467142239,
      "learning_rate": 2.449069279025577e-06,
      "loss": 0.2421,
      "num_input_tokens_seen": 339477488,
      "step": 5890,
      "train_runtime": 33478.1673,
      "train_tokens_per_second": 10140.265
    },
    {
      "epoch": 1.3422131147540983,
      "grad_norm": 0.6132800513535002,
      "learning_rate": 2.4413833266361005e-06,
      "loss": 0.2378,
      "num_input_tokens_seen": 339758224,
      "step": 5895,
      "train_runtime": 33504.0992,
      "train_tokens_per_second": 10140.796
    },
    {
      "epoch": 1.343351548269581,
      "grad_norm": 0.6417535422561211,
      "learning_rate": 2.433705556260463e-06,
      "loss": 0.212,
      "num_input_tokens_seen": 340075160,
      "step": 5900,
      "train_runtime": 33529.3093,
      "train_tokens_per_second": 10142.623
    },
    {
      "epoch": 1.3444899817850637,
      "grad_norm": 0.7053606568914449,
      "learning_rate": 2.426035992450848e-06,
      "loss": 0.2606,
      "num_input_tokens_seen": 340372880,
      "step": 5905,
      "train_runtime": 33558.0873,
      "train_tokens_per_second": 10142.797
    },
    {
      "epoch": 1.3456284153005464,
      "grad_norm": 0.532202852966944,
      "learning_rate": 2.418374659733195e-06,
      "loss": 0.2397,
      "num_input_tokens_seen": 340674704,
      "step": 5910,
      "train_runtime": 33585.5146,
      "train_tokens_per_second": 10143.501
    },
    {
      "epoch": 1.3467668488160291,
      "grad_norm": 0.5128399360657443,
      "learning_rate": 2.4107215826071195e-06,
      "loss": 0.2526,
      "num_input_tokens_seen": 340955968,
      "step": 5915,
      "train_runtime": 33610.8446,
      "train_tokens_per_second": 10144.225
    },
    {
      "epoch": 1.3479052823315119,
      "grad_norm": 0.5870566566643344,
      "learning_rate": 2.4030767855458436e-06,
      "loss": 0.2387,
      "num_input_tokens_seen": 341248144,
      "step": 5920,
      "train_runtime": 33642.4111,
      "train_tokens_per_second": 10143.391
    },
    {
      "epoch": 1.3490437158469946,
      "grad_norm": 0.5536368956189419,
      "learning_rate": 2.395440292996104e-06,
      "loss": 0.2612,
      "num_input_tokens_seen": 341549360,
      "step": 5925,
      "train_runtime": 33671.6511,
      "train_tokens_per_second": 10143.529
    },
    {
      "epoch": 1.3501821493624773,
      "grad_norm": 0.6285864766246649,
      "learning_rate": 2.3878121293780843e-06,
      "loss": 0.2263,
      "num_input_tokens_seen": 341849568,
      "step": 5930,
      "train_runtime": 33701.3139,
      "train_tokens_per_second": 10143.509
    },
    {
      "epoch": 1.35132058287796,
      "grad_norm": 0.646066370614563,
      "learning_rate": 2.3801923190853334e-06,
      "loss": 0.2512,
      "num_input_tokens_seen": 342142344,
      "step": 5935,
      "train_runtime": 33736.3414,
      "train_tokens_per_second": 10141.655
    },
    {
      "epoch": 1.3524590163934427,
      "grad_norm": 0.6296200108592718,
      "learning_rate": 2.3725808864846865e-06,
      "loss": 0.2637,
      "num_input_tokens_seen": 342446920,
      "step": 5940,
      "train_runtime": 33770.3329,
      "train_tokens_per_second": 10140.466
    },
    {
      "epoch": 1.3535974499089254,
      "grad_norm": 0.5902831731202708,
      "learning_rate": 2.3649778559161907e-06,
      "loss": 0.2505,
      "num_input_tokens_seen": 342757400,
      "step": 5945,
      "train_runtime": 33797.0275,
      "train_tokens_per_second": 10141.643
    },
    {
      "epoch": 1.354735883424408,
      "grad_norm": 0.65821238969203,
      "learning_rate": 2.357383251693023e-06,
      "loss": 0.2274,
      "num_input_tokens_seen": 343039296,
      "step": 5950,
      "train_runtime": 33822.522,
      "train_tokens_per_second": 10142.333
    },
    {
      "epoch": 1.3558743169398908,
      "grad_norm": 0.650259689808907,
      "learning_rate": 2.349797098101414e-06,
      "loss": 0.253,
      "num_input_tokens_seen": 343336504,
      "step": 5955,
      "train_runtime": 33848.2081,
      "train_tokens_per_second": 10143.417
    },
    {
      "epoch": 1.3570127504553735,
      "grad_norm": 0.584592677395859,
      "learning_rate": 2.342219419400571e-06,
      "loss": 0.2744,
      "num_input_tokens_seen": 343629664,
      "step": 5960,
      "train_runtime": 33878.7434,
      "train_tokens_per_second": 10142.928
    },
    {
      "epoch": 1.3581511839708562,
      "grad_norm": 0.6184541630314334,
      "learning_rate": 2.3346502398226035e-06,
      "loss": 0.2459,
      "num_input_tokens_seen": 343928576,
      "step": 5965,
      "train_runtime": 33902.2602,
      "train_tokens_per_second": 10144.709
    },
    {
      "epoch": 1.359289617486339,
      "grad_norm": 0.6260918123853367,
      "learning_rate": 2.3270895835724377e-06,
      "loss": 0.2454,
      "num_input_tokens_seen": 344222480,
      "step": 5970,
      "train_runtime": 33927.4793,
      "train_tokens_per_second": 10145.831
    },
    {
      "epoch": 1.3604280510018216,
      "grad_norm": 0.6049378243413113,
      "learning_rate": 2.3195374748277453e-06,
      "loss": 0.2599,
      "num_input_tokens_seen": 344519648,
      "step": 5975,
      "train_runtime": 33955.5188,
      "train_tokens_per_second": 10146.205
    },
    {
      "epoch": 1.361566484517304,
      "grad_norm": 0.6280226306419981,
      "learning_rate": 2.3119939377388655e-06,
      "loss": 0.2553,
      "num_input_tokens_seen": 344808816,
      "step": 5980,
      "train_runtime": 33982.3572,
      "train_tokens_per_second": 10146.701
    },
    {
      "epoch": 1.3627049180327868,
      "grad_norm": 0.5864861778402481,
      "learning_rate": 2.3044589964287245e-06,
      "loss": 0.2588,
      "num_input_tokens_seen": 345114616,
      "step": 5985,
      "train_runtime": 34008.3103,
      "train_tokens_per_second": 10147.95
    },
    {
      "epoch": 1.3638433515482695,
      "grad_norm": 0.49348411265947306,
      "learning_rate": 2.296932674992765e-06,
      "loss": 0.2264,
      "num_input_tokens_seen": 345427280,
      "step": 5990,
      "train_runtime": 34035.108,
      "train_tokens_per_second": 10149.146
    },
    {
      "epoch": 1.3649817850637522,
      "grad_norm": 0.6727311830103408,
      "learning_rate": 2.289414997498856e-06,
      "loss": 0.2559,
      "num_input_tokens_seen": 345722656,
      "step": 5995,
      "train_runtime": 34060.0196,
      "train_tokens_per_second": 10150.395
    },
    {
      "epoch": 1.366120218579235,
      "grad_norm": 0.6208357244455375,
      "learning_rate": 2.2819059879872363e-06,
      "loss": 0.2325,
      "num_input_tokens_seen": 346003384,
      "step": 6000,
      "train_runtime": 34084.9067,
      "train_tokens_per_second": 10151.22
    },
    {
      "epoch": 1.3672586520947176,
      "grad_norm": 0.9224274504517604,
      "learning_rate": 2.274405670470418e-06,
      "loss": 0.2595,
      "num_input_tokens_seen": 346297736,
      "step": 6005,
      "train_runtime": 34115.4749,
      "train_tokens_per_second": 10150.752
    },
    {
      "epoch": 1.3683970856102003,
      "grad_norm": 0.6337386632268462,
      "learning_rate": 2.2669140689331208e-06,
      "loss": 0.2342,
      "num_input_tokens_seen": 346582680,
      "step": 6010,
      "train_runtime": 34142.5495,
      "train_tokens_per_second": 10151.049
    },
    {
      "epoch": 1.369535519125683,
      "grad_norm": 0.6855226099555336,
      "learning_rate": 2.2594312073321916e-06,
      "loss": 0.2788,
      "num_input_tokens_seen": 346882792,
      "step": 6015,
      "train_runtime": 34169.0009,
      "train_tokens_per_second": 10151.974
    },
    {
      "epoch": 1.3706739526411658,
      "grad_norm": 0.5575860164928159,
      "learning_rate": 2.251957109596527e-06,
      "loss": 0.2622,
      "num_input_tokens_seen": 347180984,
      "step": 6020,
      "train_runtime": 34202.6456,
      "train_tokens_per_second": 10150.706
    },
    {
      "epoch": 1.3718123861566485,
      "grad_norm": 0.6852393116834301,
      "learning_rate": 2.2444917996270003e-06,
      "loss": 0.2697,
      "num_input_tokens_seen": 347444744,
      "step": 6025,
      "train_runtime": 34235.0236,
      "train_tokens_per_second": 10148.81
    },
    {
      "epoch": 1.3729508196721312,
      "grad_norm": 0.5873670380078109,
      "learning_rate": 2.237035301296383e-06,
      "loss": 0.2374,
      "num_input_tokens_seen": 347748896,
      "step": 6030,
      "train_runtime": 34260.4101,
      "train_tokens_per_second": 10150.167
    },
    {
      "epoch": 1.3740892531876139,
      "grad_norm": 0.6452102734096556,
      "learning_rate": 2.2295876384492663e-06,
      "loss": 0.2467,
      "num_input_tokens_seen": 348031328,
      "step": 6035,
      "train_runtime": 34294.262,
      "train_tokens_per_second": 10148.384
    },
    {
      "epoch": 1.3752276867030966,
      "grad_norm": 0.5912667909733584,
      "learning_rate": 2.2221488349019903e-06,
      "loss": 0.2438,
      "num_input_tokens_seen": 348312432,
      "step": 6040,
      "train_runtime": 34320.4736,
      "train_tokens_per_second": 10148.824
    },
    {
      "epoch": 1.3763661202185793,
      "grad_norm": 0.580797503772085,
      "learning_rate": 2.2147189144425603e-06,
      "loss": 0.242,
      "num_input_tokens_seen": 348597520,
      "step": 6045,
      "train_runtime": 34346.6799,
      "train_tokens_per_second": 10149.38
    },
    {
      "epoch": 1.377504553734062,
      "grad_norm": 0.6606988785927155,
      "learning_rate": 2.2072979008305806e-06,
      "loss": 0.2562,
      "num_input_tokens_seen": 348880504,
      "step": 6050,
      "train_runtime": 34373.2609,
      "train_tokens_per_second": 10149.765
    },
    {
      "epoch": 1.3786429872495447,
      "grad_norm": 0.5760318490662635,
      "learning_rate": 2.199885817797169e-06,
      "loss": 0.2239,
      "num_input_tokens_seen": 349178160,
      "step": 6055,
      "train_runtime": 34404.4434,
      "train_tokens_per_second": 10149.217
    },
    {
      "epoch": 1.3797814207650272,
      "grad_norm": 0.5515324838095865,
      "learning_rate": 2.192482689044885e-06,
      "loss": 0.2327,
      "num_input_tokens_seen": 349468976,
      "step": 6060,
      "train_runtime": 34430.7468,
      "train_tokens_per_second": 10149.91
    },
    {
      "epoch": 1.38091985428051,
      "grad_norm": 0.6386569656816568,
      "learning_rate": 2.185088538247656e-06,
      "loss": 0.2555,
      "num_input_tokens_seen": 349755928,
      "step": 6065,
      "train_runtime": 34461.7289,
      "train_tokens_per_second": 10149.111
    },
    {
      "epoch": 1.3820582877959926,
      "grad_norm": 0.5369894991278035,
      "learning_rate": 2.1777033890506973e-06,
      "loss": 0.2304,
      "num_input_tokens_seen": 350047544,
      "step": 6070,
      "train_runtime": 34487.8367,
      "train_tokens_per_second": 10149.884
    },
    {
      "epoch": 1.3831967213114753,
      "grad_norm": 0.6647221680415647,
      "learning_rate": 2.1703272650704387e-06,
      "loss": 0.2594,
      "num_input_tokens_seen": 350333000,
      "step": 6075,
      "train_runtime": 34519.803,
      "train_tokens_per_second": 10148.754
    },
    {
      "epoch": 1.384335154826958,
      "grad_norm": 0.5505377528521929,
      "learning_rate": 2.16296018989445e-06,
      "loss": 0.229,
      "num_input_tokens_seen": 350645888,
      "step": 6080,
      "train_runtime": 34549.7601,
      "train_tokens_per_second": 10149.011
    },
    {
      "epoch": 1.3854735883424407,
      "grad_norm": 0.5929311990065055,
      "learning_rate": 2.1556021870813653e-06,
      "loss": 0.2487,
      "num_input_tokens_seen": 350938720,
      "step": 6085,
      "train_runtime": 34574.5915,
      "train_tokens_per_second": 10150.191
    },
    {
      "epoch": 1.3866120218579234,
      "grad_norm": 0.683789479698647,
      "learning_rate": 2.148253280160803e-06,
      "loss": 0.2581,
      "num_input_tokens_seen": 351236152,
      "step": 6090,
      "train_runtime": 34607.915,
      "train_tokens_per_second": 10149.012
    },
    {
      "epoch": 1.3877504553734061,
      "grad_norm": 0.6199724652519499,
      "learning_rate": 2.1409134926333023e-06,
      "loss": 0.2436,
      "num_input_tokens_seen": 351528832,
      "step": 6095,
      "train_runtime": 34633.2081,
      "train_tokens_per_second": 10150.051
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 0.663076192692642,
      "learning_rate": 2.1335828479702304e-06,
      "loss": 0.257,
      "num_input_tokens_seen": 351819400,
      "step": 6100,
      "train_runtime": 34665.7397,
      "train_tokens_per_second": 10148.908
    },
    {
      "epoch": 1.3900273224043715,
      "grad_norm": 0.6863608569307176,
      "learning_rate": 2.1262613696137226e-06,
      "loss": 0.2495,
      "num_input_tokens_seen": 352098568,
      "step": 6105,
      "train_runtime": 34694.0878,
      "train_tokens_per_second": 10148.662
    },
    {
      "epoch": 1.3911657559198543,
      "grad_norm": 0.5464717147394285,
      "learning_rate": 2.118949080976601e-06,
      "loss": 0.2659,
      "num_input_tokens_seen": 352390104,
      "step": 6110,
      "train_runtime": 34722.5735,
      "train_tokens_per_second": 10148.732
    },
    {
      "epoch": 1.392304189435337,
      "grad_norm": 0.5514966560454454,
      "learning_rate": 2.111646005442302e-06,
      "loss": 0.2344,
      "num_input_tokens_seen": 352705256,
      "step": 6115,
      "train_runtime": 34748.8987,
      "train_tokens_per_second": 10150.113
    },
    {
      "epoch": 1.3934426229508197,
      "grad_norm": 0.6259506142408993,
      "learning_rate": 2.1043521663647977e-06,
      "loss": 0.2437,
      "num_input_tokens_seen": 353018080,
      "step": 6120,
      "train_runtime": 34770.0107,
      "train_tokens_per_second": 10152.947
    },
    {
      "epoch": 1.3945810564663024,
      "grad_norm": 0.5946209401576331,
      "learning_rate": 2.097067587068523e-06,
      "loss": 0.2386,
      "num_input_tokens_seen": 353315040,
      "step": 6125,
      "train_runtime": 34796.5616,
      "train_tokens_per_second": 10153.734
    },
    {
      "epoch": 1.395719489981785,
      "grad_norm": 0.6194165981723914,
      "learning_rate": 2.089792290848307e-06,
      "loss": 0.2563,
      "num_input_tokens_seen": 353590832,
      "step": 6130,
      "train_runtime": 34823.9154,
      "train_tokens_per_second": 10153.678
    },
    {
      "epoch": 1.3968579234972678,
      "grad_norm": 0.5185023158418115,
      "learning_rate": 2.0825263009692874e-06,
      "loss": 0.2456,
      "num_input_tokens_seen": 353883136,
      "step": 6135,
      "train_runtime": 34854.5899,
      "train_tokens_per_second": 10153.129
    },
    {
      "epoch": 1.3979963570127505,
      "grad_norm": 0.5673493006682085,
      "learning_rate": 2.075269640666844e-06,
      "loss": 0.2417,
      "num_input_tokens_seen": 354177720,
      "step": 6140,
      "train_runtime": 34882.1406,
      "train_tokens_per_second": 10153.555
    },
    {
      "epoch": 1.3991347905282332,
      "grad_norm": 0.6673095541035615,
      "learning_rate": 2.0680223331465223e-06,
      "loss": 0.264,
      "num_input_tokens_seen": 354460512,
      "step": 6145,
      "train_runtime": 34915.0408,
      "train_tokens_per_second": 10152.086
    },
    {
      "epoch": 1.400273224043716,
      "grad_norm": 0.6801187474626353,
      "learning_rate": 2.0607844015839584e-06,
      "loss": 0.2678,
      "num_input_tokens_seen": 354749744,
      "step": 6150,
      "train_runtime": 34943.8596,
      "train_tokens_per_second": 10151.991
    },
    {
      "epoch": 1.4014116575591986,
      "grad_norm": 0.6444613114606411,
      "learning_rate": 2.053555869124807e-06,
      "loss": 0.2334,
      "num_input_tokens_seen": 355045256,
      "step": 6155,
      "train_runtime": 34979.5363,
      "train_tokens_per_second": 10150.085
    },
    {
      "epoch": 1.4025500910746813,
      "grad_norm": 0.5684297874910791,
      "learning_rate": 2.0463367588846644e-06,
      "loss": 0.2174,
      "num_input_tokens_seen": 355353968,
      "step": 6160,
      "train_runtime": 35015.8618,
      "train_tokens_per_second": 10148.371
    },
    {
      "epoch": 1.403688524590164,
      "grad_norm": 0.6645193962912541,
      "learning_rate": 2.0391270939489988e-06,
      "loss": 0.2615,
      "num_input_tokens_seen": 355653712,
      "step": 6165,
      "train_runtime": 35044.6116,
      "train_tokens_per_second": 10148.599
    },
    {
      "epoch": 1.4048269581056467,
      "grad_norm": 0.6945428811207489,
      "learning_rate": 2.031926897373069e-06,
      "loss": 0.2627,
      "num_input_tokens_seen": 355921680,
      "step": 6170,
      "train_runtime": 35079.4907,
      "train_tokens_per_second": 10146.147
    },
    {
      "epoch": 1.4059653916211294,
      "grad_norm": 0.6362288283804102,
      "learning_rate": 2.0247361921818638e-06,
      "loss": 0.2506,
      "num_input_tokens_seen": 356230040,
      "step": 6175,
      "train_runtime": 35104.7362,
      "train_tokens_per_second": 10147.635
    },
    {
      "epoch": 1.4071038251366121,
      "grad_norm": 0.6639355790290074,
      "learning_rate": 2.0175550013700125e-06,
      "loss": 0.2489,
      "num_input_tokens_seen": 356503912,
      "step": 6180,
      "train_runtime": 35131.6919,
      "train_tokens_per_second": 10147.644
    },
    {
      "epoch": 1.4082422586520948,
      "grad_norm": 0.7095161602575195,
      "learning_rate": 2.010383347901723e-06,
      "loss": 0.2596,
      "num_input_tokens_seen": 356771056,
      "step": 6185,
      "train_runtime": 35161.2921,
      "train_tokens_per_second": 10146.699
    },
    {
      "epoch": 1.4093806921675773,
      "grad_norm": 0.6417671512023683,
      "learning_rate": 2.0032212547107025e-06,
      "loss": 0.2515,
      "num_input_tokens_seen": 357071504,
      "step": 6190,
      "train_runtime": 35192.6104,
      "train_tokens_per_second": 10146.207
    },
    {
      "epoch": 1.41051912568306,
      "grad_norm": 0.6520304756983221,
      "learning_rate": 1.9960687447000884e-06,
      "loss": 0.2416,
      "num_input_tokens_seen": 357348944,
      "step": 6195,
      "train_runtime": 35222.98,
      "train_tokens_per_second": 10145.335
    },
    {
      "epoch": 1.4116575591985427,
      "grad_norm": 0.6222983477132735,
      "learning_rate": 1.9889258407423717e-06,
      "loss": 0.2537,
      "num_input_tokens_seen": 357624080,
      "step": 6200,
      "train_runtime": 35250.4766,
      "train_tokens_per_second": 10145.227
    },
    {
      "epoch": 1.4127959927140255,
      "grad_norm": 0.6282705221427642,
      "learning_rate": 1.981792565679321e-06,
      "loss": 0.2608,
      "num_input_tokens_seen": 357901392,
      "step": 6205,
      "train_runtime": 35272.984,
      "train_tokens_per_second": 10146.615
    },
    {
      "epoch": 1.4139344262295082,
      "grad_norm": 0.625618592543389,
      "learning_rate": 1.9746689423219184e-06,
      "loss": 0.2565,
      "num_input_tokens_seen": 358190344,
      "step": 6210,
      "train_runtime": 35302.4754,
      "train_tokens_per_second": 10146.324
    },
    {
      "epoch": 1.4150728597449909,
      "grad_norm": 0.586076187308115,
      "learning_rate": 1.967554993450282e-06,
      "loss": 0.2408,
      "num_input_tokens_seen": 358455480,
      "step": 6215,
      "train_runtime": 35328.5435,
      "train_tokens_per_second": 10146.342
    },
    {
      "epoch": 1.4162112932604736,
      "grad_norm": 0.7114194287009781,
      "learning_rate": 1.96045074181359e-06,
      "loss": 0.2647,
      "num_input_tokens_seen": 358725952,
      "step": 6220,
      "train_runtime": 35362.727,
      "train_tokens_per_second": 10144.182
    },
    {
      "epoch": 1.4173497267759563,
      "grad_norm": 0.6063040647607151,
      "learning_rate": 1.95335621013001e-06,
      "loss": 0.2566,
      "num_input_tokens_seen": 359019480,
      "step": 6225,
      "train_runtime": 35395.9175,
      "train_tokens_per_second": 10142.963
    },
    {
      "epoch": 1.418488160291439,
      "grad_norm": 0.5961600293870616,
      "learning_rate": 1.946271421086628e-06,
      "loss": 0.2279,
      "num_input_tokens_seen": 359280352,
      "step": 6230,
      "train_runtime": 35421.5039,
      "train_tokens_per_second": 10143.001
    },
    {
      "epoch": 1.4196265938069217,
      "grad_norm": 0.6531560561474384,
      "learning_rate": 1.939196397339375e-06,
      "loss": 0.2605,
      "num_input_tokens_seen": 359559944,
      "step": 6235,
      "train_runtime": 35455.4831,
      "train_tokens_per_second": 10141.166
    },
    {
      "epoch": 1.4207650273224044,
      "grad_norm": 0.6354419735890257,
      "learning_rate": 1.9321311615129546e-06,
      "loss": 0.228,
      "num_input_tokens_seen": 359841880,
      "step": 6240,
      "train_runtime": 35485.6684,
      "train_tokens_per_second": 10140.485
    },
    {
      "epoch": 1.421903460837887,
      "grad_norm": 0.7183978899580762,
      "learning_rate": 1.925075736200769e-06,
      "loss": 0.2858,
      "num_input_tokens_seen": 360131264,
      "step": 6245,
      "train_runtime": 35518.5696,
      "train_tokens_per_second": 10139.239
    },
    {
      "epoch": 1.4230418943533698,
      "grad_norm": 0.6690412223868244,
      "learning_rate": 1.918030143964849e-06,
      "loss": 0.2587,
      "num_input_tokens_seen": 360405168,
      "step": 6250,
      "train_runtime": 35548.0265,
      "train_tokens_per_second": 10138.542
    },
    {
      "epoch": 1.4241803278688525,
      "grad_norm": 0.5779687387961643,
      "learning_rate": 1.910994407335779e-06,
      "loss": 0.2601,
      "num_input_tokens_seen": 360694872,
      "step": 6255,
      "train_runtime": 35579.1716,
      "train_tokens_per_second": 10137.81
    },
    {
      "epoch": 1.4253187613843352,
      "grad_norm": 0.5945522713777023,
      "learning_rate": 1.9039685488126313e-06,
      "loss": 0.2183,
      "num_input_tokens_seen": 360991376,
      "step": 6260,
      "train_runtime": 35607.7617,
      "train_tokens_per_second": 10137.997
    },
    {
      "epoch": 1.426457194899818,
      "grad_norm": 0.624225308022321,
      "learning_rate": 1.896952590862886e-06,
      "loss": 0.2758,
      "num_input_tokens_seen": 361298608,
      "step": 6265,
      "train_runtime": 35634.8909,
      "train_tokens_per_second": 10138.9
    },
    {
      "epoch": 1.4275956284153004,
      "grad_norm": 0.6664286905631913,
      "learning_rate": 1.8899465559223635e-06,
      "loss": 0.2538,
      "num_input_tokens_seen": 361573248,
      "step": 6270,
      "train_runtime": 35665.2118,
      "train_tokens_per_second": 10137.981
    },
    {
      "epoch": 1.4287340619307831,
      "grad_norm": 0.5787358547052777,
      "learning_rate": 1.882950466395153e-06,
      "loss": 0.2447,
      "num_input_tokens_seen": 361861880,
      "step": 6275,
      "train_runtime": 35698.9538,
      "train_tokens_per_second": 10136.484
    },
    {
      "epoch": 1.4298724954462658,
      "grad_norm": 0.7314094082875362,
      "learning_rate": 1.8759643446535386e-06,
      "loss": 0.2387,
      "num_input_tokens_seen": 362128472,
      "step": 6280,
      "train_runtime": 35725.1047,
      "train_tokens_per_second": 10136.527
    },
    {
      "epoch": 1.4310109289617485,
      "grad_norm": 0.6844379395462504,
      "learning_rate": 1.8689882130379305e-06,
      "loss": 0.2538,
      "num_input_tokens_seen": 362394656,
      "step": 6285,
      "train_runtime": 35753.3583,
      "train_tokens_per_second": 10135.961
    },
    {
      "epoch": 1.4321493624772312,
      "grad_norm": 0.6818100620155881,
      "learning_rate": 1.8620220938567924e-06,
      "loss": 0.2516,
      "num_input_tokens_seen": 362684056,
      "step": 6290,
      "train_runtime": 35779.7405,
      "train_tokens_per_second": 10136.576
    },
    {
      "epoch": 1.433287795992714,
      "grad_norm": 0.601963440630329,
      "learning_rate": 1.855066009386568e-06,
      "loss": 0.2486,
      "num_input_tokens_seen": 362978176,
      "step": 6295,
      "train_runtime": 35814.2983,
      "train_tokens_per_second": 10135.007
    },
    {
      "epoch": 1.4344262295081966,
      "grad_norm": 0.5579641259279018,
      "learning_rate": 1.8481199818716116e-06,
      "loss": 0.2713,
      "num_input_tokens_seen": 363273048,
      "step": 6300,
      "train_runtime": 35844.8479,
      "train_tokens_per_second": 10134.596
    },
    {
      "epoch": 1.4355646630236794,
      "grad_norm": 0.5722389008267739,
      "learning_rate": 1.8411840335241243e-06,
      "loss": 0.2426,
      "num_input_tokens_seen": 363565040,
      "step": 6305,
      "train_runtime": 35870.928,
      "train_tokens_per_second": 10135.368
    },
    {
      "epoch": 1.436703096539162,
      "grad_norm": 0.657641305069205,
      "learning_rate": 1.834258186524065e-06,
      "loss": 0.2471,
      "num_input_tokens_seen": 363844400,
      "step": 6310,
      "train_runtime": 35895.8436,
      "train_tokens_per_second": 10136.115
    },
    {
      "epoch": 1.4378415300546448,
      "grad_norm": 0.5352283708323896,
      "learning_rate": 1.8273424630190956e-06,
      "loss": 0.2519,
      "num_input_tokens_seen": 364154280,
      "step": 6315,
      "train_runtime": 35930.1687,
      "train_tokens_per_second": 10135.056
    },
    {
      "epoch": 1.4389799635701275,
      "grad_norm": 0.5795798858572819,
      "learning_rate": 1.8204368851245052e-06,
      "loss": 0.2475,
      "num_input_tokens_seen": 364422024,
      "step": 6320,
      "train_runtime": 35960.6458,
      "train_tokens_per_second": 10133.912
    },
    {
      "epoch": 1.4401183970856102,
      "grad_norm": 0.6525486646732753,
      "learning_rate": 1.8135414749231384e-06,
      "loss": 0.2633,
      "num_input_tokens_seen": 364712688,
      "step": 6325,
      "train_runtime": 35992.9927,
      "train_tokens_per_second": 10132.88
    },
    {
      "epoch": 1.4412568306010929,
      "grad_norm": 0.6379675720301663,
      "learning_rate": 1.8066562544653249e-06,
      "loss": 0.2495,
      "num_input_tokens_seen": 365009136,
      "step": 6330,
      "train_runtime": 36018.5526,
      "train_tokens_per_second": 10133.92
    },
    {
      "epoch": 1.4423952641165756,
      "grad_norm": 0.6381778877958568,
      "learning_rate": 1.7997812457688096e-06,
      "loss": 0.2391,
      "num_input_tokens_seen": 365297576,
      "step": 6335,
      "train_runtime": 36046.749,
      "train_tokens_per_second": 10133.995
    },
    {
      "epoch": 1.4435336976320583,
      "grad_norm": 0.7463224272145978,
      "learning_rate": 1.7929164708186803e-06,
      "loss": 0.2451,
      "num_input_tokens_seen": 365571872,
      "step": 6340,
      "train_runtime": 36080.71,
      "train_tokens_per_second": 10132.059
    },
    {
      "epoch": 1.444672131147541,
      "grad_norm": 0.6981834415137032,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 0.251,
      "num_input_tokens_seen": 365841016,
      "step": 6345,
      "train_runtime": 36108.7265,
      "train_tokens_per_second": 10131.651
    },
    {
      "epoch": 1.4458105646630237,
      "grad_norm": 0.6542135655632019,
      "learning_rate": 1.7792177099342456e-06,
      "loss": 0.2323,
      "num_input_tokens_seen": 366116944,
      "step": 6350,
      "train_runtime": 36140.0034,
      "train_tokens_per_second": 10130.518
    },
    {
      "epoch": 1.4469489981785064,
      "grad_norm": 0.7213897996080726,
      "learning_rate": 1.7723837678062083e-06,
      "loss": 0.2642,
      "num_input_tokens_seen": 366414072,
      "step": 6355,
      "train_runtime": 36172.0132,
      "train_tokens_per_second": 10129.767
    },
    {
      "epoch": 1.4480874316939891,
      "grad_norm": 0.6763851366948624,
      "learning_rate": 1.7655601470369576e-06,
      "loss": 0.2421,
      "num_input_tokens_seen": 366696944,
      "step": 6360,
      "train_runtime": 36195.3947,
      "train_tokens_per_second": 10131.039
    },
    {
      "epoch": 1.4492258652094718,
      "grad_norm": 0.6608326181460072,
      "learning_rate": 1.758746869447252e-06,
      "loss": 0.2677,
      "num_input_tokens_seen": 366982112,
      "step": 6365,
      "train_runtime": 36230.341,
      "train_tokens_per_second": 10129.138
    },
    {
      "epoch": 1.4503642987249545,
      "grad_norm": 0.6152377139118828,
      "learning_rate": 1.7519439568247764e-06,
      "loss": 0.2774,
      "num_input_tokens_seen": 367271336,
      "step": 6370,
      "train_runtime": 36260.3597,
      "train_tokens_per_second": 10128.728
    },
    {
      "epoch": 1.4515027322404372,
      "grad_norm": 0.5761026944646921,
      "learning_rate": 1.7451514309240698e-06,
      "loss": 0.2403,
      "num_input_tokens_seen": 367555336,
      "step": 6375,
      "train_runtime": 36285.971,
      "train_tokens_per_second": 10129.406
    },
    {
      "epoch": 1.45264116575592,
      "grad_norm": 0.7082496570930572,
      "learning_rate": 1.7383693134664548e-06,
      "loss": 0.2739,
      "num_input_tokens_seen": 367848024,
      "step": 6380,
      "train_runtime": 36315.2797,
      "train_tokens_per_second": 10129.291
    },
    {
      "epoch": 1.4537795992714027,
      "grad_norm": 0.6892545592697764,
      "learning_rate": 1.7315976261399698e-06,
      "loss": 0.2539,
      "num_input_tokens_seen": 368148264,
      "step": 6385,
      "train_runtime": 36341.1302,
      "train_tokens_per_second": 10130.347
    },
    {
      "epoch": 1.4549180327868854,
      "grad_norm": 0.6413003168342092,
      "learning_rate": 1.7248363905993032e-06,
      "loss": 0.2332,
      "num_input_tokens_seen": 368453560,
      "step": 6390,
      "train_runtime": 36371.6668,
      "train_tokens_per_second": 10130.236
    },
    {
      "epoch": 1.456056466302368,
      "grad_norm": 0.6282453205265823,
      "learning_rate": 1.718085628465716e-06,
      "loss": 0.2792,
      "num_input_tokens_seen": 368722896,
      "step": 6395,
      "train_runtime": 36402.873,
      "train_tokens_per_second": 10128.95
    },
    {
      "epoch": 1.4571948998178508,
      "grad_norm": 0.552678202911407,
      "learning_rate": 1.7113453613269788e-06,
      "loss": 0.2294,
      "num_input_tokens_seen": 368995696,
      "step": 6400,
      "train_runtime": 36427.3304,
      "train_tokens_per_second": 10129.639
    },
    {
      "epoch": 1.4583333333333333,
      "grad_norm": 0.5899907250164695,
      "learning_rate": 1.7046156107373007e-06,
      "loss": 0.2463,
      "num_input_tokens_seen": 369285128,
      "step": 6405,
      "train_runtime": 36460.0937,
      "train_tokens_per_second": 10128.474
    },
    {
      "epoch": 1.459471766848816,
      "grad_norm": 0.6545305562463091,
      "learning_rate": 1.697896398217262e-06,
      "loss": 0.2498,
      "num_input_tokens_seen": 369580056,
      "step": 6410,
      "train_runtime": 36484.1283,
      "train_tokens_per_second": 10129.886
    },
    {
      "epoch": 1.4606102003642987,
      "grad_norm": 0.5903667380218601,
      "learning_rate": 1.69118774525374e-06,
      "loss": 0.2562,
      "num_input_tokens_seen": 369877432,
      "step": 6415,
      "train_runtime": 36514.5318,
      "train_tokens_per_second": 10129.595
    },
    {
      "epoch": 1.4617486338797814,
      "grad_norm": 0.6320893757818454,
      "learning_rate": 1.6844896732998472e-06,
      "loss": 0.2206,
      "num_input_tokens_seen": 370162312,
      "step": 6420,
      "train_runtime": 36545.6222,
      "train_tokens_per_second": 10128.773
    },
    {
      "epoch": 1.462887067395264,
      "grad_norm": 0.6044086313005436,
      "learning_rate": 1.6778022037748627e-06,
      "loss": 0.2349,
      "num_input_tokens_seen": 370458552,
      "step": 6425,
      "train_runtime": 36577.6037,
      "train_tokens_per_second": 10128.016
    },
    {
      "epoch": 1.4640255009107468,
      "grad_norm": 0.49190850488106713,
      "learning_rate": 1.6711253580641573e-06,
      "loss": 0.2338,
      "num_input_tokens_seen": 370761440,
      "step": 6430,
      "train_runtime": 36604.4425,
      "train_tokens_per_second": 10128.865
    },
    {
      "epoch": 1.4651639344262295,
      "grad_norm": 0.6437671149981549,
      "learning_rate": 1.6644591575191288e-06,
      "loss": 0.2688,
      "num_input_tokens_seen": 371009248,
      "step": 6435,
      "train_runtime": 36632.1673,
      "train_tokens_per_second": 10127.963
    },
    {
      "epoch": 1.4663023679417122,
      "grad_norm": 0.5440024319143629,
      "learning_rate": 1.6578036234571337e-06,
      "loss": 0.2681,
      "num_input_tokens_seen": 371320880,
      "step": 6440,
      "train_runtime": 36665.2057,
      "train_tokens_per_second": 10127.337
    },
    {
      "epoch": 1.467440801457195,
      "grad_norm": 0.5696996615618398,
      "learning_rate": 1.6511587771614208e-06,
      "loss": 0.2574,
      "num_input_tokens_seen": 371605800,
      "step": 6445,
      "train_runtime": 36691.1291,
      "train_tokens_per_second": 10127.947
    },
    {
      "epoch": 1.4685792349726776,
      "grad_norm": 0.6054407070840849,
      "learning_rate": 1.6445246398810593e-06,
      "loss": 0.2568,
      "num_input_tokens_seen": 371882672,
      "step": 6450,
      "train_runtime": 36713.8679,
      "train_tokens_per_second": 10129.215
    },
    {
      "epoch": 1.4697176684881603,
      "grad_norm": 0.5918244365169727,
      "learning_rate": 1.637901232830873e-06,
      "loss": 0.2443,
      "num_input_tokens_seen": 372190728,
      "step": 6455,
      "train_runtime": 36742.7263,
      "train_tokens_per_second": 10129.644
    },
    {
      "epoch": 1.470856102003643,
      "grad_norm": 0.6628846729025601,
      "learning_rate": 1.6312885771913732e-06,
      "loss": 0.2694,
      "num_input_tokens_seen": 372472232,
      "step": 6460,
      "train_runtime": 36768.2512,
      "train_tokens_per_second": 10130.268
    },
    {
      "epoch": 1.4719945355191257,
      "grad_norm": 0.5976533763657561,
      "learning_rate": 1.6246866941086897e-06,
      "loss": 0.2346,
      "num_input_tokens_seen": 372771280,
      "step": 6465,
      "train_runtime": 36801.5537,
      "train_tokens_per_second": 10129.227
    },
    {
      "epoch": 1.4731329690346084,
      "grad_norm": 0.6391227763301437,
      "learning_rate": 1.6180956046945013e-06,
      "loss": 0.2277,
      "num_input_tokens_seen": 373046232,
      "step": 6470,
      "train_runtime": 36827.232,
      "train_tokens_per_second": 10129.63
    },
    {
      "epoch": 1.4742714025500911,
      "grad_norm": 2.316838596543321,
      "learning_rate": 1.6115153300259762e-06,
      "loss": 0.2649,
      "num_input_tokens_seen": 373327840,
      "step": 6475,
      "train_runtime": 36856.7036,
      "train_tokens_per_second": 10129.171
    },
    {
      "epoch": 1.4754098360655736,
      "grad_norm": 0.6631069137643082,
      "learning_rate": 1.6049458911456933e-06,
      "loss": 0.2574,
      "num_input_tokens_seen": 373624728,
      "step": 6480,
      "train_runtime": 36888.1517,
      "train_tokens_per_second": 10128.584
    },
    {
      "epoch": 1.4765482695810563,
      "grad_norm": 0.6997683356562581,
      "learning_rate": 1.5983873090615831e-06,
      "loss": 0.2481,
      "num_input_tokens_seen": 373901808,
      "step": 6485,
      "train_runtime": 36912.1846,
      "train_tokens_per_second": 10129.496
    },
    {
      "epoch": 1.477686703096539,
      "grad_norm": 0.6394739039946139,
      "learning_rate": 1.5918396047468572e-06,
      "loss": 0.2472,
      "num_input_tokens_seen": 374203328,
      "step": 6490,
      "train_runtime": 36937.7833,
      "train_tokens_per_second": 10130.638
    },
    {
      "epoch": 1.4788251366120218,
      "grad_norm": 0.6272455827630014,
      "learning_rate": 1.5853027991399423e-06,
      "loss": 0.2358,
      "num_input_tokens_seen": 374501984,
      "step": 6495,
      "train_runtime": 36968.2007,
      "train_tokens_per_second": 10130.382
    },
    {
      "epoch": 1.4799635701275045,
      "grad_norm": 0.6005441234038956,
      "learning_rate": 1.5787769131444126e-06,
      "loss": 0.2417,
      "num_input_tokens_seen": 374780288,
      "step": 6500,
      "train_runtime": 36994.7933,
      "train_tokens_per_second": 10130.623
    },
    {
      "epoch": 1.4811020036429872,
      "grad_norm": 0.599669832923083,
      "learning_rate": 1.572261967628923e-06,
      "loss": 0.2525,
      "num_input_tokens_seen": 375081992,
      "step": 6505,
      "train_runtime": 37023.0953,
      "train_tokens_per_second": 10131.027
    },
    {
      "epoch": 1.4822404371584699,
      "grad_norm": 0.6793827049847567,
      "learning_rate": 1.5657579834271435e-06,
      "loss": 0.2672,
      "num_input_tokens_seen": 375363216,
      "step": 6510,
      "train_runtime": 37052.5344,
      "train_tokens_per_second": 10130.568
    },
    {
      "epoch": 1.4833788706739526,
      "grad_norm": 0.6255925679415474,
      "learning_rate": 1.5592649813376904e-06,
      "loss": 0.2496,
      "num_input_tokens_seen": 375655264,
      "step": 6515,
      "train_runtime": 37084.0554,
      "train_tokens_per_second": 10129.832
    },
    {
      "epoch": 1.4845173041894353,
      "grad_norm": 0.654418977211696,
      "learning_rate": 1.5527829821240658e-06,
      "loss": 0.2665,
      "num_input_tokens_seen": 375951712,
      "step": 6520,
      "train_runtime": 37117.231,
      "train_tokens_per_second": 10128.765
    },
    {
      "epoch": 1.485655737704918,
      "grad_norm": 0.7121475547975372,
      "learning_rate": 1.546312006514578e-06,
      "loss": 0.2699,
      "num_input_tokens_seen": 376213136,
      "step": 6525,
      "train_runtime": 37140.5151,
      "train_tokens_per_second": 10129.454
    },
    {
      "epoch": 1.4867941712204007,
      "grad_norm": 0.7063728358224849,
      "learning_rate": 1.5398520752022905e-06,
      "loss": 0.2694,
      "num_input_tokens_seen": 376507120,
      "step": 6530,
      "train_runtime": 37168.9903,
      "train_tokens_per_second": 10129.603
    },
    {
      "epoch": 1.4879326047358834,
      "grad_norm": 0.5686047797478687,
      "learning_rate": 1.533403208844947e-06,
      "loss": 0.2364,
      "num_input_tokens_seen": 376805128,
      "step": 6535,
      "train_runtime": 37197.4395,
      "train_tokens_per_second": 10129.867
    },
    {
      "epoch": 1.489071038251366,
      "grad_norm": 0.5723970323865668,
      "learning_rate": 1.5269654280649066e-06,
      "loss": 0.2592,
      "num_input_tokens_seen": 377106160,
      "step": 6540,
      "train_runtime": 37220.9444,
      "train_tokens_per_second": 10131.558
    },
    {
      "epoch": 1.4902094717668488,
      "grad_norm": 0.6968338031933203,
      "learning_rate": 1.5205387534490807e-06,
      "loss": 0.2597,
      "num_input_tokens_seen": 377405968,
      "step": 6545,
      "train_runtime": 37253.8479,
      "train_tokens_per_second": 10130.657
    },
    {
      "epoch": 1.4913479052823315,
      "grad_norm": 0.5525660425869732,
      "learning_rate": 1.5141232055488609e-06,
      "loss": 0.2364,
      "num_input_tokens_seen": 377681600,
      "step": 6550,
      "train_runtime": 37284.0437,
      "train_tokens_per_second": 10129.845
    },
    {
      "epoch": 1.4924863387978142,
      "grad_norm": 0.7471242980026919,
      "learning_rate": 1.5077188048800639e-06,
      "loss": 0.2635,
      "num_input_tokens_seen": 377928392,
      "step": 6555,
      "train_runtime": 37311.1334,
      "train_tokens_per_second": 10129.105
    },
    {
      "epoch": 1.493624772313297,
      "grad_norm": 0.6056891012270484,
      "learning_rate": 1.5013255719228547e-06,
      "loss": 0.2204,
      "num_input_tokens_seen": 378211008,
      "step": 6560,
      "train_runtime": 37337.8392,
      "train_tokens_per_second": 10129.429
    },
    {
      "epoch": 1.4947632058287796,
      "grad_norm": 0.5268515982190045,
      "learning_rate": 1.4949435271216878e-06,
      "loss": 0.2442,
      "num_input_tokens_seen": 378471760,
      "step": 6565,
      "train_runtime": 37367.605,
      "train_tokens_per_second": 10128.339
    },
    {
      "epoch": 1.4959016393442623,
      "grad_norm": 0.5834641534407602,
      "learning_rate": 1.4885726908852378e-06,
      "loss": 0.2516,
      "num_input_tokens_seen": 378775008,
      "step": 6570,
      "train_runtime": 37398.0049,
      "train_tokens_per_second": 10128.214
    },
    {
      "epoch": 1.497040072859745,
      "grad_norm": 0.5408933693453757,
      "learning_rate": 1.4822130835863385e-06,
      "loss": 0.2234,
      "num_input_tokens_seen": 379080928,
      "step": 6575,
      "train_runtime": 37425.5127,
      "train_tokens_per_second": 10128.944
    },
    {
      "epoch": 1.4981785063752278,
      "grad_norm": 0.6735382199748232,
      "learning_rate": 1.4758647255619153e-06,
      "loss": 0.2519,
      "num_input_tokens_seen": 379377288,
      "step": 6580,
      "train_runtime": 37452.8898,
      "train_tokens_per_second": 10129.453
    },
    {
      "epoch": 1.4993169398907105,
      "grad_norm": 0.5533498721125989,
      "learning_rate": 1.4695276371129187e-06,
      "loss": 0.2519,
      "num_input_tokens_seen": 379666200,
      "step": 6585,
      "train_runtime": 37479.9432,
      "train_tokens_per_second": 10129.85
    },
    {
      "epoch": 1.5004553734061932,
      "grad_norm": 0.5534289738888422,
      "learning_rate": 1.4632018385042623e-06,
      "loss": 0.2397,
      "num_input_tokens_seen": 379968040,
      "step": 6590,
      "train_runtime": 37507.8203,
      "train_tokens_per_second": 10130.368
    },
    {
      "epoch": 1.5015938069216759,
      "grad_norm": 0.5603669334183689,
      "learning_rate": 1.4568873499647573e-06,
      "loss": 0.2358,
      "num_input_tokens_seen": 380259736,
      "step": 6595,
      "train_runtime": 37535.0622,
      "train_tokens_per_second": 10130.787
    },
    {
      "epoch": 1.5027322404371586,
      "grad_norm": 0.662179622609062,
      "learning_rate": 1.4505841916870444e-06,
      "loss": 0.2761,
      "num_input_tokens_seen": 380548584,
      "step": 6600,
      "train_runtime": 37562.9776,
      "train_tokens_per_second": 10130.948
    },
    {
      "epoch": 1.5038706739526413,
      "grad_norm": 0.5980208795834857,
      "learning_rate": 1.444292383827538e-06,
      "loss": 0.2492,
      "num_input_tokens_seen": 380835600,
      "step": 6605,
      "train_runtime": 37584.3931,
      "train_tokens_per_second": 10132.812
    },
    {
      "epoch": 1.505009107468124,
      "grad_norm": 0.6212088968561429,
      "learning_rate": 1.4380119465063497e-06,
      "loss": 0.2363,
      "num_input_tokens_seen": 381149248,
      "step": 6610,
      "train_runtime": 37614.6607,
      "train_tokens_per_second": 10132.997
    },
    {
      "epoch": 1.5061475409836067,
      "grad_norm": 0.7321024828048336,
      "learning_rate": 1.4317428998072324e-06,
      "loss": 0.231,
      "num_input_tokens_seen": 381454728,
      "step": 6615,
      "train_runtime": 37635.7986,
      "train_tokens_per_second": 10135.423
    },
    {
      "epoch": 1.5072859744990894,
      "grad_norm": 0.5914691249627513,
      "learning_rate": 1.4254852637775142e-06,
      "loss": 0.2385,
      "num_input_tokens_seen": 381756448,
      "step": 6620,
      "train_runtime": 37666.2285,
      "train_tokens_per_second": 10135.245
    },
    {
      "epoch": 1.508424408014572,
      "grad_norm": 0.5640525639054546,
      "learning_rate": 1.4192390584280347e-06,
      "loss": 0.2459,
      "num_input_tokens_seen": 382042568,
      "step": 6625,
      "train_runtime": 37693.4096,
      "train_tokens_per_second": 10135.527
    },
    {
      "epoch": 1.5095628415300546,
      "grad_norm": 0.6336181133289996,
      "learning_rate": 1.4130043037330764e-06,
      "loss": 0.2417,
      "num_input_tokens_seen": 382338440,
      "step": 6630,
      "train_runtime": 37720.8747,
      "train_tokens_per_second": 10135.991
    },
    {
      "epoch": 1.5107012750455373,
      "grad_norm": 0.5509922100201632,
      "learning_rate": 1.4067810196303061e-06,
      "loss": 0.2442,
      "num_input_tokens_seen": 382638152,
      "step": 6635,
      "train_runtime": 37740.2879,
      "train_tokens_per_second": 10138.718
    },
    {
      "epoch": 1.51183970856102,
      "grad_norm": 0.7428030605182623,
      "learning_rate": 1.400569226020715e-06,
      "loss": 0.2555,
      "num_input_tokens_seen": 382921952,
      "step": 6640,
      "train_runtime": 37775.3783,
      "train_tokens_per_second": 10136.813
    },
    {
      "epoch": 1.5129781420765027,
      "grad_norm": 0.6023697568975682,
      "learning_rate": 1.394368942768542e-06,
      "loss": 0.2432,
      "num_input_tokens_seen": 383206536,
      "step": 6645,
      "train_runtime": 37804.6534,
      "train_tokens_per_second": 10136.491
    },
    {
      "epoch": 1.5141165755919854,
      "grad_norm": 0.5447915000329002,
      "learning_rate": 1.3881801897012225e-06,
      "loss": 0.2272,
      "num_input_tokens_seen": 383511176,
      "step": 6650,
      "train_runtime": 37838.6964,
      "train_tokens_per_second": 10135.423
    },
    {
      "epoch": 1.5152550091074681,
      "grad_norm": 0.6318359591032876,
      "learning_rate": 1.3820029866093188e-06,
      "loss": 0.2714,
      "num_input_tokens_seen": 383818168,
      "step": 6655,
      "train_runtime": 37862.1141,
      "train_tokens_per_second": 10137.262
    },
    {
      "epoch": 1.5163934426229508,
      "grad_norm": 0.5608487910094851,
      "learning_rate": 1.3758373532464585e-06,
      "loss": 0.2374,
      "num_input_tokens_seen": 384087680,
      "step": 6660,
      "train_runtime": 37888.2552,
      "train_tokens_per_second": 10137.381
    },
    {
      "epoch": 1.5175318761384335,
      "grad_norm": 0.5848910711844347,
      "learning_rate": 1.369683309329271e-06,
      "loss": 0.2364,
      "num_input_tokens_seen": 384372016,
      "step": 6665,
      "train_runtime": 37921.3246,
      "train_tokens_per_second": 10136.039
    },
    {
      "epoch": 1.5186703096539163,
      "grad_norm": 0.6133440213614728,
      "learning_rate": 1.3635408745373253e-06,
      "loss": 0.2408,
      "num_input_tokens_seen": 384652640,
      "step": 6670,
      "train_runtime": 37954.385,
      "train_tokens_per_second": 10134.603
    },
    {
      "epoch": 1.519808743169399,
      "grad_norm": 0.5573687973178996,
      "learning_rate": 1.3574100685130659e-06,
      "loss": 0.2712,
      "num_input_tokens_seen": 384927696,
      "step": 6675,
      "train_runtime": 37979.3977,
      "train_tokens_per_second": 10135.171
    },
    {
      "epoch": 1.5209471766848814,
      "grad_norm": 0.6796241134886291,
      "learning_rate": 1.3512909108617494e-06,
      "loss": 0.244,
      "num_input_tokens_seen": 385203304,
      "step": 6680,
      "train_runtime": 38001.5295,
      "train_tokens_per_second": 10136.521
    },
    {
      "epoch": 1.5220856102003641,
      "grad_norm": 0.6341146093859592,
      "learning_rate": 1.3451834211513876e-06,
      "loss": 0.2664,
      "num_input_tokens_seen": 385497016,
      "step": 6685,
      "train_runtime": 38029.4282,
      "train_tokens_per_second": 10136.808
    },
    {
      "epoch": 1.5232240437158469,
      "grad_norm": 0.610439688094552,
      "learning_rate": 1.339087618912674e-06,
      "loss": 0.2348,
      "num_input_tokens_seen": 385766368,
      "step": 6690,
      "train_runtime": 38056.8859,
      "train_tokens_per_second": 10136.572
    },
    {
      "epoch": 1.5243624772313296,
      "grad_norm": 0.5909230408659082,
      "learning_rate": 1.3330035236389322e-06,
      "loss": 0.2418,
      "num_input_tokens_seen": 386068064,
      "step": 6695,
      "train_runtime": 38082.2587,
      "train_tokens_per_second": 10137.741
    },
    {
      "epoch": 1.5255009107468123,
      "grad_norm": 0.5741198907132,
      "learning_rate": 1.3269311547860452e-06,
      "loss": 0.2576,
      "num_input_tokens_seen": 386347048,
      "step": 6700,
      "train_runtime": 38109.23,
      "train_tokens_per_second": 10137.886
    },
    {
      "epoch": 1.526639344262295,
      "grad_norm": 0.6467838796434272,
      "learning_rate": 1.3208705317724007e-06,
      "loss": 0.2424,
      "num_input_tokens_seen": 386666160,
      "step": 6705,
      "train_runtime": 38139.2816,
      "train_tokens_per_second": 10138.265
    },
    {
      "epoch": 1.5277777777777777,
      "grad_norm": 0.5416390072105584,
      "learning_rate": 1.3148216739788233e-06,
      "loss": 0.2474,
      "num_input_tokens_seen": 386957664,
      "step": 6710,
      "train_runtime": 38167.5994,
      "train_tokens_per_second": 10138.381
    },
    {
      "epoch": 1.5289162112932604,
      "grad_norm": 0.6338062802730171,
      "learning_rate": 1.3087846007485134e-06,
      "loss": 0.2442,
      "num_input_tokens_seen": 387251872,
      "step": 6715,
      "train_runtime": 38197.2036,
      "train_tokens_per_second": 10138.226
    },
    {
      "epoch": 1.530054644808743,
      "grad_norm": 0.6825567247899664,
      "learning_rate": 1.3027593313869886e-06,
      "loss": 0.2767,
      "num_input_tokens_seen": 387551136,
      "step": 6720,
      "train_runtime": 38230.2379,
      "train_tokens_per_second": 10137.293
    },
    {
      "epoch": 1.5311930783242258,
      "grad_norm": 0.6945096511061462,
      "learning_rate": 1.2967458851620185e-06,
      "loss": 0.2843,
      "num_input_tokens_seen": 387842664,
      "step": 6725,
      "train_runtime": 38256.2228,
      "train_tokens_per_second": 10138.028
    },
    {
      "epoch": 1.5323315118397085,
      "grad_norm": 0.6216503659922062,
      "learning_rate": 1.2907442813035643e-06,
      "loss": 0.2378,
      "num_input_tokens_seen": 388153160,
      "step": 6730,
      "train_runtime": 38280.9186,
      "train_tokens_per_second": 10139.599
    },
    {
      "epoch": 1.5334699453551912,
      "grad_norm": 0.5266351976889567,
      "learning_rate": 1.2847545390037186e-06,
      "loss": 0.2458,
      "num_input_tokens_seen": 388457400,
      "step": 6735,
      "train_runtime": 38306.3575,
      "train_tokens_per_second": 10140.808
    },
    {
      "epoch": 1.534608378870674,
      "grad_norm": 0.6526667365119126,
      "learning_rate": 1.278776677416641e-06,
      "loss": 0.2598,
      "num_input_tokens_seen": 388759072,
      "step": 6740,
      "train_runtime": 38330.2906,
      "train_tokens_per_second": 10142.346
    },
    {
      "epoch": 1.5357468123861566,
      "grad_norm": 0.662315333578438,
      "learning_rate": 1.2728107156585e-06,
      "loss": 0.253,
      "num_input_tokens_seen": 389056544,
      "step": 6745,
      "train_runtime": 38360.6223,
      "train_tokens_per_second": 10142.081
    },
    {
      "epoch": 1.5368852459016393,
      "grad_norm": 0.5683380187422552,
      "learning_rate": 1.26685667280741e-06,
      "loss": 0.2628,
      "num_input_tokens_seen": 389343824,
      "step": 6750,
      "train_runtime": 38386.6203,
      "train_tokens_per_second": 10142.696
    },
    {
      "epoch": 1.538023679417122,
      "grad_norm": 0.6042763677416196,
      "learning_rate": 1.2609145679033707e-06,
      "loss": 0.2483,
      "num_input_tokens_seen": 389653288,
      "step": 6755,
      "train_runtime": 38408.1894,
      "train_tokens_per_second": 10145.057
    },
    {
      "epoch": 1.5391621129326047,
      "grad_norm": 0.6007046232430606,
      "learning_rate": 1.2549844199482063e-06,
      "loss": 0.2476,
      "num_input_tokens_seen": 389954232,
      "step": 6760,
      "train_runtime": 38437.0263,
      "train_tokens_per_second": 10145.276
    },
    {
      "epoch": 1.5403005464480874,
      "grad_norm": 0.5617261281250675,
      "learning_rate": 1.2490662479055037e-06,
      "loss": 0.2368,
      "num_input_tokens_seen": 390234408,
      "step": 6765,
      "train_runtime": 38464.9529,
      "train_tokens_per_second": 10145.194
    },
    {
      "epoch": 1.5414389799635702,
      "grad_norm": 0.716302243313803,
      "learning_rate": 1.243160070700557e-06,
      "loss": 0.2494,
      "num_input_tokens_seen": 390506400,
      "step": 6770,
      "train_runtime": 38486.574,
      "train_tokens_per_second": 10146.562
    },
    {
      "epoch": 1.5425774134790529,
      "grad_norm": 0.5826338074076364,
      "learning_rate": 1.2372659072202985e-06,
      "loss": 0.2253,
      "num_input_tokens_seen": 390807088,
      "step": 6775,
      "train_runtime": 38518.9068,
      "train_tokens_per_second": 10145.851
    },
    {
      "epoch": 1.5437158469945356,
      "grad_norm": 0.7275797787122177,
      "learning_rate": 1.2313837763132442e-06,
      "loss": 0.2374,
      "num_input_tokens_seen": 391093904,
      "step": 6780,
      "train_runtime": 38543.8914,
      "train_tokens_per_second": 10146.716
    },
    {
      "epoch": 1.5448542805100183,
      "grad_norm": 0.5856564021474205,
      "learning_rate": 1.2255136967894316e-06,
      "loss": 0.2482,
      "num_input_tokens_seen": 391389896,
      "step": 6785,
      "train_runtime": 38567.3833,
      "train_tokens_per_second": 10148.21
    },
    {
      "epoch": 1.545992714025501,
      "grad_norm": 0.6172201069942818,
      "learning_rate": 1.2196556874203602e-06,
      "loss": 0.2464,
      "num_input_tokens_seen": 391674776,
      "step": 6790,
      "train_runtime": 38594.0316,
      "train_tokens_per_second": 10148.584
    },
    {
      "epoch": 1.5471311475409837,
      "grad_norm": 0.5672123485983878,
      "learning_rate": 1.213809766938932e-06,
      "loss": 0.2381,
      "num_input_tokens_seen": 391972032,
      "step": 6795,
      "train_runtime": 38621.9296,
      "train_tokens_per_second": 10148.95
    },
    {
      "epoch": 1.5482695810564664,
      "grad_norm": 0.6148639367123417,
      "learning_rate": 1.20797595403939e-06,
      "loss": 0.2518,
      "num_input_tokens_seen": 392256088,
      "step": 6800,
      "train_runtime": 38649.1318,
      "train_tokens_per_second": 10149.157
    },
    {
      "epoch": 1.549408014571949,
      "grad_norm": 0.5993432217128827,
      "learning_rate": 1.2021542673772584e-06,
      "loss": 0.249,
      "num_input_tokens_seen": 392541384,
      "step": 6805,
      "train_runtime": 38680.6601,
      "train_tokens_per_second": 10148.26
    },
    {
      "epoch": 1.5505464480874318,
      "grad_norm": 0.6399726448883781,
      "learning_rate": 1.1963447255692851e-06,
      "loss": 0.2525,
      "num_input_tokens_seen": 392831312,
      "step": 6810,
      "train_runtime": 38707.7036,
      "train_tokens_per_second": 10148.66
    },
    {
      "epoch": 1.5516848816029145,
      "grad_norm": 0.6109603068971353,
      "learning_rate": 1.1905473471933827e-06,
      "loss": 0.2521,
      "num_input_tokens_seen": 393121072,
      "step": 6815,
      "train_runtime": 38743.136,
      "train_tokens_per_second": 10146.857
    },
    {
      "epoch": 1.5528233151183972,
      "grad_norm": 0.6220442227305935,
      "learning_rate": 1.184762150788563e-06,
      "loss": 0.219,
      "num_input_tokens_seen": 393398808,
      "step": 6820,
      "train_runtime": 38766.0423,
      "train_tokens_per_second": 10148.026
    },
    {
      "epoch": 1.55396174863388,
      "grad_norm": 0.6630597957988684,
      "learning_rate": 1.1789891548548848e-06,
      "loss": 0.2546,
      "num_input_tokens_seen": 393689512,
      "step": 6825,
      "train_runtime": 38793.4864,
      "train_tokens_per_second": 10148.341
    },
    {
      "epoch": 1.5551001821493626,
      "grad_norm": 0.6760074310166717,
      "learning_rate": 1.173228377853391e-06,
      "loss": 0.2477,
      "num_input_tokens_seen": 393976568,
      "step": 6830,
      "train_runtime": 38814.4588,
      "train_tokens_per_second": 10150.253
    },
    {
      "epoch": 1.5562386156648453,
      "grad_norm": 0.6544932310541774,
      "learning_rate": 1.167479838206052e-06,
      "loss": 0.2552,
      "num_input_tokens_seen": 394262864,
      "step": 6835,
      "train_runtime": 38838.9903,
      "train_tokens_per_second": 10151.213
    },
    {
      "epoch": 1.5573770491803278,
      "grad_norm": 0.5735844807625747,
      "learning_rate": 1.1617435542957018e-06,
      "loss": 0.2653,
      "num_input_tokens_seen": 394571448,
      "step": 6840,
      "train_runtime": 38869.1412,
      "train_tokens_per_second": 10151.278
    },
    {
      "epoch": 1.5585154826958105,
      "grad_norm": 0.655545768380942,
      "learning_rate": 1.156019544465985e-06,
      "loss": 0.2635,
      "num_input_tokens_seen": 394864528,
      "step": 6845,
      "train_runtime": 38896.9576,
      "train_tokens_per_second": 10151.553
    },
    {
      "epoch": 1.5596539162112932,
      "grad_norm": 0.6156739497159732,
      "learning_rate": 1.1503078270212941e-06,
      "loss": 0.2615,
      "num_input_tokens_seen": 395147608,
      "step": 6850,
      "train_runtime": 38925.9206,
      "train_tokens_per_second": 10151.272
    },
    {
      "epoch": 1.560792349726776,
      "grad_norm": 0.6085442077776226,
      "learning_rate": 1.1446084202267188e-06,
      "loss": 0.234,
      "num_input_tokens_seen": 395441072,
      "step": 6855,
      "train_runtime": 38956.0354,
      "train_tokens_per_second": 10150.958
    },
    {
      "epoch": 1.5619307832422586,
      "grad_norm": 0.6576577009139875,
      "learning_rate": 1.138921342307973e-06,
      "loss": 0.2754,
      "num_input_tokens_seen": 395727104,
      "step": 6860,
      "train_runtime": 38987.6915,
      "train_tokens_per_second": 10150.052
    },
    {
      "epoch": 1.5630692167577414,
      "grad_norm": 0.6263466678638524,
      "learning_rate": 1.1332466114513512e-06,
      "loss": 0.2532,
      "num_input_tokens_seen": 395987224,
      "step": 6865,
      "train_runtime": 39016.3168,
      "train_tokens_per_second": 10149.272
    },
    {
      "epoch": 1.564207650273224,
      "grad_norm": 0.5538341763724142,
      "learning_rate": 1.1275842458036601e-06,
      "loss": 0.2553,
      "num_input_tokens_seen": 396270736,
      "step": 6870,
      "train_runtime": 39045.6614,
      "train_tokens_per_second": 10148.906
    },
    {
      "epoch": 1.5653460837887068,
      "grad_norm": 0.7089466741269095,
      "learning_rate": 1.1219342634721663e-06,
      "loss": 0.2573,
      "num_input_tokens_seen": 396557824,
      "step": 6875,
      "train_runtime": 39078.3654,
      "train_tokens_per_second": 10147.759
    },
    {
      "epoch": 1.5664845173041895,
      "grad_norm": 0.6566737737500602,
      "learning_rate": 1.1162966825245375e-06,
      "loss": 0.2488,
      "num_input_tokens_seen": 396859288,
      "step": 6880,
      "train_runtime": 39106.9672,
      "train_tokens_per_second": 10148.046
    },
    {
      "epoch": 1.5676229508196722,
      "grad_norm": 0.6652859586242653,
      "learning_rate": 1.110671520988783e-06,
      "loss": 0.2531,
      "num_input_tokens_seen": 397151248,
      "step": 6885,
      "train_runtime": 39139.596,
      "train_tokens_per_second": 10147.045
    },
    {
      "epoch": 1.5687613843351547,
      "grad_norm": 0.6604191428626092,
      "learning_rate": 1.1050587968531962e-06,
      "loss": 0.2523,
      "num_input_tokens_seen": 397443728,
      "step": 6890,
      "train_runtime": 39173.0613,
      "train_tokens_per_second": 10145.843
    },
    {
      "epoch": 1.5698998178506374,
      "grad_norm": 0.6134524945782629,
      "learning_rate": 1.0994585280662978e-06,
      "loss": 0.2527,
      "num_input_tokens_seen": 397711840,
      "step": 6895,
      "train_runtime": 39205.4909,
      "train_tokens_per_second": 10144.289
    },
    {
      "epoch": 1.57103825136612,
      "grad_norm": 0.6410628153078229,
      "learning_rate": 1.093870732536782e-06,
      "loss": 0.2363,
      "num_input_tokens_seen": 398008656,
      "step": 6900,
      "train_runtime": 39231.1935,
      "train_tokens_per_second": 10145.209
    },
    {
      "epoch": 1.5721766848816028,
      "grad_norm": 0.6370653456971064,
      "learning_rate": 1.0882954281334517e-06,
      "loss": 0.229,
      "num_input_tokens_seen": 398307744,
      "step": 6905,
      "train_runtime": 39255.8976,
      "train_tokens_per_second": 10146.443
    },
    {
      "epoch": 1.5733151183970855,
      "grad_norm": 0.690300662412627,
      "learning_rate": 1.0827326326851667e-06,
      "loss": 0.2352,
      "num_input_tokens_seen": 398612520,
      "step": 6910,
      "train_runtime": 39287.7899,
      "train_tokens_per_second": 10145.964
    },
    {
      "epoch": 1.5744535519125682,
      "grad_norm": 0.6463123023513266,
      "learning_rate": 1.0771823639807849e-06,
      "loss": 0.2467,
      "num_input_tokens_seen": 398888640,
      "step": 6915,
      "train_runtime": 39313.7602,
      "train_tokens_per_second": 10146.286
    },
    {
      "epoch": 1.575591985428051,
      "grad_norm": 0.6367182242617424,
      "learning_rate": 1.0716446397691066e-06,
      "loss": 0.2524,
      "num_input_tokens_seen": 399191672,
      "step": 6920,
      "train_runtime": 39344.3557,
      "train_tokens_per_second": 10146.098
    },
    {
      "epoch": 1.5767304189435336,
      "grad_norm": 0.6177497099679792,
      "learning_rate": 1.0661194777588162e-06,
      "loss": 0.2464,
      "num_input_tokens_seen": 399477480,
      "step": 6925,
      "train_runtime": 39373.5651,
      "train_tokens_per_second": 10145.83
    },
    {
      "epoch": 1.5778688524590163,
      "grad_norm": 0.6312710796461775,
      "learning_rate": 1.0606068956184274e-06,
      "loss": 0.2575,
      "num_input_tokens_seen": 399753104,
      "step": 6930,
      "train_runtime": 39402.5027,
      "train_tokens_per_second": 10145.373
    },
    {
      "epoch": 1.579007285974499,
      "grad_norm": 0.5589274860426858,
      "learning_rate": 1.055106910976224e-06,
      "loss": 0.2417,
      "num_input_tokens_seen": 400052872,
      "step": 6935,
      "train_runtime": 39433.6514,
      "train_tokens_per_second": 10144.961
    },
    {
      "epoch": 1.5801457194899817,
      "grad_norm": 0.6177720010044508,
      "learning_rate": 1.049619541420207e-06,
      "loss": 0.2638,
      "num_input_tokens_seen": 400342760,
      "step": 6940,
      "train_runtime": 39459.4436,
      "train_tokens_per_second": 10145.677
    },
    {
      "epoch": 1.5812841530054644,
      "grad_norm": 0.5864650120135761,
      "learning_rate": 1.0441448044980357e-06,
      "loss": 0.2188,
      "num_input_tokens_seen": 400649336,
      "step": 6945,
      "train_runtime": 39485.392,
      "train_tokens_per_second": 10146.774
    },
    {
      "epoch": 1.5824225865209471,
      "grad_norm": 0.7025518984197849,
      "learning_rate": 1.0386827177169717e-06,
      "loss": 0.2486,
      "num_input_tokens_seen": 400922112,
      "step": 6950,
      "train_runtime": 39514.9002,
      "train_tokens_per_second": 10146.1
    },
    {
      "epoch": 1.5835610200364298,
      "grad_norm": 0.6618438266332473,
      "learning_rate": 1.0332332985438248e-06,
      "loss": 0.2283,
      "num_input_tokens_seen": 401229376,
      "step": 6955,
      "train_runtime": 39545.5434,
      "train_tokens_per_second": 10146.007
    },
    {
      "epoch": 1.5846994535519126,
      "grad_norm": 0.6130144476518622,
      "learning_rate": 1.0277965644048959e-06,
      "loss": 0.2548,
      "num_input_tokens_seen": 401519256,
      "step": 6960,
      "train_runtime": 39573.5891,
      "train_tokens_per_second": 10146.142
    },
    {
      "epoch": 1.5858378870673953,
      "grad_norm": 0.5453620138020282,
      "learning_rate": 1.0223725326859201e-06,
      "loss": 0.2417,
      "num_input_tokens_seen": 401817672,
      "step": 6965,
      "train_runtime": 39602.4794,
      "train_tokens_per_second": 10146.276
    },
    {
      "epoch": 1.586976320582878,
      "grad_norm": 0.7353879628669323,
      "learning_rate": 1.016961220732014e-06,
      "loss": 0.2507,
      "num_input_tokens_seen": 402072488,
      "step": 6970,
      "train_runtime": 39631.4797,
      "train_tokens_per_second": 10145.281
    },
    {
      "epoch": 1.5881147540983607,
      "grad_norm": 0.6390203611558507,
      "learning_rate": 1.0115626458476162e-06,
      "loss": 0.2551,
      "num_input_tokens_seen": 402372480,
      "step": 6975,
      "train_runtime": 39660.0843,
      "train_tokens_per_second": 10145.528
    },
    {
      "epoch": 1.5892531876138434,
      "grad_norm": 5.227617582824699,
      "learning_rate": 1.0061768252964388e-06,
      "loss": 0.2467,
      "num_input_tokens_seen": 402654824,
      "step": 6980,
      "train_runtime": 39690.6888,
      "train_tokens_per_second": 10144.818
    },
    {
      "epoch": 1.590391621129326,
      "grad_norm": 0.5364427418498926,
      "learning_rate": 1.0008037763014033e-06,
      "loss": 0.2359,
      "num_input_tokens_seen": 402955584,
      "step": 6985,
      "train_runtime": 39718.8719,
      "train_tokens_per_second": 10145.192
    },
    {
      "epoch": 1.5915300546448088,
      "grad_norm": 0.6581973380417357,
      "learning_rate": 9.954435160445913e-07,
      "loss": 0.2495,
      "num_input_tokens_seen": 403260032,
      "step": 6990,
      "train_runtime": 39745.161,
      "train_tokens_per_second": 10146.142
    },
    {
      "epoch": 1.5926684881602915,
      "grad_norm": 0.6014326217050916,
      "learning_rate": 9.900960616671895e-07,
      "loss": 0.2511,
      "num_input_tokens_seen": 403561040,
      "step": 6995,
      "train_runtime": 39773.8622,
      "train_tokens_per_second": 10146.388
    },
    {
      "epoch": 1.5938069216757742,
      "grad_norm": 0.6655714026086997,
      "learning_rate": 9.847614302694324e-07,
      "loss": 0.2501,
      "num_input_tokens_seen": 403855784,
      "step": 7000,
      "train_runtime": 39796.6051,
      "train_tokens_per_second": 10147.996
    },
    {
      "epoch": 1.594945355191257,
      "grad_norm": 0.6036257998476097,
      "learning_rate": 9.794396389105487e-07,
      "loss": 0.2543,
      "num_input_tokens_seen": 404164744,
      "step": 7005,
      "train_runtime": 39827.3986,
      "train_tokens_per_second": 10147.907
    },
    {
      "epoch": 1.5960837887067396,
      "grad_norm": 0.5809900824120195,
      "learning_rate": 9.741307046087072e-07,
      "loss": 0.2473,
      "num_input_tokens_seen": 404445240,
      "step": 7010,
      "train_runtime": 39853.993,
      "train_tokens_per_second": 10148.174
    },
    {
      "epoch": 1.5972222222222223,
      "grad_norm": 0.5860064313075095,
      "learning_rate": 9.68834644340962e-07,
      "loss": 0.2603,
      "num_input_tokens_seen": 404742960,
      "step": 7015,
      "train_runtime": 39885.5568,
      "train_tokens_per_second": 10147.607
    },
    {
      "epoch": 1.598360655737705,
      "grad_norm": 0.615453821013076,
      "learning_rate": 9.63551475043199e-07,
      "loss": 0.2482,
      "num_input_tokens_seen": 405054432,
      "step": 7020,
      "train_runtime": 39910.5652,
      "train_tokens_per_second": 10149.053
    },
    {
      "epoch": 1.5994990892531877,
      "grad_norm": 0.7212550189326133,
      "learning_rate": 9.582812136100784e-07,
      "loss": 0.2489,
      "num_input_tokens_seen": 405336400,
      "step": 7025,
      "train_runtime": 39934.6675,
      "train_tokens_per_second": 10149.988
    },
    {
      "epoch": 1.6006375227686704,
      "grad_norm": 0.6220658120452512,
      "learning_rate": 9.530238768949885e-07,
      "loss": 0.2166,
      "num_input_tokens_seen": 405638984,
      "step": 7030,
      "train_runtime": 39961.6197,
      "train_tokens_per_second": 10150.714
    },
    {
      "epoch": 1.6017759562841531,
      "grad_norm": 0.5386025916096612,
      "learning_rate": 9.477794817099812e-07,
      "loss": 0.2305,
      "num_input_tokens_seen": 405952728,
      "step": 7035,
      "train_runtime": 39991.1518,
      "train_tokens_per_second": 10151.064
    },
    {
      "epoch": 1.6029143897996359,
      "grad_norm": 0.5605819404826327,
      "learning_rate": 9.425480448257268e-07,
      "loss": 0.2338,
      "num_input_tokens_seen": 406242192,
      "step": 7040,
      "train_runtime": 40013.8635,
      "train_tokens_per_second": 10152.536
    },
    {
      "epoch": 1.6040528233151186,
      "grad_norm": 0.5534106958510162,
      "learning_rate": 9.373295829714562e-07,
      "loss": 0.2626,
      "num_input_tokens_seen": 406541800,
      "step": 7045,
      "train_runtime": 40033.4962,
      "train_tokens_per_second": 10155.041
    },
    {
      "epoch": 1.605191256830601,
      "grad_norm": 0.6040219324245127,
      "learning_rate": 9.321241128349068e-07,
      "loss": 0.2443,
      "num_input_tokens_seen": 406837176,
      "step": 7050,
      "train_runtime": 40064.1822,
      "train_tokens_per_second": 10154.636
    },
    {
      "epoch": 1.6063296903460837,
      "grad_norm": 0.6782063224390903,
      "learning_rate": 9.269316510622733e-07,
      "loss": 0.2694,
      "num_input_tokens_seen": 407136480,
      "step": 7055,
      "train_runtime": 40095.3426,
      "train_tokens_per_second": 10154.209
    },
    {
      "epoch": 1.6074681238615665,
      "grad_norm": 0.6699682754874458,
      "learning_rate": 9.217522142581492e-07,
      "loss": 0.2335,
      "num_input_tokens_seen": 407430240,
      "step": 7060,
      "train_runtime": 40121.0198,
      "train_tokens_per_second": 10155.032
    },
    {
      "epoch": 1.6086065573770492,
      "grad_norm": 0.6144432217882563,
      "learning_rate": 9.165858189854815e-07,
      "loss": 0.2458,
      "num_input_tokens_seen": 407688368,
      "step": 7065,
      "train_runtime": 40146.1837,
      "train_tokens_per_second": 10155.096
    },
    {
      "epoch": 1.6097449908925319,
      "grad_norm": 0.5882162968842995,
      "learning_rate": 9.114324817655068e-07,
      "loss": 0.2466,
      "num_input_tokens_seen": 407984000,
      "step": 7070,
      "train_runtime": 40173.706,
      "train_tokens_per_second": 10155.498
    },
    {
      "epoch": 1.6108834244080146,
      "grad_norm": 0.728691306624718,
      "learning_rate": 9.062922190777079e-07,
      "loss": 0.2235,
      "num_input_tokens_seen": 408278104,
      "step": 7075,
      "train_runtime": 40201.4591,
      "train_tokens_per_second": 10155.803
    },
    {
      "epoch": 1.6120218579234973,
      "grad_norm": 0.7246426900875359,
      "learning_rate": 9.01165047359755e-07,
      "loss": 0.252,
      "num_input_tokens_seen": 408558144,
      "step": 7080,
      "train_runtime": 40228.6368,
      "train_tokens_per_second": 10155.903
    },
    {
      "epoch": 1.61316029143898,
      "grad_norm": 0.7234443360391667,
      "learning_rate": 8.960509830074582e-07,
      "loss": 0.2508,
      "num_input_tokens_seen": 408849080,
      "step": 7085,
      "train_runtime": 40256.5099,
      "train_tokens_per_second": 10156.099
    },
    {
      "epoch": 1.6142987249544627,
      "grad_norm": 0.60755110343456,
      "learning_rate": 8.909500423747108e-07,
      "loss": 0.2612,
      "num_input_tokens_seen": 409143168,
      "step": 7090,
      "train_runtime": 40286.5359,
      "train_tokens_per_second": 10155.829
    },
    {
      "epoch": 1.6154371584699454,
      "grad_norm": 0.5813823337111967,
      "learning_rate": 8.858622417734391e-07,
      "loss": 0.2446,
      "num_input_tokens_seen": 409435184,
      "step": 7095,
      "train_runtime": 40316.0143,
      "train_tokens_per_second": 10155.646
    },
    {
      "epoch": 1.616575591985428,
      "grad_norm": 0.5853745717747535,
      "learning_rate": 8.807875974735503e-07,
      "loss": 0.2265,
      "num_input_tokens_seen": 409713208,
      "step": 7100,
      "train_runtime": 40342.8909,
      "train_tokens_per_second": 10155.772
    },
    {
      "epoch": 1.6177140255009106,
      "grad_norm": 0.6476151746231777,
      "learning_rate": 8.757261257028776e-07,
      "loss": 0.2479,
      "num_input_tokens_seen": 409980384,
      "step": 7105,
      "train_runtime": 40366.5945,
      "train_tokens_per_second": 10156.427
    },
    {
      "epoch": 1.6188524590163933,
      "grad_norm": 0.6753531171021958,
      "learning_rate": 8.706778426471363e-07,
      "loss": 0.2536,
      "num_input_tokens_seen": 410277144,
      "step": 7110,
      "train_runtime": 40401.3032,
      "train_tokens_per_second": 10155.047
    },
    {
      "epoch": 1.619990892531876,
      "grad_norm": 0.565632557507784,
      "learning_rate": 8.656427644498617e-07,
      "loss": 0.2294,
      "num_input_tokens_seen": 410571784,
      "step": 7115,
      "train_runtime": 40430.5558,
      "train_tokens_per_second": 10154.987
    },
    {
      "epoch": 1.6211293260473587,
      "grad_norm": 0.6331232556727413,
      "learning_rate": 8.606209072123623e-07,
      "loss": 0.2445,
      "num_input_tokens_seen": 410856040,
      "step": 7120,
      "train_runtime": 40458.6064,
      "train_tokens_per_second": 10154.973
    },
    {
      "epoch": 1.6222677595628414,
      "grad_norm": 0.6484728162884074,
      "learning_rate": 8.556122869936706e-07,
      "loss": 0.2585,
      "num_input_tokens_seen": 411109640,
      "step": 7125,
      "train_runtime": 40487.3842,
      "train_tokens_per_second": 10154.018
    },
    {
      "epoch": 1.6234061930783241,
      "grad_norm": 0.7052086141098372,
      "learning_rate": 8.506169198104879e-07,
      "loss": 0.2652,
      "num_input_tokens_seen": 411405992,
      "step": 7130,
      "train_runtime": 40514.3173,
      "train_tokens_per_second": 10154.583
    },
    {
      "epoch": 1.6245446265938068,
      "grad_norm": 0.5006078215863179,
      "learning_rate": 8.456348216371352e-07,
      "loss": 0.2349,
      "num_input_tokens_seen": 411703448,
      "step": 7135,
      "train_runtime": 40539.7841,
      "train_tokens_per_second": 10155.541
    },
    {
      "epoch": 1.6256830601092895,
      "grad_norm": 0.5665030554087213,
      "learning_rate": 8.406660084055007e-07,
      "loss": 0.2355,
      "num_input_tokens_seen": 411984872,
      "step": 7140,
      "train_runtime": 40560.1021,
      "train_tokens_per_second": 10157.392
    },
    {
      "epoch": 1.6268214936247722,
      "grad_norm": 0.5342536567050857,
      "learning_rate": 8.357104960049906e-07,
      "loss": 0.2439,
      "num_input_tokens_seen": 412285128,
      "step": 7145,
      "train_runtime": 40593.7095,
      "train_tokens_per_second": 10156.38
    },
    {
      "epoch": 1.627959927140255,
      "grad_norm": 0.5335336312426865,
      "learning_rate": 8.307683002824774e-07,
      "loss": 0.2409,
      "num_input_tokens_seen": 412558208,
      "step": 7150,
      "train_runtime": 40621.2455,
      "train_tokens_per_second": 10156.218
    },
    {
      "epoch": 1.6290983606557377,
      "grad_norm": 0.637405176496244,
      "learning_rate": 8.258394370422479e-07,
      "loss": 0.2353,
      "num_input_tokens_seen": 412849328,
      "step": 7155,
      "train_runtime": 40642.8466,
      "train_tokens_per_second": 10157.983
    },
    {
      "epoch": 1.6302367941712204,
      "grad_norm": 0.8048570004847423,
      "learning_rate": 8.209239220459547e-07,
      "loss": 0.2548,
      "num_input_tokens_seen": 413117304,
      "step": 7160,
      "train_runtime": 40671.5131,
      "train_tokens_per_second": 10157.412
    },
    {
      "epoch": 1.631375227686703,
      "grad_norm": 0.5869041160699315,
      "learning_rate": 8.160217710125661e-07,
      "loss": 0.2506,
      "num_input_tokens_seen": 413412776,
      "step": 7165,
      "train_runtime": 40698.6488,
      "train_tokens_per_second": 10157.899
    },
    {
      "epoch": 1.6325136612021858,
      "grad_norm": 0.5913843673029897,
      "learning_rate": 8.111329996183126e-07,
      "loss": 0.2435,
      "num_input_tokens_seen": 413699856,
      "step": 7170,
      "train_runtime": 40727.9524,
      "train_tokens_per_second": 10157.639
    },
    {
      "epoch": 1.6336520947176685,
      "grad_norm": 0.7095862318566484,
      "learning_rate": 8.062576234966407e-07,
      "loss": 0.2246,
      "num_input_tokens_seen": 413975200,
      "step": 7175,
      "train_runtime": 40757.874,
      "train_tokens_per_second": 10156.938
    },
    {
      "epoch": 1.6347905282331512,
      "grad_norm": 0.6474649056152475,
      "learning_rate": 8.013956582381599e-07,
      "loss": 0.2475,
      "num_input_tokens_seen": 414278280,
      "step": 7180,
      "train_runtime": 40785.1607,
      "train_tokens_per_second": 10157.574
    },
    {
      "epoch": 1.635928961748634,
      "grad_norm": 0.7123715261113813,
      "learning_rate": 7.965471193905955e-07,
      "loss": 0.2545,
      "num_input_tokens_seen": 414563184,
      "step": 7185,
      "train_runtime": 40810.7755,
      "train_tokens_per_second": 10158.18
    },
    {
      "epoch": 1.6370673952641166,
      "grad_norm": 0.6708331339180332,
      "learning_rate": 7.917120224587349e-07,
      "loss": 0.2603,
      "num_input_tokens_seen": 414835952,
      "step": 7190,
      "train_runtime": 40845.0764,
      "train_tokens_per_second": 10156.327
    },
    {
      "epoch": 1.6382058287795993,
      "grad_norm": 0.6670000934824174,
      "learning_rate": 7.868903829043845e-07,
      "loss": 0.2501,
      "num_input_tokens_seen": 415140896,
      "step": 7195,
      "train_runtime": 40877.343,
      "train_tokens_per_second": 10155.77
    },
    {
      "epoch": 1.639344262295082,
      "grad_norm": 0.6641508658774579,
      "learning_rate": 7.820822161463132e-07,
      "loss": 0.2474,
      "num_input_tokens_seen": 415444144,
      "step": 7200,
      "train_runtime": 40907.2382,
      "train_tokens_per_second": 10155.761
    },
    {
      "epoch": 1.6404826958105647,
      "grad_norm": 0.6691167394010623,
      "learning_rate": 7.772875375602073e-07,
      "loss": 0.2379,
      "num_input_tokens_seen": 415719000,
      "step": 7205,
      "train_runtime": 40939.4641,
      "train_tokens_per_second": 10154.481
    },
    {
      "epoch": 1.6416211293260474,
      "grad_norm": 0.6020787300118979,
      "learning_rate": 7.7250636247862e-07,
      "loss": 0.2319,
      "num_input_tokens_seen": 416005880,
      "step": 7210,
      "train_runtime": 40970.9141,
      "train_tokens_per_second": 10153.688
    },
    {
      "epoch": 1.6427595628415301,
      "grad_norm": 0.6300308619788163,
      "learning_rate": 7.677387061909225e-07,
      "loss": 0.2526,
      "num_input_tokens_seen": 416287088,
      "step": 7215,
      "train_runtime": 41002.6508,
      "train_tokens_per_second": 10152.687
    },
    {
      "epoch": 1.6438979963570128,
      "grad_norm": 0.6099397455418852,
      "learning_rate": 7.629845839432559e-07,
      "loss": 0.2761,
      "num_input_tokens_seen": 416558648,
      "step": 7220,
      "train_runtime": 41027.8505,
      "train_tokens_per_second": 10153.07
    },
    {
      "epoch": 1.6450364298724955,
      "grad_norm": 0.7728117526808749,
      "learning_rate": 7.582440109384808e-07,
      "loss": 0.2887,
      "num_input_tokens_seen": 416839792,
      "step": 7225,
      "train_runtime": 41055.7677,
      "train_tokens_per_second": 10153.014
    },
    {
      "epoch": 1.6461748633879782,
      "grad_norm": 0.595525270724856,
      "learning_rate": 7.53517002336131e-07,
      "loss": 0.2484,
      "num_input_tokens_seen": 417111376,
      "step": 7230,
      "train_runtime": 41086.51,
      "train_tokens_per_second": 10152.027
    },
    {
      "epoch": 1.647313296903461,
      "grad_norm": 0.6677936002983921,
      "learning_rate": 7.488035732523607e-07,
      "loss": 0.2269,
      "num_input_tokens_seen": 417395136,
      "step": 7235,
      "train_runtime": 41114.8331,
      "train_tokens_per_second": 10151.936
    },
    {
      "epoch": 1.6484517304189437,
      "grad_norm": 0.5079411343857689,
      "learning_rate": 7.441037387599048e-07,
      "loss": 0.2167,
      "num_input_tokens_seen": 417687504,
      "step": 7240,
      "train_runtime": 41145.3012,
      "train_tokens_per_second": 10151.524
    },
    {
      "epoch": 1.6495901639344264,
      "grad_norm": 0.6743122678369717,
      "learning_rate": 7.394175138880195e-07,
      "loss": 0.2445,
      "num_input_tokens_seen": 417969528,
      "step": 7245,
      "train_runtime": 41175.4168,
      "train_tokens_per_second": 10150.948
    },
    {
      "epoch": 1.650728597449909,
      "grad_norm": 0.6697857706325355,
      "learning_rate": 7.347449136224422e-07,
      "loss": 0.2806,
      "num_input_tokens_seen": 418259944,
      "step": 7250,
      "train_runtime": 41202.6451,
      "train_tokens_per_second": 10151.289
    },
    {
      "epoch": 1.6518670309653918,
      "grad_norm": 0.6321955078998158,
      "learning_rate": 7.300859529053422e-07,
      "loss": 0.2437,
      "num_input_tokens_seen": 418535768,
      "step": 7255,
      "train_runtime": 41226.2556,
      "train_tokens_per_second": 10152.165
    },
    {
      "epoch": 1.6530054644808743,
      "grad_norm": 0.647243667134972,
      "learning_rate": 7.254406466352682e-07,
      "loss": 0.2484,
      "num_input_tokens_seen": 418828168,
      "step": 7260,
      "train_runtime": 41255.0743,
      "train_tokens_per_second": 10152.161
    },
    {
      "epoch": 1.654143897996357,
      "grad_norm": 0.6577209558554395,
      "learning_rate": 7.208090096671072e-07,
      "loss": 0.2661,
      "num_input_tokens_seen": 419122320,
      "step": 7265,
      "train_runtime": 41283.0292,
      "train_tokens_per_second": 10152.412
    },
    {
      "epoch": 1.6552823315118397,
      "grad_norm": 0.5482870756680917,
      "learning_rate": 7.161910568120345e-07,
      "loss": 0.2335,
      "num_input_tokens_seen": 419412024,
      "step": 7270,
      "train_runtime": 41313.1199,
      "train_tokens_per_second": 10152.03
    },
    {
      "epoch": 1.6564207650273224,
      "grad_norm": 0.550188783201533,
      "learning_rate": 7.115868028374634e-07,
      "loss": 0.2401,
      "num_input_tokens_seen": 419714880,
      "step": 7275,
      "train_runtime": 41340.5212,
      "train_tokens_per_second": 10152.627
    },
    {
      "epoch": 1.657559198542805,
      "grad_norm": 0.6140780205143517,
      "learning_rate": 7.069962624670051e-07,
      "loss": 0.2421,
      "num_input_tokens_seen": 420024280,
      "step": 7280,
      "train_runtime": 41367.715,
      "train_tokens_per_second": 10153.432
    },
    {
      "epoch": 1.6586976320582878,
      "grad_norm": 0.665517945602968,
      "learning_rate": 7.024194503804121e-07,
      "loss": 0.2667,
      "num_input_tokens_seen": 420300320,
      "step": 7285,
      "train_runtime": 41400.0851,
      "train_tokens_per_second": 10152.161
    },
    {
      "epoch": 1.6598360655737705,
      "grad_norm": 0.6024211614500167,
      "learning_rate": 6.978563812135386e-07,
      "loss": 0.2256,
      "num_input_tokens_seen": 420572344,
      "step": 7290,
      "train_runtime": 41422.4763,
      "train_tokens_per_second": 10153.24
    },
    {
      "epoch": 1.6609744990892532,
      "grad_norm": 0.6529749584778654,
      "learning_rate": 6.933070695582916e-07,
      "loss": 0.2644,
      "num_input_tokens_seen": 420872008,
      "step": 7295,
      "train_runtime": 41447.6675,
      "train_tokens_per_second": 10154.299
    },
    {
      "epoch": 1.662112932604736,
      "grad_norm": 0.5914165071372024,
      "learning_rate": 6.88771529962583e-07,
      "loss": 0.2366,
      "num_input_tokens_seen": 421156064,
      "step": 7300,
      "train_runtime": 41479.3338,
      "train_tokens_per_second": 10153.395
    },
    {
      "epoch": 1.6632513661202186,
      "grad_norm": 0.6364993119266165,
      "learning_rate": 6.842497769302842e-07,
      "loss": 0.2315,
      "num_input_tokens_seen": 421459960,
      "step": 7305,
      "train_runtime": 41501.6695,
      "train_tokens_per_second": 10155.253
    },
    {
      "epoch": 1.6643897996357013,
      "grad_norm": 0.5972039296851418,
      "learning_rate": 6.797418249211801e-07,
      "loss": 0.2295,
      "num_input_tokens_seen": 421749344,
      "step": 7310,
      "train_runtime": 41531.7157,
      "train_tokens_per_second": 10154.874
    },
    {
      "epoch": 1.6655282331511838,
      "grad_norm": 0.5505189524683765,
      "learning_rate": 6.752476883509218e-07,
      "loss": 0.2232,
      "num_input_tokens_seen": 422051464,
      "step": 7315,
      "train_runtime": 41560.251,
      "train_tokens_per_second": 10155.171
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.5165833617118742,
      "learning_rate": 6.707673815909799e-07,
      "loss": 0.2168,
      "num_input_tokens_seen": 422351504,
      "step": 7320,
      "train_runtime": 41590.5125,
      "train_tokens_per_second": 10154.996
    },
    {
      "epoch": 1.6678051001821492,
      "grad_norm": 0.6398681865315481,
      "learning_rate": 6.663009189686037e-07,
      "loss": 0.2543,
      "num_input_tokens_seen": 422621992,
      "step": 7325,
      "train_runtime": 41616.3933,
      "train_tokens_per_second": 10155.181
    },
    {
      "epoch": 1.668943533697632,
      "grad_norm": 0.570166157686179,
      "learning_rate": 6.61848314766767e-07,
      "loss": 0.2538,
      "num_input_tokens_seen": 422914504,
      "step": 7330,
      "train_runtime": 41643.4056,
      "train_tokens_per_second": 10155.618
    },
    {
      "epoch": 1.6700819672131146,
      "grad_norm": 0.5693964828507665,
      "learning_rate": 6.574095832241289e-07,
      "loss": 0.2449,
      "num_input_tokens_seen": 423190768,
      "step": 7335,
      "train_runtime": 41668.9897,
      "train_tokens_per_second": 10156.012
    },
    {
      "epoch": 1.6712204007285973,
      "grad_norm": 0.6125706133600172,
      "learning_rate": 6.529847385349847e-07,
      "loss": 0.236,
      "num_input_tokens_seen": 423473936,
      "step": 7340,
      "train_runtime": 41699.0869,
      "train_tokens_per_second": 10155.473
    },
    {
      "epoch": 1.67235883424408,
      "grad_norm": 0.5668757231569517,
      "learning_rate": 6.485737948492237e-07,
      "loss": 0.2353,
      "num_input_tokens_seen": 423772544,
      "step": 7345,
      "train_runtime": 41720.3397,
      "train_tokens_per_second": 10157.457
    },
    {
      "epoch": 1.6734972677595628,
      "grad_norm": 0.6665589371264055,
      "learning_rate": 6.441767662722814e-07,
      "loss": 0.26,
      "num_input_tokens_seen": 424053496,
      "step": 7350,
      "train_runtime": 41749.627,
      "train_tokens_per_second": 10157.061
    },
    {
      "epoch": 1.6746357012750455,
      "grad_norm": 0.5238150903375042,
      "learning_rate": 6.397936668650944e-07,
      "loss": 0.2828,
      "num_input_tokens_seen": 424339920,
      "step": 7355,
      "train_runtime": 41786.6574,
      "train_tokens_per_second": 10154.914
    },
    {
      "epoch": 1.6757741347905282,
      "grad_norm": 0.6804916659706369,
      "learning_rate": 6.354245106440577e-07,
      "loss": 0.2444,
      "num_input_tokens_seen": 424649104,
      "step": 7360,
      "train_runtime": 41811.2849,
      "train_tokens_per_second": 10156.328
    },
    {
      "epoch": 1.6769125683060109,
      "grad_norm": 0.6411793609153961,
      "learning_rate": 6.310693115809769e-07,
      "loss": 0.2558,
      "num_input_tokens_seen": 424914392,
      "step": 7365,
      "train_runtime": 41839.7559,
      "train_tokens_per_second": 10155.757
    },
    {
      "epoch": 1.6780510018214936,
      "grad_norm": 0.6087036542945676,
      "learning_rate": 6.267280836030259e-07,
      "loss": 0.2287,
      "num_input_tokens_seen": 425217824,
      "step": 7370,
      "train_runtime": 41865.6758,
      "train_tokens_per_second": 10156.717
    },
    {
      "epoch": 1.6791894353369763,
      "grad_norm": 0.527913152951376,
      "learning_rate": 6.224008405927018e-07,
      "loss": 0.2279,
      "num_input_tokens_seen": 425513440,
      "step": 7375,
      "train_runtime": 41893.8963,
      "train_tokens_per_second": 10156.932
    },
    {
      "epoch": 1.680327868852459,
      "grad_norm": 0.5672808809662798,
      "learning_rate": 6.180875963877792e-07,
      "loss": 0.2282,
      "num_input_tokens_seen": 425797080,
      "step": 7380,
      "train_runtime": 41921.786,
      "train_tokens_per_second": 10156.94
    },
    {
      "epoch": 1.6814663023679417,
      "grad_norm": 0.5797440829658635,
      "learning_rate": 6.137883647812681e-07,
      "loss": 0.2648,
      "num_input_tokens_seen": 426094544,
      "step": 7385,
      "train_runtime": 41952.866,
      "train_tokens_per_second": 10156.506
    },
    {
      "epoch": 1.6826047358834244,
      "grad_norm": 0.8619444874314824,
      "learning_rate": 6.095031595213674e-07,
      "loss": 0.2722,
      "num_input_tokens_seen": 426347232,
      "step": 7390,
      "train_runtime": 41985.4758,
      "train_tokens_per_second": 10154.636
    },
    {
      "epoch": 1.6837431693989071,
      "grad_norm": 0.6271338756201452,
      "learning_rate": 6.052319943114238e-07,
      "loss": 0.2197,
      "num_input_tokens_seen": 426647824,
      "step": 7395,
      "train_runtime": 42009.7029,
      "train_tokens_per_second": 10155.935
    },
    {
      "epoch": 1.6848816029143898,
      "grad_norm": 0.6718311625319994,
      "learning_rate": 6.009748828098838e-07,
      "loss": 0.2257,
      "num_input_tokens_seen": 426937088,
      "step": 7400,
      "train_runtime": 42038.8057,
      "train_tokens_per_second": 10155.785
    },
    {
      "epoch": 1.6860200364298725,
      "grad_norm": 0.8728671950399761,
      "learning_rate": 5.967318386302539e-07,
      "loss": 0.2434,
      "num_input_tokens_seen": 427238040,
      "step": 7405,
      "train_runtime": 42064.3689,
      "train_tokens_per_second": 10156.768
    },
    {
      "epoch": 1.6871584699453552,
      "grad_norm": 0.5922540694726464,
      "learning_rate": 5.925028753410583e-07,
      "loss": 0.2364,
      "num_input_tokens_seen": 427515704,
      "step": 7410,
      "train_runtime": 42090.4589,
      "train_tokens_per_second": 10157.069
    },
    {
      "epoch": 1.688296903460838,
      "grad_norm": 0.6344596723882897,
      "learning_rate": 5.882880064657892e-07,
      "loss": 0.2422,
      "num_input_tokens_seen": 427809968,
      "step": 7415,
      "train_runtime": 42120.242,
      "train_tokens_per_second": 10156.873
    },
    {
      "epoch": 1.6894353369763206,
      "grad_norm": 0.5320459894707016,
      "learning_rate": 5.840872454828689e-07,
      "loss": 0.2448,
      "num_input_tokens_seen": 428102840,
      "step": 7420,
      "train_runtime": 42144.5717,
      "train_tokens_per_second": 10157.959
    },
    {
      "epoch": 1.6905737704918034,
      "grad_norm": 0.6910688475157037,
      "learning_rate": 5.799006058256046e-07,
      "loss": 0.2427,
      "num_input_tokens_seen": 428389944,
      "step": 7425,
      "train_runtime": 42174.1923,
      "train_tokens_per_second": 10157.632
    },
    {
      "epoch": 1.691712204007286,
      "grad_norm": 0.6145960407531474,
      "learning_rate": 5.757281008821458e-07,
      "loss": 0.2379,
      "num_input_tokens_seen": 428667608,
      "step": 7430,
      "train_runtime": 42202.4316,
      "train_tokens_per_second": 10157.415
    },
    {
      "epoch": 1.6928506375227688,
      "grad_norm": 0.5900963645150759,
      "learning_rate": 5.715697439954432e-07,
      "loss": 0.2673,
      "num_input_tokens_seen": 428954448,
      "step": 7435,
      "train_runtime": 42228.5377,
      "train_tokens_per_second": 10157.928
    },
    {
      "epoch": 1.6939890710382515,
      "grad_norm": 0.5575220982701504,
      "learning_rate": 5.674255484632024e-07,
      "loss": 0.2523,
      "num_input_tokens_seen": 429236960,
      "step": 7440,
      "train_runtime": 42253.239,
      "train_tokens_per_second": 10158.676
    },
    {
      "epoch": 1.6951275045537342,
      "grad_norm": 0.6187475068814297,
      "learning_rate": 5.632955275378449e-07,
      "loss": 0.2246,
      "num_input_tokens_seen": 429509472,
      "step": 7445,
      "train_runtime": 42279.2338,
      "train_tokens_per_second": 10158.875
    },
    {
      "epoch": 1.6962659380692169,
      "grad_norm": 0.6341009609491899,
      "learning_rate": 5.591796944264616e-07,
      "loss": 0.2471,
      "num_input_tokens_seen": 429812560,
      "step": 7450,
      "train_runtime": 42307.4218,
      "train_tokens_per_second": 10159.271
    },
    {
      "epoch": 1.6974043715846996,
      "grad_norm": 0.6112747685699345,
      "learning_rate": 5.550780622907792e-07,
      "loss": 0.2303,
      "num_input_tokens_seen": 430113800,
      "step": 7455,
      "train_runtime": 42337.9378,
      "train_tokens_per_second": 10159.064
    },
    {
      "epoch": 1.6985428051001823,
      "grad_norm": 0.6348003295498496,
      "learning_rate": 5.509906442471058e-07,
      "loss": 0.2509,
      "num_input_tokens_seen": 430405664,
      "step": 7460,
      "train_runtime": 42371.4976,
      "train_tokens_per_second": 10157.905
    },
    {
      "epoch": 1.699681238615665,
      "grad_norm": 0.5342753985485985,
      "learning_rate": 5.46917453366298e-07,
      "loss": 0.2203,
      "num_input_tokens_seen": 430712136,
      "step": 7465,
      "train_runtime": 42399.3914,
      "train_tokens_per_second": 10158.451
    },
    {
      "epoch": 1.7008196721311475,
      "grad_norm": 0.661251178070617,
      "learning_rate": 5.428585026737171e-07,
      "loss": 0.2354,
      "num_input_tokens_seen": 430986840,
      "step": 7470,
      "train_runtime": 42425.5954,
      "train_tokens_per_second": 10158.652
    },
    {
      "epoch": 1.7019581056466302,
      "grad_norm": 0.6099644034779629,
      "learning_rate": 5.388138051491848e-07,
      "loss": 0.2435,
      "num_input_tokens_seen": 431290576,
      "step": 7475,
      "train_runtime": 42451.4674,
      "train_tokens_per_second": 10159.615
    },
    {
      "epoch": 1.703096539162113,
      "grad_norm": 0.5218852055150947,
      "learning_rate": 5.347833737269442e-07,
      "loss": 0.2306,
      "num_input_tokens_seen": 431581736,
      "step": 7480,
      "train_runtime": 42484.5992,
      "train_tokens_per_second": 10158.546
    },
    {
      "epoch": 1.7042349726775956,
      "grad_norm": 0.8177829046221388,
      "learning_rate": 5.307672212956177e-07,
      "loss": 0.245,
      "num_input_tokens_seen": 431847600,
      "step": 7485,
      "train_runtime": 42513.4995,
      "train_tokens_per_second": 10157.894
    },
    {
      "epoch": 1.7053734061930783,
      "grad_norm": 0.5898866905941736,
      "learning_rate": 5.267653606981693e-07,
      "loss": 0.2396,
      "num_input_tokens_seen": 432144056,
      "step": 7490,
      "train_runtime": 42545.5103,
      "train_tokens_per_second": 10157.219
    },
    {
      "epoch": 1.706511839708561,
      "grad_norm": 0.7545199332365502,
      "learning_rate": 5.22777804731856e-07,
      "loss": 0.257,
      "num_input_tokens_seen": 432450176,
      "step": 7495,
      "train_runtime": 42571.5195,
      "train_tokens_per_second": 10158.204
    },
    {
      "epoch": 1.7076502732240437,
      "grad_norm": 0.7019418757615417,
      "learning_rate": 5.188045661481922e-07,
      "loss": 0.2353,
      "num_input_tokens_seen": 432732536,
      "step": 7500,
      "train_runtime": 42598.8658,
      "train_tokens_per_second": 10158.311
    },
    {
      "epoch": 1.7087887067395264,
      "grad_norm": 0.5917864529026189,
      "learning_rate": 5.148456576529082e-07,
      "loss": 0.2349,
      "num_input_tokens_seen": 433025976,
      "step": 7505,
      "train_runtime": 42625.8284,
      "train_tokens_per_second": 10158.77
    },
    {
      "epoch": 1.7099271402550091,
      "grad_norm": 0.607758460811708,
      "learning_rate": 5.109010919059087e-07,
      "loss": 0.2411,
      "num_input_tokens_seen": 433312952,
      "step": 7510,
      "train_runtime": 42649.6669,
      "train_tokens_per_second": 10159.82
    },
    {
      "epoch": 1.7110655737704918,
      "grad_norm": 0.4966773270656375,
      "learning_rate": 5.069708815212343e-07,
      "loss": 0.2281,
      "num_input_tokens_seen": 433605456,
      "step": 7515,
      "train_runtime": 42684.1829,
      "train_tokens_per_second": 10158.457
    },
    {
      "epoch": 1.7122040072859745,
      "grad_norm": 0.5274455055890709,
      "learning_rate": 5.030550390670169e-07,
      "loss": 0.2355,
      "num_input_tokens_seen": 433895064,
      "step": 7520,
      "train_runtime": 42716.1506,
      "train_tokens_per_second": 10157.635
    },
    {
      "epoch": 1.713342440801457,
      "grad_norm": 0.7423123429867762,
      "learning_rate": 4.991535770654449e-07,
      "loss": 0.2772,
      "num_input_tokens_seen": 434192968,
      "step": 7525,
      "train_runtime": 42745.893,
      "train_tokens_per_second": 10157.536
    },
    {
      "epoch": 1.7144808743169397,
      "grad_norm": 0.632743956895355,
      "learning_rate": 4.952665079927188e-07,
      "loss": 0.2398,
      "num_input_tokens_seen": 434484384,
      "step": 7530,
      "train_runtime": 42782.2542,
      "train_tokens_per_second": 10155.715
    },
    {
      "epoch": 1.7156193078324224,
      "grad_norm": 0.6651510294079065,
      "learning_rate": 4.913938442790128e-07,
      "loss": 0.2724,
      "num_input_tokens_seen": 434769344,
      "step": 7535,
      "train_runtime": 42806.8317,
      "train_tokens_per_second": 10156.541
    },
    {
      "epoch": 1.7167577413479052,
      "grad_norm": 0.6055683637140316,
      "learning_rate": 4.875355983084384e-07,
      "loss": 0.2494,
      "num_input_tokens_seen": 435064240,
      "step": 7540,
      "train_runtime": 42838.8918,
      "train_tokens_per_second": 10155.824
    },
    {
      "epoch": 1.7178961748633879,
      "grad_norm": 0.5657389405234465,
      "learning_rate": 4.836917824189974e-07,
      "loss": 0.2556,
      "num_input_tokens_seen": 435383720,
      "step": 7545,
      "train_runtime": 42868.7813,
      "train_tokens_per_second": 10156.195
    },
    {
      "epoch": 1.7190346083788706,
      "grad_norm": 0.6450303526520849,
      "learning_rate": 4.798624089025495e-07,
      "loss": 0.246,
      "num_input_tokens_seen": 435649696,
      "step": 7550,
      "train_runtime": 42897.185,
      "train_tokens_per_second": 10155.671
    },
    {
      "epoch": 1.7201730418943533,
      "grad_norm": 0.6154488664497993,
      "learning_rate": 4.7604749000476846e-07,
      "loss": 0.2552,
      "num_input_tokens_seen": 435939992,
      "step": 7555,
      "train_runtime": 42925.4765,
      "train_tokens_per_second": 10155.74
    },
    {
      "epoch": 1.721311475409836,
      "grad_norm": 0.8176711321962297,
      "learning_rate": 4.7224703792510594e-07,
      "loss": 0.2548,
      "num_input_tokens_seen": 436223280,
      "step": 7560,
      "train_runtime": 42955.0373,
      "train_tokens_per_second": 10155.346
    },
    {
      "epoch": 1.7224499089253187,
      "grad_norm": 0.5610322834864184,
      "learning_rate": 4.6846106481675035e-07,
      "loss": 0.2305,
      "num_input_tokens_seen": 436532200,
      "step": 7565,
      "train_runtime": 42981.8051,
      "train_tokens_per_second": 10156.209
    },
    {
      "epoch": 1.7235883424408014,
      "grad_norm": 0.6911568064480084,
      "learning_rate": 4.6468958278658993e-07,
      "loss": 0.2504,
      "num_input_tokens_seen": 436801576,
      "step": 7570,
      "train_runtime": 43006.7517,
      "train_tokens_per_second": 10156.581
    },
    {
      "epoch": 1.724726775956284,
      "grad_norm": 0.4998411931543715,
      "learning_rate": 4.609326038951717e-07,
      "loss": 0.215,
      "num_input_tokens_seen": 437092672,
      "step": 7575,
      "train_runtime": 43035.1914,
      "train_tokens_per_second": 10156.634
    },
    {
      "epoch": 1.7258652094717668,
      "grad_norm": 0.6087993792057537,
      "learning_rate": 4.5719014015666474e-07,
      "loss": 0.2562,
      "num_input_tokens_seen": 437391592,
      "step": 7580,
      "train_runtime": 43056.6995,
      "train_tokens_per_second": 10158.503
    },
    {
      "epoch": 1.7270036429872495,
      "grad_norm": 0.6459114825299229,
      "learning_rate": 4.534622035388214e-07,
      "loss": 0.2624,
      "num_input_tokens_seen": 437701704,
      "step": 7585,
      "train_runtime": 43087.0545,
      "train_tokens_per_second": 10158.543
    },
    {
      "epoch": 1.7281420765027322,
      "grad_norm": 0.6675450248849955,
      "learning_rate": 4.4974880596293847e-07,
      "loss": 0.2391,
      "num_input_tokens_seen": 437958320,
      "step": 7590,
      "train_runtime": 43113.4828,
      "train_tokens_per_second": 10158.268
    },
    {
      "epoch": 1.729280510018215,
      "grad_norm": 0.7602257952353703,
      "learning_rate": 4.4604995930381977e-07,
      "loss": 0.2543,
      "num_input_tokens_seen": 438254160,
      "step": 7595,
      "train_runtime": 43138.644,
      "train_tokens_per_second": 10159.201
    },
    {
      "epoch": 1.7304189435336976,
      "grad_norm": 0.6037476689276308,
      "learning_rate": 4.4236567538973706e-07,
      "loss": 0.2127,
      "num_input_tokens_seen": 438549864,
      "step": 7600,
      "train_runtime": 43168.2466,
      "train_tokens_per_second": 10159.084
    },
    {
      "epoch": 1.7315573770491803,
      "grad_norm": 0.6900974804384848,
      "learning_rate": 4.386959660023937e-07,
      "loss": 0.2556,
      "num_input_tokens_seen": 438844632,
      "step": 7605,
      "train_runtime": 43195.131,
      "train_tokens_per_second": 10159.586
    },
    {
      "epoch": 1.732695810564663,
      "grad_norm": 0.6763022349704024,
      "learning_rate": 4.35040842876886e-07,
      "loss": 0.2739,
      "num_input_tokens_seen": 439115088,
      "step": 7610,
      "train_runtime": 43224.6864,
      "train_tokens_per_second": 10158.896
    },
    {
      "epoch": 1.7338342440801457,
      "grad_norm": 0.8826718377352689,
      "learning_rate": 4.3140031770166457e-07,
      "loss": 0.2599,
      "num_input_tokens_seen": 439349664,
      "step": 7615,
      "train_runtime": 43248.2512,
      "train_tokens_per_second": 10158.785
    },
    {
      "epoch": 1.7349726775956285,
      "grad_norm": 0.6972494102605169,
      "learning_rate": 4.277744021185021e-07,
      "loss": 0.2402,
      "num_input_tokens_seen": 439629360,
      "step": 7620,
      "train_runtime": 43280.2315,
      "train_tokens_per_second": 10157.74
    },
    {
      "epoch": 1.7361111111111112,
      "grad_norm": 0.6898665279562479,
      "learning_rate": 4.241631077224484e-07,
      "loss": 0.2665,
      "num_input_tokens_seen": 439906376,
      "step": 7625,
      "train_runtime": 43304.4244,
      "train_tokens_per_second": 10158.463
    },
    {
      "epoch": 1.7372495446265939,
      "grad_norm": 0.6999878167531347,
      "learning_rate": 4.205664460617992e-07,
      "loss": 0.2572,
      "num_input_tokens_seen": 440199776,
      "step": 7630,
      "train_runtime": 43329.0788,
      "train_tokens_per_second": 10159.454
    },
    {
      "epoch": 1.7383879781420766,
      "grad_norm": 0.648384774666678,
      "learning_rate": 4.1698442863805646e-07,
      "loss": 0.2447,
      "num_input_tokens_seen": 440482416,
      "step": 7635,
      "train_runtime": 43356.0566,
      "train_tokens_per_second": 10159.651
    },
    {
      "epoch": 1.7395264116575593,
      "grad_norm": 0.6888146197431915,
      "learning_rate": 4.134170669058929e-07,
      "loss": 0.2622,
      "num_input_tokens_seen": 440768312,
      "step": 7640,
      "train_runtime": 43387.4144,
      "train_tokens_per_second": 10158.898
    },
    {
      "epoch": 1.740664845173042,
      "grad_norm": 0.5986215810019514,
      "learning_rate": 4.0986437227311436e-07,
      "loss": 0.2327,
      "num_input_tokens_seen": 441068144,
      "step": 7645,
      "train_runtime": 43420.3314,
      "train_tokens_per_second": 10158.102
    },
    {
      "epoch": 1.7418032786885247,
      "grad_norm": 0.6938601049904567,
      "learning_rate": 4.063263561006242e-07,
      "loss": 0.2541,
      "num_input_tokens_seen": 441335520,
      "step": 7650,
      "train_runtime": 43447.7901,
      "train_tokens_per_second": 10157.836
    },
    {
      "epoch": 1.7429417122040074,
      "grad_norm": 0.7188414797877842,
      "learning_rate": 4.028030297023866e-07,
      "loss": 0.2641,
      "num_input_tokens_seen": 441634632,
      "step": 7655,
      "train_runtime": 43473.0662,
      "train_tokens_per_second": 10158.81
    },
    {
      "epoch": 1.74408014571949,
      "grad_norm": 0.648254019399287,
      "learning_rate": 3.9929440434539114e-07,
      "loss": 0.2526,
      "num_input_tokens_seen": 441927480,
      "step": 7660,
      "train_runtime": 43499.2144,
      "train_tokens_per_second": 10159.436
    },
    {
      "epoch": 1.7452185792349728,
      "grad_norm": 0.5563155389777217,
      "learning_rate": 3.9580049124961274e-07,
      "loss": 0.2318,
      "num_input_tokens_seen": 442208408,
      "step": 7665,
      "train_runtime": 43524.1317,
      "train_tokens_per_second": 10160.074
    },
    {
      "epoch": 1.7463570127504555,
      "grad_norm": 0.817217165006333,
      "learning_rate": 3.9232130158798456e-07,
      "loss": 0.2274,
      "num_input_tokens_seen": 442489072,
      "step": 7670,
      "train_runtime": 43557.3253,
      "train_tokens_per_second": 10158.775
    },
    {
      "epoch": 1.7474954462659382,
      "grad_norm": 0.6122448732247877,
      "learning_rate": 3.888568464863529e-07,
      "loss": 0.2771,
      "num_input_tokens_seen": 442797528,
      "step": 7675,
      "train_runtime": 43588.2873,
      "train_tokens_per_second": 10158.636
    },
    {
      "epoch": 1.748633879781421,
      "grad_norm": 0.7132125965175948,
      "learning_rate": 3.8540713702344635e-07,
      "loss": 0.2553,
      "num_input_tokens_seen": 443084000,
      "step": 7680,
      "train_runtime": 43618.7935,
      "train_tokens_per_second": 10158.098
    },
    {
      "epoch": 1.7497723132969034,
      "grad_norm": 0.5296769814538738,
      "learning_rate": 3.819721842308377e-07,
      "loss": 0.2553,
      "num_input_tokens_seen": 443378920,
      "step": 7685,
      "train_runtime": 43653.2752,
      "train_tokens_per_second": 10156.831
    },
    {
      "epoch": 1.7509107468123861,
      "grad_norm": 0.6254683099469766,
      "learning_rate": 3.785519990929143e-07,
      "loss": 0.2597,
      "num_input_tokens_seen": 443665776,
      "step": 7690,
      "train_runtime": 43686.6039,
      "train_tokens_per_second": 10155.648
    },
    {
      "epoch": 1.7520491803278688,
      "grad_norm": 0.5448226643794144,
      "learning_rate": 3.751465925468356e-07,
      "loss": 0.2281,
      "num_input_tokens_seen": 443957184,
      "step": 7695,
      "train_runtime": 43721.8932,
      "train_tokens_per_second": 10154.116
    },
    {
      "epoch": 1.7531876138433515,
      "grad_norm": 0.594630857142188,
      "learning_rate": 3.7175597548250284e-07,
      "loss": 0.2713,
      "num_input_tokens_seen": 444234416,
      "step": 7700,
      "train_runtime": 43743.7936,
      "train_tokens_per_second": 10155.37
    },
    {
      "epoch": 1.7543260473588342,
      "grad_norm": 0.6319744000706948,
      "learning_rate": 3.683801587425251e-07,
      "loss": 0.2599,
      "num_input_tokens_seen": 444521936,
      "step": 7705,
      "train_runtime": 43766.9879,
      "train_tokens_per_second": 10156.558
    },
    {
      "epoch": 1.755464480874317,
      "grad_norm": 0.6040860297032866,
      "learning_rate": 3.650191531221803e-07,
      "loss": 0.2431,
      "num_input_tokens_seen": 444806608,
      "step": 7710,
      "train_runtime": 43796.4934,
      "train_tokens_per_second": 10156.215
    },
    {
      "epoch": 1.7566029143897997,
      "grad_norm": 0.5953621545219973,
      "learning_rate": 3.6167296936938213e-07,
      "loss": 0.2454,
      "num_input_tokens_seen": 445081808,
      "step": 7715,
      "train_runtime": 43824.6874,
      "train_tokens_per_second": 10155.961
    },
    {
      "epoch": 1.7577413479052824,
      "grad_norm": 0.6760382511133434,
      "learning_rate": 3.5834161818464885e-07,
      "loss": 0.2397,
      "num_input_tokens_seen": 445371192,
      "step": 7720,
      "train_runtime": 43854.1303,
      "train_tokens_per_second": 10155.741
    },
    {
      "epoch": 1.758879781420765,
      "grad_norm": 0.5896013753000227,
      "learning_rate": 3.550251102210661e-07,
      "loss": 0.2476,
      "num_input_tokens_seen": 445645376,
      "step": 7725,
      "train_runtime": 43889.4277,
      "train_tokens_per_second": 10153.821
    },
    {
      "epoch": 1.7600182149362478,
      "grad_norm": 0.6886770489196392,
      "learning_rate": 3.517234560842514e-07,
      "loss": 0.2711,
      "num_input_tokens_seen": 445949568,
      "step": 7730,
      "train_runtime": 43923.3548,
      "train_tokens_per_second": 10152.903
    },
    {
      "epoch": 1.7611566484517303,
      "grad_norm": 0.6141857047621949,
      "learning_rate": 3.4843666633232574e-07,
      "loss": 0.2348,
      "num_input_tokens_seen": 446239800,
      "step": 7735,
      "train_runtime": 43945.2724,
      "train_tokens_per_second": 10154.444
    },
    {
      "epoch": 1.762295081967213,
      "grad_norm": 0.809741521171814,
      "learning_rate": 3.451647514758738e-07,
      "loss": 0.2259,
      "num_input_tokens_seen": 446527464,
      "step": 7740,
      "train_runtime": 43972.9582,
      "train_tokens_per_second": 10154.592
    },
    {
      "epoch": 1.7634335154826957,
      "grad_norm": 0.7492373860227964,
      "learning_rate": 3.419077219779127e-07,
      "loss": 0.2362,
      "num_input_tokens_seen": 446791576,
      "step": 7745,
      "train_runtime": 44002.8056,
      "train_tokens_per_second": 10153.707
    },
    {
      "epoch": 1.7645719489981784,
      "grad_norm": 0.5898892226323419,
      "learning_rate": 3.386655882538614e-07,
      "loss": 0.2478,
      "num_input_tokens_seen": 447092384,
      "step": 7750,
      "train_runtime": 44038.2164,
      "train_tokens_per_second": 10152.373
    },
    {
      "epoch": 1.765710382513661,
      "grad_norm": 0.6222895506359561,
      "learning_rate": 3.354383606715028e-07,
      "loss": 0.2428,
      "num_input_tokens_seen": 447389496,
      "step": 7755,
      "train_runtime": 44061.9951,
      "train_tokens_per_second": 10153.637
    },
    {
      "epoch": 1.7668488160291438,
      "grad_norm": 0.575593303770524,
      "learning_rate": 3.322260495509527e-07,
      "loss": 0.255,
      "num_input_tokens_seen": 447684776,
      "step": 7760,
      "train_runtime": 44089.36,
      "train_tokens_per_second": 10154.032
    },
    {
      "epoch": 1.7679872495446265,
      "grad_norm": 0.5868859726897377,
      "learning_rate": 3.2902866516462626e-07,
      "loss": 0.2348,
      "num_input_tokens_seen": 447990024,
      "step": 7765,
      "train_runtime": 44121.4775,
      "train_tokens_per_second": 10153.559
    },
    {
      "epoch": 1.7691256830601092,
      "grad_norm": 0.5731346865483872,
      "learning_rate": 3.258462177372068e-07,
      "loss": 0.2664,
      "num_input_tokens_seen": 448264152,
      "step": 7770,
      "train_runtime": 44147.3925,
      "train_tokens_per_second": 10153.808
    },
    {
      "epoch": 1.770264116575592,
      "grad_norm": 0.6112403387273122,
      "learning_rate": 3.226787174456103e-07,
      "loss": 0.2573,
      "num_input_tokens_seen": 448563456,
      "step": 7775,
      "train_runtime": 44175.9617,
      "train_tokens_per_second": 10154.017
    },
    {
      "epoch": 1.7714025500910746,
      "grad_norm": 0.788901479188707,
      "learning_rate": 3.1952617441895594e-07,
      "loss": 0.2305,
      "num_input_tokens_seen": 448859176,
      "step": 7780,
      "train_runtime": 44207.2829,
      "train_tokens_per_second": 10153.512
    },
    {
      "epoch": 1.7725409836065573,
      "grad_norm": 0.7218152372237217,
      "learning_rate": 3.1638859873853124e-07,
      "loss": 0.2307,
      "num_input_tokens_seen": 449125720,
      "step": 7785,
      "train_runtime": 44240.8747,
      "train_tokens_per_second": 10151.827
    },
    {
      "epoch": 1.77367941712204,
      "grad_norm": 0.5582492421469119,
      "learning_rate": 3.1326600043776023e-07,
      "loss": 0.2391,
      "num_input_tokens_seen": 449427848,
      "step": 7790,
      "train_runtime": 44273.3116,
      "train_tokens_per_second": 10151.214
    },
    {
      "epoch": 1.7748178506375227,
      "grad_norm": 0.6296323477839733,
      "learning_rate": 3.101583895021731e-07,
      "loss": 0.2433,
      "num_input_tokens_seen": 449713824,
      "step": 7795,
      "train_runtime": 44301.9785,
      "train_tokens_per_second": 10151.1
    },
    {
      "epoch": 1.7759562841530054,
      "grad_norm": 0.6985614628496678,
      "learning_rate": 3.070657758693718e-07,
      "loss": 0.2441,
      "num_input_tokens_seen": 449996712,
      "step": 7800,
      "train_runtime": 44333.9324,
      "train_tokens_per_second": 10150.165
    },
    {
      "epoch": 1.7770947176684881,
      "grad_norm": 0.5650634229653492,
      "learning_rate": 3.039881694290009e-07,
      "loss": 0.2383,
      "num_input_tokens_seen": 450280120,
      "step": 7805,
      "train_runtime": 44359.0371,
      "train_tokens_per_second": 10150.809
    },
    {
      "epoch": 1.7782331511839709,
      "grad_norm": 0.6180372655133335,
      "learning_rate": 3.0092558002271397e-07,
      "loss": 0.2429,
      "num_input_tokens_seen": 450576048,
      "step": 7810,
      "train_runtime": 44384.7414,
      "train_tokens_per_second": 10151.598
    },
    {
      "epoch": 1.7793715846994536,
      "grad_norm": 0.7058337455056695,
      "learning_rate": 2.978780174441415e-07,
      "loss": 0.2529,
      "num_input_tokens_seen": 450864752,
      "step": 7815,
      "train_runtime": 44408.2981,
      "train_tokens_per_second": 10152.714
    },
    {
      "epoch": 1.7805100182149363,
      "grad_norm": 0.5239124274959888,
      "learning_rate": 2.9484549143886267e-07,
      "loss": 0.2221,
      "num_input_tokens_seen": 451165944,
      "step": 7820,
      "train_runtime": 44432.1717,
      "train_tokens_per_second": 10154.038
    },
    {
      "epoch": 1.781648451730419,
      "grad_norm": 0.595005812127358,
      "learning_rate": 2.918280117043709e-07,
      "loss": 0.2456,
      "num_input_tokens_seen": 451467648,
      "step": 7825,
      "train_runtime": 44459.6497,
      "train_tokens_per_second": 10154.548
    },
    {
      "epoch": 1.7827868852459017,
      "grad_norm": 0.6306470519199422,
      "learning_rate": 2.8882558789004423e-07,
      "loss": 0.2378,
      "num_input_tokens_seen": 451753728,
      "step": 7830,
      "train_runtime": 44492.9084,
      "train_tokens_per_second": 10153.387
    },
    {
      "epoch": 1.7839253187613844,
      "grad_norm": 0.6288345186386275,
      "learning_rate": 2.8583822959711683e-07,
      "loss": 0.2279,
      "num_input_tokens_seen": 452043208,
      "step": 7835,
      "train_runtime": 44518.2146,
      "train_tokens_per_second": 10154.118
    },
    {
      "epoch": 1.785063752276867,
      "grad_norm": 0.7662386228332995,
      "learning_rate": 2.8286594637864316e-07,
      "loss": 0.2559,
      "num_input_tokens_seen": 452324776,
      "step": 7840,
      "train_runtime": 44543.6078,
      "train_tokens_per_second": 10154.651
    },
    {
      "epoch": 1.7862021857923498,
      "grad_norm": 0.6076524091850782,
      "learning_rate": 2.7990874773947075e-07,
      "loss": 0.2573,
      "num_input_tokens_seen": 452619136,
      "step": 7845,
      "train_runtime": 44577.9051,
      "train_tokens_per_second": 10153.441
    },
    {
      "epoch": 1.7873406193078325,
      "grad_norm": 0.7311468225104598,
      "learning_rate": 2.769666431362106e-07,
      "loss": 0.2665,
      "num_input_tokens_seen": 452904064,
      "step": 7850,
      "train_runtime": 44605.0746,
      "train_tokens_per_second": 10153.644
    },
    {
      "epoch": 1.7884790528233152,
      "grad_norm": 0.6269443072948218,
      "learning_rate": 2.7403964197720344e-07,
      "loss": 0.2382,
      "num_input_tokens_seen": 453204224,
      "step": 7855,
      "train_runtime": 44635.7388,
      "train_tokens_per_second": 10153.394
    },
    {
      "epoch": 1.789617486338798,
      "grad_norm": 0.6588342121244368,
      "learning_rate": 2.7112775362249433e-07,
      "loss": 0.2473,
      "num_input_tokens_seen": 453517704,
      "step": 7860,
      "train_runtime": 44666.9161,
      "train_tokens_per_second": 10153.325
    },
    {
      "epoch": 1.7907559198542806,
      "grad_norm": 0.6233853575021853,
      "learning_rate": 2.6823098738379793e-07,
      "loss": 0.2712,
      "num_input_tokens_seen": 453808312,
      "step": 7865,
      "train_runtime": 44699.5702,
      "train_tokens_per_second": 10152.409
    },
    {
      "epoch": 1.7918943533697633,
      "grad_norm": 0.6445474006666734,
      "learning_rate": 2.653493525244721e-07,
      "loss": 0.2548,
      "num_input_tokens_seen": 454064776,
      "step": 7870,
      "train_runtime": 44733.0251,
      "train_tokens_per_second": 10150.549
    },
    {
      "epoch": 1.793032786885246,
      "grad_norm": 0.6700321747544038,
      "learning_rate": 2.6248285825948626e-07,
      "loss": 0.2585,
      "num_input_tokens_seen": 454358160,
      "step": 7875,
      "train_runtime": 44764.0306,
      "train_tokens_per_second": 10150.073
    },
    {
      "epoch": 1.7941712204007287,
      "grad_norm": 0.5197945464968109,
      "learning_rate": 2.596315137553956e-07,
      "loss": 0.2123,
      "num_input_tokens_seen": 454634656,
      "step": 7880,
      "train_runtime": 44792.3349,
      "train_tokens_per_second": 10149.832
    },
    {
      "epoch": 1.7953096539162114,
      "grad_norm": 0.631173818068363,
      "learning_rate": 2.567953281303059e-07,
      "loss": 0.2304,
      "num_input_tokens_seen": 454923760,
      "step": 7885,
      "train_runtime": 44819.6178,
      "train_tokens_per_second": 10150.104
    },
    {
      "epoch": 1.7964480874316942,
      "grad_norm": 0.694868216639529,
      "learning_rate": 2.539743104538489e-07,
      "loss": 0.2263,
      "num_input_tokens_seen": 455212592,
      "step": 7890,
      "train_runtime": 44848.1184,
      "train_tokens_per_second": 10150.093
    },
    {
      "epoch": 1.7975865209471766,
      "grad_norm": 0.6123374722174,
      "learning_rate": 2.5116846974715e-07,
      "loss": 0.2295,
      "num_input_tokens_seen": 455504040,
      "step": 7895,
      "train_runtime": 44878.241,
      "train_tokens_per_second": 10149.775
    },
    {
      "epoch": 1.7987249544626593,
      "grad_norm": 0.6839679087146967,
      "learning_rate": 2.4837781498280424e-07,
      "loss": 0.2513,
      "num_input_tokens_seen": 455791120,
      "step": 7900,
      "train_runtime": 44910.3274,
      "train_tokens_per_second": 10148.916
    },
    {
      "epoch": 1.799863387978142,
      "grad_norm": 0.6480986461925203,
      "learning_rate": 2.456023550848424e-07,
      "loss": 0.2644,
      "num_input_tokens_seen": 456065616,
      "step": 7905,
      "train_runtime": 44940.8914,
      "train_tokens_per_second": 10148.121
    },
    {
      "epoch": 1.8010018214936248,
      "grad_norm": 0.6623998484771872,
      "learning_rate": 2.4284209892870557e-07,
      "loss": 0.2576,
      "num_input_tokens_seen": 456332656,
      "step": 7910,
      "train_runtime": 44971.8983,
      "train_tokens_per_second": 10147.062
    },
    {
      "epoch": 1.8021402550091075,
      "grad_norm": 0.5307143581637035,
      "learning_rate": 2.400970553412163e-07,
      "loss": 0.2394,
      "num_input_tokens_seen": 456640048,
      "step": 7915,
      "train_runtime": 44994.516,
      "train_tokens_per_second": 10148.793
    },
    {
      "epoch": 1.8032786885245902,
      "grad_norm": 0.5536032299236313,
      "learning_rate": 2.3736723310054888e-07,
      "loss": 0.2518,
      "num_input_tokens_seen": 456916168,
      "step": 7920,
      "train_runtime": 45025.1704,
      "train_tokens_per_second": 10148.016
    },
    {
      "epoch": 1.8044171220400729,
      "grad_norm": 0.5780155871762948,
      "learning_rate": 2.3465264093620378e-07,
      "loss": 0.2431,
      "num_input_tokens_seen": 457198608,
      "step": 7925,
      "train_runtime": 45054.0989,
      "train_tokens_per_second": 10147.769
    },
    {
      "epoch": 1.8055555555555556,
      "grad_norm": 0.6421943270186443,
      "learning_rate": 2.3195328752897705e-07,
      "loss": 0.2373,
      "num_input_tokens_seen": 457487328,
      "step": 7930,
      "train_runtime": 45089.9098,
      "train_tokens_per_second": 10146.113
    },
    {
      "epoch": 1.8066939890710383,
      "grad_norm": 0.536129237646269,
      "learning_rate": 2.292691815109338e-07,
      "loss": 0.2386,
      "num_input_tokens_seen": 457765664,
      "step": 7935,
      "train_runtime": 45116.5331,
      "train_tokens_per_second": 10146.295
    },
    {
      "epoch": 1.807832422586521,
      "grad_norm": 0.5959016773153697,
      "learning_rate": 2.2660033146538196e-07,
      "loss": 0.2311,
      "num_input_tokens_seen": 458036824,
      "step": 7940,
      "train_runtime": 45146.21,
      "train_tokens_per_second": 10145.632
    },
    {
      "epoch": 1.8089708561020035,
      "grad_norm": 0.6596316185868915,
      "learning_rate": 2.2394674592684184e-07,
      "loss": 0.2455,
      "num_input_tokens_seen": 458322840,
      "step": 7945,
      "train_runtime": 45177.545,
      "train_tokens_per_second": 10144.926
    },
    {
      "epoch": 1.8101092896174862,
      "grad_norm": 0.6498968508408964,
      "learning_rate": 2.213084333810217e-07,
      "loss": 0.2338,
      "num_input_tokens_seen": 458599552,
      "step": 7950,
      "train_runtime": 45205.417,
      "train_tokens_per_second": 10144.792
    },
    {
      "epoch": 1.811247723132969,
      "grad_norm": 0.5559210759288865,
      "learning_rate": 2.186854022647894e-07,
      "loss": 0.2427,
      "num_input_tokens_seen": 458891176,
      "step": 7955,
      "train_runtime": 45232.4199,
      "train_tokens_per_second": 10145.183
    },
    {
      "epoch": 1.8123861566484516,
      "grad_norm": 0.5237076055801724,
      "learning_rate": 2.160776609661436e-07,
      "loss": 0.2402,
      "num_input_tokens_seen": 459185968,
      "step": 7960,
      "train_runtime": 45259.765,
      "train_tokens_per_second": 10145.567
    },
    {
      "epoch": 1.8135245901639343,
      "grad_norm": 0.613859236379917,
      "learning_rate": 2.1348521782419196e-07,
      "loss": 0.272,
      "num_input_tokens_seen": 459469344,
      "step": 7965,
      "train_runtime": 45295.7476,
      "train_tokens_per_second": 10143.763
    },
    {
      "epoch": 1.814663023679417,
      "grad_norm": 0.6386730500067086,
      "learning_rate": 2.1090808112911866e-07,
      "loss": 0.2596,
      "num_input_tokens_seen": 459740408,
      "step": 7970,
      "train_runtime": 45325.6571,
      "train_tokens_per_second": 10143.05
    },
    {
      "epoch": 1.8158014571948997,
      "grad_norm": 0.6019039542131698,
      "learning_rate": 2.0834625912216133e-07,
      "loss": 0.2218,
      "num_input_tokens_seen": 460038544,
      "step": 7975,
      "train_runtime": 45353.5836,
      "train_tokens_per_second": 10143.378
    },
    {
      "epoch": 1.8169398907103824,
      "grad_norm": 0.6365847452289327,
      "learning_rate": 2.057997599955841e-07,
      "loss": 0.2373,
      "num_input_tokens_seen": 460339480,
      "step": 7980,
      "train_runtime": 45377.9967,
      "train_tokens_per_second": 10144.553
    },
    {
      "epoch": 1.8180783242258651,
      "grad_norm": 0.7150106618302905,
      "learning_rate": 2.032685918926508e-07,
      "loss": 0.2432,
      "num_input_tokens_seen": 460635992,
      "step": 7985,
      "train_runtime": 45402.2297,
      "train_tokens_per_second": 10145.669
    },
    {
      "epoch": 1.8192167577413478,
      "grad_norm": 0.684180686215791,
      "learning_rate": 2.007527629075995e-07,
      "loss": 0.2329,
      "num_input_tokens_seen": 460926344,
      "step": 7990,
      "train_runtime": 45438.5201,
      "train_tokens_per_second": 10143.956
    },
    {
      "epoch": 1.8203551912568305,
      "grad_norm": 0.7200158557888217,
      "learning_rate": 1.9825228108561688e-07,
      "loss": 0.2521,
      "num_input_tokens_seen": 461212112,
      "step": 7995,
      "train_runtime": 45471.2186,
      "train_tokens_per_second": 10142.946
    },
    {
      "epoch": 1.8214936247723132,
      "grad_norm": 0.6686633725764196,
      "learning_rate": 1.9576715442281125e-07,
      "loss": 0.2519,
      "num_input_tokens_seen": 461510712,
      "step": 8000,
      "train_runtime": 45492.9107,
      "train_tokens_per_second": 10144.673
    },
    {
      "epoch": 1.822632058287796,
      "grad_norm": 0.6229195548565526,
      "learning_rate": 1.9329739086618783e-07,
      "loss": 0.2458,
      "num_input_tokens_seen": 461802864,
      "step": 8005,
      "train_runtime": 45518.6021,
      "train_tokens_per_second": 10145.366
    },
    {
      "epoch": 1.8237704918032787,
      "grad_norm": 0.5792149286794758,
      "learning_rate": 1.9084299831362396e-07,
      "loss": 0.214,
      "num_input_tokens_seen": 462100768,
      "step": 8010,
      "train_runtime": 45543.0864,
      "train_tokens_per_second": 10146.453
    },
    {
      "epoch": 1.8249089253187614,
      "grad_norm": 0.5866388444918534,
      "learning_rate": 1.8840398461384247e-07,
      "loss": 0.2315,
      "num_input_tokens_seen": 462391912,
      "step": 8015,
      "train_runtime": 45565.7566,
      "train_tokens_per_second": 10147.794
    },
    {
      "epoch": 1.826047358834244,
      "grad_norm": 0.9015502377049001,
      "learning_rate": 1.8598035756638766e-07,
      "loss": 0.2289,
      "num_input_tokens_seen": 462662432,
      "step": 8020,
      "train_runtime": 45591.4121,
      "train_tokens_per_second": 10148.017
    },
    {
      "epoch": 1.8271857923497268,
      "grad_norm": 0.7230913854032843,
      "learning_rate": 1.8357212492159993e-07,
      "loss": 0.2789,
      "num_input_tokens_seen": 462945408,
      "step": 8025,
      "train_runtime": 45623.9987,
      "train_tokens_per_second": 10146.971
    },
    {
      "epoch": 1.8283242258652095,
      "grad_norm": 0.6494371096373169,
      "learning_rate": 1.8117929438059122e-07,
      "loss": 0.2434,
      "num_input_tokens_seen": 463231752,
      "step": 8030,
      "train_runtime": 45653.4864,
      "train_tokens_per_second": 10146.69
    },
    {
      "epoch": 1.8294626593806922,
      "grad_norm": 0.6366677877460561,
      "learning_rate": 1.788018735952196e-07,
      "loss": 0.2335,
      "num_input_tokens_seen": 463512592,
      "step": 8035,
      "train_runtime": 45681.6342,
      "train_tokens_per_second": 10146.585
    },
    {
      "epoch": 1.830601092896175,
      "grad_norm": 0.6861246555240829,
      "learning_rate": 1.764398701680653e-07,
      "loss": 0.2464,
      "num_input_tokens_seen": 463776584,
      "step": 8040,
      "train_runtime": 45711.0959,
      "train_tokens_per_second": 10145.821
    },
    {
      "epoch": 1.8317395264116576,
      "grad_norm": 0.601213719184848,
      "learning_rate": 1.7409329165240796e-07,
      "loss": 0.2426,
      "num_input_tokens_seen": 464073464,
      "step": 8045,
      "train_runtime": 45740.7058,
      "train_tokens_per_second": 10145.743
    },
    {
      "epoch": 1.8328779599271403,
      "grad_norm": 0.6261850703038222,
      "learning_rate": 1.717621455521995e-07,
      "loss": 0.2435,
      "num_input_tokens_seen": 464355616,
      "step": 8050,
      "train_runtime": 45765.6745,
      "train_tokens_per_second": 10146.373
    },
    {
      "epoch": 1.834016393442623,
      "grad_norm": 0.8010216694740472,
      "learning_rate": 1.6944643932204296e-07,
      "loss": 0.2694,
      "num_input_tokens_seen": 464639432,
      "step": 8055,
      "train_runtime": 45790.3969,
      "train_tokens_per_second": 10147.093
    },
    {
      "epoch": 1.8351548269581057,
      "grad_norm": 0.6473100134844367,
      "learning_rate": 1.6714618036716535e-07,
      "loss": 0.2195,
      "num_input_tokens_seen": 464937480,
      "step": 8060,
      "train_runtime": 45815.1994,
      "train_tokens_per_second": 10148.106
    },
    {
      "epoch": 1.8362932604735884,
      "grad_norm": 0.6249962388341578,
      "learning_rate": 1.6486137604339813e-07,
      "loss": 0.2651,
      "num_input_tokens_seen": 465219872,
      "step": 8065,
      "train_runtime": 45844.8821,
      "train_tokens_per_second": 10147.695
    },
    {
      "epoch": 1.8374316939890711,
      "grad_norm": 0.682019397196056,
      "learning_rate": 1.625920336571507e-07,
      "loss": 0.2478,
      "num_input_tokens_seen": 465519752,
      "step": 8070,
      "train_runtime": 45872.7227,
      "train_tokens_per_second": 10148.073
    },
    {
      "epoch": 1.8385701275045538,
      "grad_norm": 0.5797637661778924,
      "learning_rate": 1.603381604653864e-07,
      "loss": 0.2323,
      "num_input_tokens_seen": 465835160,
      "step": 8075,
      "train_runtime": 45905.4537,
      "train_tokens_per_second": 10147.708
    },
    {
      "epoch": 1.8397085610200365,
      "grad_norm": 0.6634145592816685,
      "learning_rate": 1.5809976367560376e-07,
      "loss": 0.218,
      "num_input_tokens_seen": 466095944,
      "step": 8080,
      "train_runtime": 45931.5789,
      "train_tokens_per_second": 10147.614
    },
    {
      "epoch": 1.8408469945355193,
      "grad_norm": 0.8401128156340085,
      "learning_rate": 1.5587685044580747e-07,
      "loss": 0.2525,
      "num_input_tokens_seen": 466398344,
      "step": 8085,
      "train_runtime": 45959.738,
      "train_tokens_per_second": 10147.977
    },
    {
      "epoch": 1.841985428051002,
      "grad_norm": 0.6888184575483771,
      "learning_rate": 1.5366942788449024e-07,
      "loss": 0.2495,
      "num_input_tokens_seen": 466684696,
      "step": 8090,
      "train_runtime": 45989.0004,
      "train_tokens_per_second": 10147.746
    },
    {
      "epoch": 1.8431238615664847,
      "grad_norm": 0.6754072535284613,
      "learning_rate": 1.5147750305060826e-07,
      "loss": 0.2256,
      "num_input_tokens_seen": 466989176,
      "step": 8095,
      "train_runtime": 46015.8851,
      "train_tokens_per_second": 10148.434
    },
    {
      "epoch": 1.8442622950819674,
      "grad_norm": 0.5723474046451744,
      "learning_rate": 1.4930108295355793e-07,
      "loss": 0.2305,
      "num_input_tokens_seen": 467295680,
      "step": 8100,
      "train_runtime": 46049.0689,
      "train_tokens_per_second": 10147.777
    },
    {
      "epoch": 1.8454007285974499,
      "grad_norm": 0.5682633395901954,
      "learning_rate": 1.4714017455315476e-07,
      "loss": 0.2184,
      "num_input_tokens_seen": 467587936,
      "step": 8105,
      "train_runtime": 46078.7911,
      "train_tokens_per_second": 10147.574
    },
    {
      "epoch": 1.8465391621129326,
      "grad_norm": 0.6186481943427812,
      "learning_rate": 1.4499478475960948e-07,
      "loss": 0.2407,
      "num_input_tokens_seen": 467873624,
      "step": 8110,
      "train_runtime": 46109.3364,
      "train_tokens_per_second": 10147.047
    },
    {
      "epoch": 1.8476775956284153,
      "grad_norm": 0.6784194826973339,
      "learning_rate": 1.4286492043350863e-07,
      "loss": 0.2464,
      "num_input_tokens_seen": 468162888,
      "step": 8115,
      "train_runtime": 46139.7672,
      "train_tokens_per_second": 10146.624
    },
    {
      "epoch": 1.848816029143898,
      "grad_norm": 0.6874695378159069,
      "learning_rate": 1.4075058838579013e-07,
      "loss": 0.2781,
      "num_input_tokens_seen": 468428240,
      "step": 8120,
      "train_runtime": 46165.8424,
      "train_tokens_per_second": 10146.641
    },
    {
      "epoch": 1.8499544626593807,
      "grad_norm": 0.7967488512685423,
      "learning_rate": 1.3865179537772112e-07,
      "loss": 0.2598,
      "num_input_tokens_seen": 468719984,
      "step": 8125,
      "train_runtime": 46189.2591,
      "train_tokens_per_second": 10147.813
    },
    {
      "epoch": 1.8510928961748634,
      "grad_norm": 0.5877497728801755,
      "learning_rate": 1.365685481208806e-07,
      "loss": 0.2567,
      "num_input_tokens_seen": 469023016,
      "step": 8130,
      "train_runtime": 46224.2249,
      "train_tokens_per_second": 10146.693
    },
    {
      "epoch": 1.852231329690346,
      "grad_norm": 0.5749558144575252,
      "learning_rate": 1.3450085327713247e-07,
      "loss": 0.2376,
      "num_input_tokens_seen": 469319504,
      "step": 8135,
      "train_runtime": 46251.5589,
      "train_tokens_per_second": 10147.107
    },
    {
      "epoch": 1.8533697632058288,
      "grad_norm": 0.6247389125483838,
      "learning_rate": 1.32448717458607e-07,
      "loss": 0.239,
      "num_input_tokens_seen": 469602480,
      "step": 8140,
      "train_runtime": 46277.0082,
      "train_tokens_per_second": 10147.641
    },
    {
      "epoch": 1.8545081967213115,
      "grad_norm": 0.750060862066216,
      "learning_rate": 1.3041214722768036e-07,
      "loss": 0.2707,
      "num_input_tokens_seen": 469888056,
      "step": 8145,
      "train_runtime": 46301.2056,
      "train_tokens_per_second": 10148.506
    },
    {
      "epoch": 1.8556466302367942,
      "grad_norm": 0.5774433343652381,
      "learning_rate": 1.2839114909695193e-07,
      "loss": 0.232,
      "num_input_tokens_seen": 470156824,
      "step": 8150,
      "train_runtime": 46329.0251,
      "train_tokens_per_second": 10148.213
    },
    {
      "epoch": 1.856785063752277,
      "grad_norm": 0.5474063165358348,
      "learning_rate": 1.2638572952922478e-07,
      "loss": 0.2501,
      "num_input_tokens_seen": 470437024,
      "step": 8155,
      "train_runtime": 46357.8536,
      "train_tokens_per_second": 10147.947
    },
    {
      "epoch": 1.8579234972677594,
      "grad_norm": 0.6643199709778639,
      "learning_rate": 1.243958949374835e-07,
      "loss": 0.2453,
      "num_input_tokens_seen": 470716920,
      "step": 8160,
      "train_runtime": 46390.397,
      "train_tokens_per_second": 10146.861
    },
    {
      "epoch": 1.8590619307832421,
      "grad_norm": 0.7023682170245591,
      "learning_rate": 1.2242165168487475e-07,
      "loss": 0.2645,
      "num_input_tokens_seen": 470969952,
      "step": 8165,
      "train_runtime": 46414.3596,
      "train_tokens_per_second": 10147.074
    },
    {
      "epoch": 1.8602003642987248,
      "grad_norm": 0.5153157389830946,
      "learning_rate": 1.2046300608468787e-07,
      "loss": 0.2304,
      "num_input_tokens_seen": 471244208,
      "step": 8170,
      "train_runtime": 46442.592,
      "train_tokens_per_second": 10146.811
    },
    {
      "epoch": 1.8613387978142075,
      "grad_norm": 0.6555833205574799,
      "learning_rate": 1.185199644003332e-07,
      "loss": 0.2863,
      "num_input_tokens_seen": 471515336,
      "step": 8175,
      "train_runtime": 46470.6393,
      "train_tokens_per_second": 10146.521
    },
    {
      "epoch": 1.8624772313296902,
      "grad_norm": 0.594770800090177,
      "learning_rate": 1.1659253284532157e-07,
      "loss": 0.247,
      "num_input_tokens_seen": 471778912,
      "step": 8180,
      "train_runtime": 46493.5337,
      "train_tokens_per_second": 10147.194
    },
    {
      "epoch": 1.863615664845173,
      "grad_norm": 0.5887038161793695,
      "learning_rate": 1.1468071758324595e-07,
      "loss": 0.2342,
      "num_input_tokens_seen": 472086032,
      "step": 8185,
      "train_runtime": 46524.2301,
      "train_tokens_per_second": 10147.1
    },
    {
      "epoch": 1.8647540983606556,
      "grad_norm": 0.633042781239872,
      "learning_rate": 1.1278452472776146e-07,
      "loss": 0.2338,
      "num_input_tokens_seen": 472357672,
      "step": 8190,
      "train_runtime": 46555.3636,
      "train_tokens_per_second": 10146.149
    },
    {
      "epoch": 1.8658925318761383,
      "grad_norm": 0.6936226131257093,
      "learning_rate": 1.109039603425649e-07,
      "loss": 0.253,
      "num_input_tokens_seen": 472661568,
      "step": 8195,
      "train_runtime": 46583.4295,
      "train_tokens_per_second": 10146.56
    },
    {
      "epoch": 1.867030965391621,
      "grad_norm": 0.6000773108942762,
      "learning_rate": 1.0903903044137576e-07,
      "loss": 0.2234,
      "num_input_tokens_seen": 472946256,
      "step": 8200,
      "train_runtime": 46616.7789,
      "train_tokens_per_second": 10145.408
    },
    {
      "epoch": 1.8681693989071038,
      "grad_norm": 0.5690019455808698,
      "learning_rate": 1.0718974098791857e-07,
      "loss": 0.26,
      "num_input_tokens_seen": 473253800,
      "step": 8205,
      "train_runtime": 46648.7635,
      "train_tokens_per_second": 10145.045
    },
    {
      "epoch": 1.8693078324225865,
      "grad_norm": 0.5428482679155647,
      "learning_rate": 1.0535609789589895e-07,
      "loss": 0.242,
      "num_input_tokens_seen": 473540712,
      "step": 8210,
      "train_runtime": 46679.0602,
      "train_tokens_per_second": 10144.607
    },
    {
      "epoch": 1.8704462659380692,
      "grad_norm": 0.6024900740878395,
      "learning_rate": 1.0353810702899315e-07,
      "loss": 0.247,
      "num_input_tokens_seen": 473836720,
      "step": 8215,
      "train_runtime": 46701.9665,
      "train_tokens_per_second": 10145.969
    },
    {
      "epoch": 1.8715846994535519,
      "grad_norm": 0.5296647976019767,
      "learning_rate": 1.017357742008207e-07,
      "loss": 0.233,
      "num_input_tokens_seen": 474139616,
      "step": 8220,
      "train_runtime": 46728.5909,
      "train_tokens_per_second": 10146.671
    },
    {
      "epoch": 1.8727231329690346,
      "grad_norm": 0.6291681627111153,
      "learning_rate": 9.994910517493017e-08,
      "loss": 0.2585,
      "num_input_tokens_seen": 474430272,
      "step": 8225,
      "train_runtime": 46761.4556,
      "train_tokens_per_second": 10145.755
    },
    {
      "epoch": 1.8738615664845173,
      "grad_norm": 0.6451164659177989,
      "learning_rate": 9.81781056647807e-08,
      "loss": 0.2667,
      "num_input_tokens_seen": 474718456,
      "step": 8230,
      "train_runtime": 46786.6348,
      "train_tokens_per_second": 10146.454
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.6009047065581845,
      "learning_rate": 9.642278133372263e-08,
      "loss": 0.2253,
      "num_input_tokens_seen": 475019264,
      "step": 8235,
      "train_runtime": 46819.8635,
      "train_tokens_per_second": 10145.678
    },
    {
      "epoch": 1.8761384335154827,
      "grad_norm": 0.530321937504643,
      "learning_rate": 9.468313779497917e-08,
      "loss": 0.2291,
      "num_input_tokens_seen": 475316272,
      "step": 8240,
      "train_runtime": 46849.319,
      "train_tokens_per_second": 10145.639
    },
    {
      "epoch": 1.8772768670309654,
      "grad_norm": 0.567567647711045,
      "learning_rate": 9.295918061163034e-08,
      "loss": 0.2347,
      "num_input_tokens_seen": 475609976,
      "step": 8245,
      "train_runtime": 46873.8033,
      "train_tokens_per_second": 10146.605
    },
    {
      "epoch": 1.8784153005464481,
      "grad_norm": 0.7537722322164958,
      "learning_rate": 9.125091529659236e-08,
      "loss": 0.2526,
      "num_input_tokens_seen": 475863440,
      "step": 8250,
      "train_runtime": 46907.4611,
      "train_tokens_per_second": 10144.728
    },
    {
      "epoch": 1.8795537340619308,
      "grad_norm": 0.6910007270485512,
      "learning_rate": 8.955834731260271e-08,
      "loss": 0.2378,
      "num_input_tokens_seen": 476167080,
      "step": 8255,
      "train_runtime": 46930.5076,
      "train_tokens_per_second": 10146.216
    },
    {
      "epoch": 1.8806921675774135,
      "grad_norm": 0.5758221987532405,
      "learning_rate": 8.788148207220182e-08,
      "loss": 0.2472,
      "num_input_tokens_seen": 476456392,
      "step": 8260,
      "train_runtime": 46956.8322,
      "train_tokens_per_second": 10146.689
    },
    {
      "epoch": 1.8818306010928962,
      "grad_norm": 0.6396792629188978,
      "learning_rate": 8.62203249377147e-08,
      "loss": 0.2573,
      "num_input_tokens_seen": 476725016,
      "step": 8265,
      "train_runtime": 46990.2693,
      "train_tokens_per_second": 10145.186
    },
    {
      "epoch": 1.882969034608379,
      "grad_norm": 0.7600961275333307,
      "learning_rate": 8.457488122123436e-08,
      "loss": 0.2424,
      "num_input_tokens_seen": 477009760,
      "step": 8270,
      "train_runtime": 47015.7093,
      "train_tokens_per_second": 10145.753
    },
    {
      "epoch": 1.8841074681238617,
      "grad_norm": 0.7542308775596297,
      "learning_rate": 8.294515618460563e-08,
      "loss": 0.231,
      "num_input_tokens_seen": 477312120,
      "step": 8275,
      "train_runtime": 47047.6265,
      "train_tokens_per_second": 10145.297
    },
    {
      "epoch": 1.8852459016393444,
      "grad_norm": 0.5765941602208563,
      "learning_rate": 8.133115503940692e-08,
      "loss": 0.2422,
      "num_input_tokens_seen": 477618088,
      "step": 8280,
      "train_runtime": 47074.0171,
      "train_tokens_per_second": 10146.109
    },
    {
      "epoch": 1.886384335154827,
      "grad_norm": 0.5032054550395297,
      "learning_rate": 7.97328829469357e-08,
      "loss": 0.2545,
      "num_input_tokens_seen": 477914160,
      "step": 8285,
      "train_runtime": 47101.3131,
      "train_tokens_per_second": 10146.515
    },
    {
      "epoch": 1.8875227686703098,
      "grad_norm": 0.5925086533309163,
      "learning_rate": 7.815034501818974e-08,
      "loss": 0.2305,
      "num_input_tokens_seen": 478188208,
      "step": 8290,
      "train_runtime": 47128.6052,
      "train_tokens_per_second": 10146.454
    },
    {
      "epoch": 1.8886612021857925,
      "grad_norm": 0.691465741004246,
      "learning_rate": 7.658354631385089e-08,
      "loss": 0.2665,
      "num_input_tokens_seen": 478478752,
      "step": 8295,
      "train_runtime": 47157.2154,
      "train_tokens_per_second": 10146.459
    },
    {
      "epoch": 1.8897996357012752,
      "grad_norm": 0.7950445181655589,
      "learning_rate": 7.50324918442713e-08,
      "loss": 0.2471,
      "num_input_tokens_seen": 478766712,
      "step": 8300,
      "train_runtime": 47180.5487,
      "train_tokens_per_second": 10147.544
    },
    {
      "epoch": 1.8909380692167579,
      "grad_norm": 0.5837221673262563,
      "learning_rate": 7.349718656945504e-08,
      "loss": 0.2249,
      "num_input_tokens_seen": 479077768,
      "step": 8305,
      "train_runtime": 47209.8982,
      "train_tokens_per_second": 10147.825
    },
    {
      "epoch": 1.8920765027322406,
      "grad_norm": 0.587813227211132,
      "learning_rate": 7.197763539904312e-08,
      "loss": 0.2539,
      "num_input_tokens_seen": 479366624,
      "step": 8310,
      "train_runtime": 47239.4731,
      "train_tokens_per_second": 10147.586
    },
    {
      "epoch": 1.893214936247723,
      "grad_norm": 0.5860986475403182,
      "learning_rate": 7.047384319229744e-08,
      "loss": 0.2595,
      "num_input_tokens_seen": 479647472,
      "step": 8315,
      "train_runtime": 47261.6181,
      "train_tokens_per_second": 10148.774
    },
    {
      "epoch": 1.8943533697632058,
      "grad_norm": 0.6497413234780836,
      "learning_rate": 6.898581475808463e-08,
      "loss": 0.2419,
      "num_input_tokens_seen": 479952552,
      "step": 8320,
      "train_runtime": 47290.1566,
      "train_tokens_per_second": 10149.101
    },
    {
      "epoch": 1.8954918032786885,
      "grad_norm": 0.67390884141606,
      "learning_rate": 6.751355485486389e-08,
      "loss": 0.2452,
      "num_input_tokens_seen": 480220336,
      "step": 8325,
      "train_runtime": 47320.5773,
      "train_tokens_per_second": 10148.235
    },
    {
      "epoch": 1.8966302367941712,
      "grad_norm": 0.7336843856329908,
      "learning_rate": 6.605706819066693e-08,
      "loss": 0.2418,
      "num_input_tokens_seen": 480495912,
      "step": 8330,
      "train_runtime": 47345.8779,
      "train_tokens_per_second": 10148.632
    },
    {
      "epoch": 1.897768670309654,
      "grad_norm": 0.6161331212709303,
      "learning_rate": 6.461635942308641e-08,
      "loss": 0.2593,
      "num_input_tokens_seen": 480797248,
      "step": 8335,
      "train_runtime": 47372.0125,
      "train_tokens_per_second": 10149.395
    },
    {
      "epoch": 1.8989071038251366,
      "grad_norm": 0.6686656505480035,
      "learning_rate": 6.319143315925924e-08,
      "loss": 0.2529,
      "num_input_tokens_seen": 481101192,
      "step": 8340,
      "train_runtime": 47400.7083,
      "train_tokens_per_second": 10149.663
    },
    {
      "epoch": 1.9000455373406193,
      "grad_norm": 0.5518423541059326,
      "learning_rate": 6.178229395585545e-08,
      "loss": 0.2295,
      "num_input_tokens_seen": 481397296,
      "step": 8345,
      "train_runtime": 47427.7869,
      "train_tokens_per_second": 10150.111
    },
    {
      "epoch": 1.901183970856102,
      "grad_norm": 0.733740502397232,
      "learning_rate": 6.038894631905656e-08,
      "loss": 0.2537,
      "num_input_tokens_seen": 481690696,
      "step": 8350,
      "train_runtime": 47450.0326,
      "train_tokens_per_second": 10151.536
    },
    {
      "epoch": 1.9023224043715847,
      "grad_norm": 0.5670664311642891,
      "learning_rate": 5.901139470454842e-08,
      "loss": 0.2472,
      "num_input_tokens_seen": 481976520,
      "step": 8355,
      "train_runtime": 47481.9843,
      "train_tokens_per_second": 10150.724
    },
    {
      "epoch": 1.9034608378870674,
      "grad_norm": 0.5639526407984988,
      "learning_rate": 5.76496435175028e-08,
      "loss": 0.2561,
      "num_input_tokens_seen": 482281440,
      "step": 8360,
      "train_runtime": 47509.2917,
      "train_tokens_per_second": 10151.308
    },
    {
      "epoch": 1.9045992714025501,
      "grad_norm": 0.7394517472059475,
      "learning_rate": 5.630369711256467e-08,
      "loss": 0.2663,
      "num_input_tokens_seen": 482561080,
      "step": 8365,
      "train_runtime": 47542.6119,
      "train_tokens_per_second": 10150.075
    },
    {
      "epoch": 1.9057377049180326,
      "grad_norm": 0.6248354463889192,
      "learning_rate": 5.4973559793837784e-08,
      "loss": 0.2382,
      "num_input_tokens_seen": 482833160,
      "step": 8370,
      "train_runtime": 47575.9697,
      "train_tokens_per_second": 10148.677
    },
    {
      "epoch": 1.9068761384335153,
      "grad_norm": 0.5360462602694144,
      "learning_rate": 5.365923581487187e-08,
      "loss": 0.2263,
      "num_input_tokens_seen": 483138928,
      "step": 8375,
      "train_runtime": 47597.5151,
      "train_tokens_per_second": 10150.507
    },
    {
      "epoch": 1.908014571948998,
      "grad_norm": 0.5745164340187959,
      "learning_rate": 5.236072937864711e-08,
      "loss": 0.242,
      "num_input_tokens_seen": 483430728,
      "step": 8380,
      "train_runtime": 47620.6183,
      "train_tokens_per_second": 10151.71
    },
    {
      "epoch": 1.9091530054644807,
      "grad_norm": 0.6912197579376506,
      "learning_rate": 5.10780446375625e-08,
      "loss": 0.2563,
      "num_input_tokens_seen": 483692056,
      "step": 8385,
      "train_runtime": 47648.387,
      "train_tokens_per_second": 10151.279
    },
    {
      "epoch": 1.9102914389799635,
      "grad_norm": 0.5827070979990939,
      "learning_rate": 4.981118569342247e-08,
      "loss": 0.2588,
      "num_input_tokens_seen": 483971400,
      "step": 8390,
      "train_runtime": 47671.94,
      "train_tokens_per_second": 10152.123
    },
    {
      "epoch": 1.9114298724954462,
      "grad_norm": 0.6244208564611616,
      "learning_rate": 4.8560156597423636e-08,
      "loss": 0.2307,
      "num_input_tokens_seen": 484270496,
      "step": 8395,
      "train_runtime": 47703.0077,
      "train_tokens_per_second": 10151.781
    },
    {
      "epoch": 1.9125683060109289,
      "grad_norm": 0.6841083024254375,
      "learning_rate": 4.7324961350139756e-08,
      "loss": 0.2713,
      "num_input_tokens_seen": 484547600,
      "step": 8400,
      "train_runtime": 47737.7624,
      "train_tokens_per_second": 10150.195
    },
    {
      "epoch": 1.9137067395264116,
      "grad_norm": 0.5808133950455927,
      "learning_rate": 4.6105603901512305e-08,
      "loss": 0.2169,
      "num_input_tokens_seen": 484832184,
      "step": 8405,
      "train_runtime": 47770.1026,
      "train_tokens_per_second": 10149.281
    },
    {
      "epoch": 1.9148451730418943,
      "grad_norm": 0.8073575767556109,
      "learning_rate": 4.490208815083552e-08,
      "loss": 0.2283,
      "num_input_tokens_seen": 485101808,
      "step": 8410,
      "train_runtime": 47790.6331,
      "train_tokens_per_second": 10150.562
    },
    {
      "epoch": 1.915983606557377,
      "grad_norm": 0.6162348044980177,
      "learning_rate": 4.371441794674414e-08,
      "loss": 0.2483,
      "num_input_tokens_seen": 485389312,
      "step": 8415,
      "train_runtime": 47817.7915,
      "train_tokens_per_second": 10150.81
    },
    {
      "epoch": 1.9171220400728597,
      "grad_norm": 0.6472733421341467,
      "learning_rate": 4.254259708720343e-08,
      "loss": 0.2403,
      "num_input_tokens_seen": 485680792,
      "step": 8420,
      "train_runtime": 47847.9405,
      "train_tokens_per_second": 10150.506
    },
    {
      "epoch": 1.9182604735883424,
      "grad_norm": 0.6342921601828159,
      "learning_rate": 4.138662931949255e-08,
      "loss": 0.2585,
      "num_input_tokens_seen": 485971096,
      "step": 8425,
      "train_runtime": 47881.2514,
      "train_tokens_per_second": 10149.507
    },
    {
      "epoch": 1.919398907103825,
      "grad_norm": 0.5376022244358796,
      "learning_rate": 4.0246518340197325e-08,
      "loss": 0.2355,
      "num_input_tokens_seen": 486271672,
      "step": 8430,
      "train_runtime": 47900.151,
      "train_tokens_per_second": 10151.777
    },
    {
      "epoch": 1.9205373406193078,
      "grad_norm": 0.6098251801491307,
      "learning_rate": 3.912226779519579e-08,
      "loss": 0.2583,
      "num_input_tokens_seen": 486537760,
      "step": 8435,
      "train_runtime": 47924.9071,
      "train_tokens_per_second": 10152.086
    },
    {
      "epoch": 1.9216757741347905,
      "grad_norm": 0.6504160505307843,
      "learning_rate": 3.801388127964656e-08,
      "loss": 0.2605,
      "num_input_tokens_seen": 486801488,
      "step": 8440,
      "train_runtime": 47950.7136,
      "train_tokens_per_second": 10152.122
    },
    {
      "epoch": 1.9228142076502732,
      "grad_norm": 0.6315847840248001,
      "learning_rate": 3.692136233797772e-08,
      "loss": 0.2595,
      "num_input_tokens_seen": 487086528,
      "step": 8445,
      "train_runtime": 47976.3917,
      "train_tokens_per_second": 10152.629
    },
    {
      "epoch": 1.923952641165756,
      "grad_norm": 0.6049081870188802,
      "learning_rate": 3.584471446387683e-08,
      "loss": 0.2291,
      "num_input_tokens_seen": 487386496,
      "step": 8450,
      "train_runtime": 48005.8239,
      "train_tokens_per_second": 10152.654
    },
    {
      "epoch": 1.9250910746812386,
      "grad_norm": 0.6402745088259635,
      "learning_rate": 3.4783941100277054e-08,
      "loss": 0.2324,
      "num_input_tokens_seen": 487677184,
      "step": 8455,
      "train_runtime": 48025.0291,
      "train_tokens_per_second": 10154.646
    },
    {
      "epoch": 1.9262295081967213,
      "grad_norm": 0.5905783943150388,
      "learning_rate": 3.373904563934882e-08,
      "loss": 0.2287,
      "num_input_tokens_seen": 487976824,
      "step": 8460,
      "train_runtime": 48053.4508,
      "train_tokens_per_second": 10154.876
    },
    {
      "epoch": 1.927367941712204,
      "grad_norm": 0.6313662742122779,
      "learning_rate": 3.271003142248652e-08,
      "loss": 0.2831,
      "num_input_tokens_seen": 488272312,
      "step": 8465,
      "train_runtime": 48078.8392,
      "train_tokens_per_second": 10155.659
    },
    {
      "epoch": 1.9285063752276868,
      "grad_norm": 0.6455529640304032,
      "learning_rate": 3.1696901740299603e-08,
      "loss": 0.2599,
      "num_input_tokens_seen": 488568472,
      "step": 8470,
      "train_runtime": 48107.6873,
      "train_tokens_per_second": 10155.726
    },
    {
      "epoch": 1.9296448087431695,
      "grad_norm": 0.6139812086867346,
      "learning_rate": 3.0699659832602594e-08,
      "loss": 0.268,
      "num_input_tokens_seen": 488844600,
      "step": 8475,
      "train_runtime": 48137.7005,
      "train_tokens_per_second": 10155.13
    },
    {
      "epoch": 1.9307832422586522,
      "grad_norm": 0.641102166905012,
      "learning_rate": 2.971830888840177e-08,
      "loss": 0.2549,
      "num_input_tokens_seen": 489105512,
      "step": 8480,
      "train_runtime": 48162.4475,
      "train_tokens_per_second": 10155.329
    },
    {
      "epoch": 1.9319216757741349,
      "grad_norm": 0.5668600768329725,
      "learning_rate": 2.8752852045889067e-08,
      "loss": 0.2356,
      "num_input_tokens_seen": 489403256,
      "step": 8485,
      "train_runtime": 48192.6846,
      "train_tokens_per_second": 10155.136
    },
    {
      "epoch": 1.9330601092896176,
      "grad_norm": 0.5447864762557715,
      "learning_rate": 2.7803292392427074e-08,
      "loss": 0.2402,
      "num_input_tokens_seen": 489711672,
      "step": 8490,
      "train_runtime": 48221.2219,
      "train_tokens_per_second": 10155.522
    },
    {
      "epoch": 1.9341985428051003,
      "grad_norm": 0.5436875927511651,
      "learning_rate": 2.686963296454459e-08,
      "loss": 0.2468,
      "num_input_tokens_seen": 490009720,
      "step": 8495,
      "train_runtime": 48246.8409,
      "train_tokens_per_second": 10156.307
    },
    {
      "epoch": 1.935336976320583,
      "grad_norm": 0.6805041414724653,
      "learning_rate": 2.595187674792221e-08,
      "loss": 0.2499,
      "num_input_tokens_seen": 490309680,
      "step": 8500,
      "train_runtime": 48275.7279,
      "train_tokens_per_second": 10156.443
    },
    {
      "epoch": 1.9364754098360657,
      "grad_norm": 0.7070862389050282,
      "learning_rate": 2.505002667738621e-08,
      "loss": 0.2533,
      "num_input_tokens_seen": 490570920,
      "step": 8505,
      "train_runtime": 48304.7039,
      "train_tokens_per_second": 10155.759
    },
    {
      "epoch": 1.9376138433515484,
      "grad_norm": 0.5621383310729223,
      "learning_rate": 2.4164085636897428e-08,
      "loss": 0.2504,
      "num_input_tokens_seen": 490860808,
      "step": 8510,
      "train_runtime": 48325.9771,
      "train_tokens_per_second": 10157.287
    },
    {
      "epoch": 1.938752276867031,
      "grad_norm": 0.6276844473088278,
      "learning_rate": 2.3294056459541302e-08,
      "loss": 0.2572,
      "num_input_tokens_seen": 491143520,
      "step": 8515,
      "train_runtime": 48356.1947,
      "train_tokens_per_second": 10156.786
    },
    {
      "epoch": 1.9398907103825138,
      "grad_norm": 0.733526781601826,
      "learning_rate": 2.2439941927521746e-08,
      "loss": 0.2397,
      "num_input_tokens_seen": 491430472,
      "step": 8520,
      "train_runtime": 48393.3822,
      "train_tokens_per_second": 10154.911
    },
    {
      "epoch": 1.9410291438979965,
      "grad_norm": 0.5607212739452502,
      "learning_rate": 2.1601744772148936e-08,
      "loss": 0.2192,
      "num_input_tokens_seen": 491729552,
      "step": 8525,
      "train_runtime": 48427.2995,
      "train_tokens_per_second": 10153.974
    },
    {
      "epoch": 1.942167577413479,
      "grad_norm": 0.6051715506732969,
      "learning_rate": 2.07794676738321e-08,
      "loss": 0.2451,
      "num_input_tokens_seen": 492006488,
      "step": 8530,
      "train_runtime": 48458.5037,
      "train_tokens_per_second": 10153.151
    },
    {
      "epoch": 1.9433060109289617,
      "grad_norm": 0.7208777286711683,
      "learning_rate": 1.9973113262072297e-08,
      "loss": 0.2586,
      "num_input_tokens_seen": 492291728,
      "step": 8535,
      "train_runtime": 48491.6233,
      "train_tokens_per_second": 10152.098
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 0.707969397928988,
      "learning_rate": 1.9182684115450212e-08,
      "loss": 0.2638,
      "num_input_tokens_seen": 492586472,
      "step": 8540,
      "train_runtime": 48522.7774,
      "train_tokens_per_second": 10151.655
    },
    {
      "epoch": 1.9455828779599271,
      "grad_norm": 0.7150797926569172,
      "learning_rate": 1.840818276162226e-08,
      "loss": 0.2635,
      "num_input_tokens_seen": 492863400,
      "step": 8545,
      "train_runtime": 48550.9961,
      "train_tokens_per_second": 10151.458
    },
    {
      "epoch": 1.9467213114754098,
      "grad_norm": 0.6960951495509664,
      "learning_rate": 1.764961167731005e-08,
      "loss": 0.2686,
      "num_input_tokens_seen": 493151440,
      "step": 8550,
      "train_runtime": 48577.8436,
      "train_tokens_per_second": 10151.777
    },
    {
      "epoch": 1.9478597449908925,
      "grad_norm": 0.8245393254607356,
      "learning_rate": 1.690697328829316e-08,
      "loss": 0.2521,
      "num_input_tokens_seen": 493441888,
      "step": 8555,
      "train_runtime": 48601.5068,
      "train_tokens_per_second": 10152.81
    },
    {
      "epoch": 1.9489981785063752,
      "grad_norm": 0.6217514429295271,
      "learning_rate": 1.6180269969399697e-08,
      "loss": 0.258,
      "num_input_tokens_seen": 493728936,
      "step": 8560,
      "train_runtime": 48627.1127,
      "train_tokens_per_second": 10153.367
    },
    {
      "epoch": 1.950136612021858,
      "grad_norm": 0.6799505024126781,
      "learning_rate": 1.546950404450187e-08,
      "loss": 0.2732,
      "num_input_tokens_seen": 494030936,
      "step": 8565,
      "train_runtime": 48659.59,
      "train_tokens_per_second": 10152.797
    },
    {
      "epoch": 1.9512750455373407,
      "grad_norm": 0.6328622193424571,
      "learning_rate": 1.4774677786506542e-08,
      "loss": 0.2406,
      "num_input_tokens_seen": 494320504,
      "step": 8570,
      "train_runtime": 48687.6926,
      "train_tokens_per_second": 10152.884
    },
    {
      "epoch": 1.9524134790528234,
      "grad_norm": 0.6074804362292668,
      "learning_rate": 1.409579341734746e-08,
      "loss": 0.2482,
      "num_input_tokens_seen": 494622312,
      "step": 8575,
      "train_runtime": 48715.2581,
      "train_tokens_per_second": 10153.335
    },
    {
      "epoch": 1.9535519125683058,
      "grad_norm": 0.6920313474968062,
      "learning_rate": 1.343285310798026e-08,
      "loss": 0.2733,
      "num_input_tokens_seen": 494896248,
      "step": 8580,
      "train_runtime": 48742.5327,
      "train_tokens_per_second": 10153.273
    },
    {
      "epoch": 1.9546903460837886,
      "grad_norm": 0.6654025453984546,
      "learning_rate": 1.2785858978371923e-08,
      "loss": 0.2635,
      "num_input_tokens_seen": 495173032,
      "step": 8585,
      "train_runtime": 48775.147,
      "train_tokens_per_second": 10152.159
    },
    {
      "epoch": 1.9558287795992713,
      "grad_norm": 0.7809347385189622,
      "learning_rate": 1.2154813097499663e-08,
      "loss": 0.2716,
      "num_input_tokens_seen": 495470512,
      "step": 8590,
      "train_runtime": 48806.8105,
      "train_tokens_per_second": 10151.668
    },
    {
      "epoch": 1.956967213114754,
      "grad_norm": 0.6132957200780976,
      "learning_rate": 1.1539717483337598e-08,
      "loss": 0.2389,
      "num_input_tokens_seen": 495778912,
      "step": 8595,
      "train_runtime": 48844.0182,
      "train_tokens_per_second": 10150.248
    },
    {
      "epoch": 1.9581056466302367,
      "grad_norm": 0.5954046976946952,
      "learning_rate": 1.0940574102856205e-08,
      "loss": 0.2446,
      "num_input_tokens_seen": 496087144,
      "step": 8600,
      "train_runtime": 48876.727,
      "train_tokens_per_second": 10149.762
    },
    {
      "epoch": 1.9592440801457194,
      "grad_norm": 0.5922244032100773,
      "learning_rate": 1.0357384872011767e-08,
      "loss": 0.2542,
      "num_input_tokens_seen": 496381616,
      "step": 8605,
      "train_runtime": 48904.3943,
      "train_tokens_per_second": 10150.041
    },
    {
      "epoch": 1.960382513661202,
      "grad_norm": 0.6007629171778818,
      "learning_rate": 9.79015165574304e-09,
      "loss": 0.2418,
      "num_input_tokens_seen": 496668568,
      "step": 8610,
      "train_runtime": 48936.2978,
      "train_tokens_per_second": 10149.288
    },
    {
      "epoch": 1.9615209471766848,
      "grad_norm": 0.7395296646336116,
      "learning_rate": 9.238876267964047e-09,
      "loss": 0.2527,
      "num_input_tokens_seen": 496940008,
      "step": 8615,
      "train_runtime": 48969.3221,
      "train_tokens_per_second": 10147.986
    },
    {
      "epoch": 1.9626593806921675,
      "grad_norm": 0.7200287910741562,
      "learning_rate": 8.703560471557405e-09,
      "loss": 0.2515,
      "num_input_tokens_seen": 497218760,
      "step": 8620,
      "train_runtime": 48998.5521,
      "train_tokens_per_second": 10147.621
    },
    {
      "epoch": 1.9637978142076502,
      "grad_norm": 0.6859365657231968,
      "learning_rate": 8.184205978370996e-09,
      "loss": 0.2398,
      "num_input_tokens_seen": 497499440,
      "step": 8625,
      "train_runtime": 49033.179,
      "train_tokens_per_second": 10146.18
    },
    {
      "epoch": 1.964936247723133,
      "grad_norm": 0.5414597454373344,
      "learning_rate": 7.680814449211315e-09,
      "loss": 0.2326,
      "num_input_tokens_seen": 497774768,
      "step": 8630,
      "train_runtime": 49058.0586,
      "train_tokens_per_second": 10146.646
    },
    {
      "epoch": 1.9660746812386156,
      "grad_norm": 0.5732762326807557,
      "learning_rate": 7.193387493835691e-09,
      "loss": 0.2361,
      "num_input_tokens_seen": 498071032,
      "step": 8635,
      "train_runtime": 49087.659,
      "train_tokens_per_second": 10146.563
    },
    {
      "epoch": 1.9672131147540983,
      "grad_norm": 0.613076668036432,
      "learning_rate": 6.7219266709528385e-09,
      "loss": 0.2495,
      "num_input_tokens_seen": 498366272,
      "step": 8640,
      "train_runtime": 49115.6836,
      "train_tokens_per_second": 10146.785
    },
    {
      "epoch": 1.968351548269581,
      "grad_norm": 0.6125963235270582,
      "learning_rate": 6.266433488212875e-09,
      "loss": 0.255,
      "num_input_tokens_seen": 498658448,
      "step": 8645,
      "train_runtime": 49150.3454,
      "train_tokens_per_second": 10145.574
    },
    {
      "epoch": 1.9694899817850637,
      "grad_norm": 0.7501111346297946,
      "learning_rate": 5.826909402203429e-09,
      "loss": 0.2303,
      "num_input_tokens_seen": 498928312,
      "step": 8650,
      "train_runtime": 49175.455,
      "train_tokens_per_second": 10145.881
    },
    {
      "epoch": 1.9706284153005464,
      "grad_norm": 0.6828961053208324,
      "learning_rate": 5.40335581844742e-09,
      "loss": 0.2663,
      "num_input_tokens_seen": 499199408,
      "step": 8655,
      "train_runtime": 49206.3466,
      "train_tokens_per_second": 10145.021
    },
    {
      "epoch": 1.9717668488160291,
      "grad_norm": 0.5802089375172034,
      "learning_rate": 4.995774091395289e-09,
      "loss": 0.2389,
      "num_input_tokens_seen": 499483136,
      "step": 8660,
      "train_runtime": 49235.4301,
      "train_tokens_per_second": 10144.791
    },
    {
      "epoch": 1.9729052823315119,
      "grad_norm": 0.6636889488140093,
      "learning_rate": 4.6041655244233315e-09,
      "loss": 0.2622,
      "num_input_tokens_seen": 499765176,
      "step": 8665,
      "train_runtime": 49266.5416,
      "train_tokens_per_second": 10144.109
    },
    {
      "epoch": 1.9740437158469946,
      "grad_norm": 0.6080062596708861,
      "learning_rate": 4.228531369827593e-09,
      "loss": 0.2443,
      "num_input_tokens_seen": 500058840,
      "step": 8670,
      "train_runtime": 49294.2294,
      "train_tokens_per_second": 10144.369
    },
    {
      "epoch": 1.9751821493624773,
      "grad_norm": 0.7895305223746949,
      "learning_rate": 3.868872828822201e-09,
      "loss": 0.2644,
      "num_input_tokens_seen": 500352416,
      "step": 8675,
      "train_runtime": 49323.441,
      "train_tokens_per_second": 10144.313
    },
    {
      "epoch": 1.97632058287796,
      "grad_norm": 0.5482774920522521,
      "learning_rate": 3.525191051532151e-09,
      "loss": 0.2492,
      "num_input_tokens_seen": 500633288,
      "step": 8680,
      "train_runtime": 49350.9174,
      "train_tokens_per_second": 10144.356
    },
    {
      "epoch": 1.9774590163934427,
      "grad_norm": 0.6703003458180462,
      "learning_rate": 3.1974871369921947e-09,
      "loss": 0.2629,
      "num_input_tokens_seen": 500921464,
      "step": 8685,
      "train_runtime": 49378.6399,
      "train_tokens_per_second": 10144.497
    },
    {
      "epoch": 1.9785974499089254,
      "grad_norm": 0.6636996105333903,
      "learning_rate": 2.8857621331435104e-09,
      "loss": 0.2603,
      "num_input_tokens_seen": 501202928,
      "step": 8690,
      "train_runtime": 49405.264,
      "train_tokens_per_second": 10144.727
    },
    {
      "epoch": 1.979735883424408,
      "grad_norm": 0.6876728249314774,
      "learning_rate": 2.5900170368281517e-09,
      "loss": 0.2486,
      "num_input_tokens_seen": 501502616,
      "step": 8695,
      "train_runtime": 49438.224,
      "train_tokens_per_second": 10144.026
    },
    {
      "epoch": 1.9808743169398908,
      "grad_norm": 0.5849834384230381,
      "learning_rate": 2.310252793787937e-09,
      "loss": 0.2318,
      "num_input_tokens_seen": 501808240,
      "step": 8700,
      "train_runtime": 49465.1564,
      "train_tokens_per_second": 10144.681
    },
    {
      "epoch": 1.9820127504553735,
      "grad_norm": 0.6709400024300086,
      "learning_rate": 2.0464702986611184e-09,
      "loss": 0.2384,
      "num_input_tokens_seen": 502102448,
      "step": 8705,
      "train_runtime": 49493.3628,
      "train_tokens_per_second": 10144.844
    },
    {
      "epoch": 1.9831511839708562,
      "grad_norm": 0.6251432102634737,
      "learning_rate": 1.7986703949779415e-09,
      "loss": 0.2502,
      "num_input_tokens_seen": 502372416,
      "step": 8710,
      "train_runtime": 49520.2411,
      "train_tokens_per_second": 10144.789
    },
    {
      "epoch": 1.984289617486339,
      "grad_norm": 0.7771857169157811,
      "learning_rate": 1.5668538751600903e-09,
      "loss": 0.2401,
      "num_input_tokens_seen": 502657816,
      "step": 8715,
      "train_runtime": 49553.7542,
      "train_tokens_per_second": 10143.688
    },
    {
      "epoch": 1.9854280510018216,
      "grad_norm": 0.6183163096036755,
      "learning_rate": 1.3510214805156908e-09,
      "loss": 0.2534,
      "num_input_tokens_seen": 502965152,
      "step": 8720,
      "train_runtime": 49588.4944,
      "train_tokens_per_second": 10142.779
    },
    {
      "epoch": 1.9865664845173043,
      "grad_norm": 0.6974396961128795,
      "learning_rate": 1.1511739012409761e-09,
      "loss": 0.2578,
      "num_input_tokens_seen": 503232808,
      "step": 8725,
      "train_runtime": 49616.883,
      "train_tokens_per_second": 10142.37
    },
    {
      "epoch": 1.987704918032787,
      "grad_norm": 0.5482389959972185,
      "learning_rate": 9.673117764125161e-10,
      "loss": 0.2356,
      "num_input_tokens_seen": 503517280,
      "step": 8730,
      "train_runtime": 49640.8362,
      "train_tokens_per_second": 10143.207
    },
    {
      "epoch": 1.9888433515482697,
      "grad_norm": 0.6810748208914952,
      "learning_rate": 7.994356939911019e-10,
      "loss": 0.2705,
      "num_input_tokens_seen": 503814688,
      "step": 8735,
      "train_runtime": 49673.2149,
      "train_tokens_per_second": 10142.583
    },
    {
      "epoch": 1.9899817850637522,
      "grad_norm": 0.64721137852851,
      "learning_rate": 6.475461908139747e-10,
      "loss": 0.2484,
      "num_input_tokens_seen": 504095408,
      "step": 8740,
      "train_runtime": 49698.3509,
      "train_tokens_per_second": 10143.101
    },
    {
      "epoch": 1.991120218579235,
      "grad_norm": 0.6951901511272359,
      "learning_rate": 5.116437525981566e-10,
      "loss": 0.2437,
      "num_input_tokens_seen": 504392720,
      "step": 8745,
      "train_runtime": 49723.609,
      "train_tokens_per_second": 10143.928
    },
    {
      "epoch": 1.9922586520947176,
      "grad_norm": 0.5996960972363836,
      "learning_rate": 3.917288139365649e-10,
      "loss": 0.2324,
      "num_input_tokens_seen": 504692424,
      "step": 8750,
      "train_runtime": 49752.2636,
      "train_tokens_per_second": 10144.11
    },
    {
      "epoch": 1.9933970856102003,
      "grad_norm": 0.721213806348224,
      "learning_rate": 2.878017582952364e-10,
      "loss": 0.2657,
      "num_input_tokens_seen": 504995736,
      "step": 8755,
      "train_runtime": 49779.0217,
      "train_tokens_per_second": 10144.75
    },
    {
      "epoch": 1.994535519125683,
      "grad_norm": 0.6405767991988569,
      "learning_rate": 1.9986291801610269e-10,
      "loss": 0.2629,
      "num_input_tokens_seen": 505272784,
      "step": 8760,
      "train_runtime": 49801.561,
      "train_tokens_per_second": 10145.722
    },
    {
      "epoch": 1.9956739526411658,
      "grad_norm": 0.6141426626073588,
      "learning_rate": 1.2791257431199467e-10,
      "loss": 0.2307,
      "num_input_tokens_seen": 505564504,
      "step": 8765,
      "train_runtime": 49832.8891,
      "train_tokens_per_second": 10145.198
    },
    {
      "epoch": 1.9968123861566485,
      "grad_norm": 0.6652078387883094,
      "learning_rate": 7.195095726775236e-11,
      "loss": 0.2311,
      "num_input_tokens_seen": 505840304,
      "step": 8770,
      "train_runtime": 49858.2513,
      "train_tokens_per_second": 10145.568
    },
    {
      "epoch": 1.9979508196721312,
      "grad_norm": 0.5586388334395597,
      "learning_rate": 3.197824583911491e-11,
      "loss": 0.2509,
      "num_input_tokens_seen": 506119448,
      "step": 8775,
      "train_runtime": 49884.492,
      "train_tokens_per_second": 10145.827
    },
    {
      "epoch": 1.9990892531876139,
      "grad_norm": 0.6324401007382205,
      "learning_rate": 7.994567851055124e-12,
      "loss": 0.2375,
      "num_input_tokens_seen": 506403488,
      "step": 8780,
      "train_runtime": 49914.9717,
      "train_tokens_per_second": 10145.323
    },
    {
      "epoch": 2.0,
      "num_input_tokens_seen": 506650592,
      "step": 8784,
      "total_flos": 936971232313344.0,
      "train_loss": 0.2898640426890229,
      "train_runtime": 50047.8007,
      "train_samples_per_second": 0.702,
      "train_steps_per_second": 0.176
    }
  ],
  "logging_steps": 5,
  "max_steps": 8784,
  "num_input_tokens_seen": 506650592,
  "num_train_epochs": 2,
  "save_steps": 0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 936971232313344.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}