{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9878213802435725,
  "eval_steps": 500,
  "global_step": 276,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010825439783491205,
      "grad_norm": 5.8079091029435865,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 0.8567,
      "step": 1
    },
    {
      "epoch": 0.02165087956698241,
      "grad_norm": 5.811672220394839,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.8604,
      "step": 2
    },
    {
      "epoch": 0.03247631935047361,
      "grad_norm": 5.449507701601476,
      "learning_rate": 8.571428571428571e-06,
      "loss": 0.8528,
      "step": 3
    },
    {
      "epoch": 0.04330175913396482,
      "grad_norm": 2.3063006110984086,
      "learning_rate": 1.1428571428571429e-05,
      "loss": 0.776,
      "step": 4
    },
    {
      "epoch": 0.05412719891745602,
      "grad_norm": 3.9302521280440543,
      "learning_rate": 1.4285714285714287e-05,
      "loss": 0.7546,
      "step": 5
    },
    {
      "epoch": 0.06495263870094722,
      "grad_norm": 4.207181909155043,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 0.751,
      "step": 6
    },
    {
      "epoch": 0.07577807848443843,
      "grad_norm": 4.360918963292297,
      "learning_rate": 2e-05,
      "loss": 0.7344,
      "step": 7
    },
    {
      "epoch": 0.08660351826792964,
      "grad_norm": 2.911250822571266,
      "learning_rate": 2.2857142857142858e-05,
      "loss": 0.729,
      "step": 8
    },
    {
      "epoch": 0.09742895805142084,
      "grad_norm": 2.7455493318657966,
      "learning_rate": 2.5714285714285718e-05,
      "loss": 0.672,
      "step": 9
    },
    {
      "epoch": 0.10825439783491204,
      "grad_norm": 2.1241241512314795,
      "learning_rate": 2.8571428571428574e-05,
      "loss": 0.6659,
      "step": 10
    },
    {
      "epoch": 0.11907983761840325,
      "grad_norm": 1.5668463996929993,
      "learning_rate": 3.142857142857143e-05,
      "loss": 0.6379,
      "step": 11
    },
    {
      "epoch": 0.12990527740189445,
      "grad_norm": 1.3619379997378058,
      "learning_rate": 3.4285714285714284e-05,
      "loss": 0.6305,
      "step": 12
    },
    {
      "epoch": 0.14073071718538566,
      "grad_norm": 1.566918593012661,
      "learning_rate": 3.714285714285715e-05,
      "loss": 0.6142,
      "step": 13
    },
    {
      "epoch": 0.15155615696887687,
      "grad_norm": 1.2044004334843796,
      "learning_rate": 4e-05,
      "loss": 0.6246,
      "step": 14
    },
    {
      "epoch": 0.16238159675236807,
      "grad_norm": 1.4087070173540086,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.6075,
      "step": 15
    },
    {
      "epoch": 0.17320703653585928,
      "grad_norm": 1.203436631182031,
      "learning_rate": 4.5714285714285716e-05,
      "loss": 0.5986,
      "step": 16
    },
    {
      "epoch": 0.18403247631935046,
      "grad_norm": 1.722190196328099,
      "learning_rate": 4.857142857142857e-05,
      "loss": 0.5943,
      "step": 17
    },
    {
      "epoch": 0.19485791610284167,
      "grad_norm": 1.1305378018180656,
      "learning_rate": 5.1428571428571436e-05,
      "loss": 0.577,
      "step": 18
    },
    {
      "epoch": 0.20568335588633288,
      "grad_norm": 1.6031033848905385,
      "learning_rate": 5.4285714285714295e-05,
      "loss": 0.5697,
      "step": 19
    },
    {
      "epoch": 0.2165087956698241,
      "grad_norm": 1.6591800967410824,
      "learning_rate": 5.714285714285715e-05,
      "loss": 0.5746,
      "step": 20
    },
    {
      "epoch": 0.2273342354533153,
      "grad_norm": 1.5025149200657286,
      "learning_rate": 6.000000000000001e-05,
      "loss": 0.5683,
      "step": 21
    },
    {
      "epoch": 0.2381596752368065,
      "grad_norm": 1.2019447493505488,
      "learning_rate": 6.285714285714286e-05,
      "loss": 0.5676,
      "step": 22
    },
    {
      "epoch": 0.2489851150202977,
      "grad_norm": 1.9839452102168298,
      "learning_rate": 6.571428571428571e-05,
      "loss": 0.5639,
      "step": 23
    },
    {
      "epoch": 0.2598105548037889,
      "grad_norm": 1.34504393296771,
      "learning_rate": 6.857142857142857e-05,
      "loss": 0.5667,
      "step": 24
    },
    {
      "epoch": 0.2706359945872801,
      "grad_norm": 2.227974826889621,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.5639,
      "step": 25
    },
    {
      "epoch": 0.2814614343707713,
      "grad_norm": 1.6397454252709596,
      "learning_rate": 7.42857142857143e-05,
      "loss": 0.5434,
      "step": 26
    },
    {
      "epoch": 0.2922868741542625,
      "grad_norm": 1.7236920014146049,
      "learning_rate": 7.714285714285715e-05,
      "loss": 0.5453,
      "step": 27
    },
    {
      "epoch": 0.30311231393775373,
      "grad_norm": 1.3679886100442236,
      "learning_rate": 8e-05,
      "loss": 0.5518,
      "step": 28
    },
    {
      "epoch": 0.31393775372124494,
      "grad_norm": 1.7636609095058786,
      "learning_rate": 7.999679062421315e-05,
      "loss": 0.5488,
      "step": 29
    },
    {
      "epoch": 0.32476319350473615,
      "grad_norm": 1.8909077915573775,
      "learning_rate": 7.998716301185722e-05,
      "loss": 0.5434,
      "step": 30
    },
    {
      "epoch": 0.33558863328822736,
      "grad_norm": 1.4831299671419245,
      "learning_rate": 7.997111870786354e-05,
      "loss": 0.543,
      "step": 31
    },
    {
      "epoch": 0.34641407307171856,
      "grad_norm": 1.5685011400813527,
      "learning_rate": 7.994866028684212e-05,
      "loss": 0.5386,
      "step": 32
    },
    {
      "epoch": 0.3572395128552097,
      "grad_norm": 1.8163117507870878,
      "learning_rate": 7.991979135266861e-05,
      "loss": 0.5426,
      "step": 33
    },
    {
      "epoch": 0.3680649526387009,
      "grad_norm": 1.3553007909847135,
      "learning_rate": 7.98845165379059e-05,
      "loss": 0.5419,
      "step": 34
    },
    {
      "epoch": 0.37889039242219213,
      "grad_norm": 1.5002520273898525,
      "learning_rate": 7.984284150306085e-05,
      "loss": 0.525,
      "step": 35
    },
    {
      "epoch": 0.38971583220568334,
      "grad_norm": 1.5788765506619407,
      "learning_rate": 7.97947729356758e-05,
      "loss": 0.5161,
      "step": 36
    },
    {
      "epoch": 0.40054127198917455,
      "grad_norm": 1.0932865412546702,
      "learning_rate": 7.974031854925562e-05,
      "loss": 0.5241,
      "step": 37
    },
    {
      "epoch": 0.41136671177266576,
      "grad_norm": 1.1169446884746776,
      "learning_rate": 7.967948708202972e-05,
      "loss": 0.5135,
      "step": 38
    },
    {
      "epoch": 0.42219215155615697,
      "grad_norm": 1.7368905888186483,
      "learning_rate": 7.961228829555003e-05,
      "loss": 0.5232,
      "step": 39
    },
    {
      "epoch": 0.4330175913396482,
      "grad_norm": 1.14825749468684,
      "learning_rate": 7.953873297312447e-05,
      "loss": 0.5114,
      "step": 40
    },
    {
      "epoch": 0.4438430311231394,
      "grad_norm": 1.681948559743569,
      "learning_rate": 7.945883291808655e-05,
      "loss": 0.5151,
      "step": 41
    },
    {
      "epoch": 0.4546684709066306,
      "grad_norm": 1.1384144753190577,
      "learning_rate": 7.937260095190137e-05,
      "loss": 0.5079,
      "step": 42
    },
    {
      "epoch": 0.4654939106901218,
      "grad_norm": 1.6677218265436802,
      "learning_rate": 7.928005091210817e-05,
      "loss": 0.5128,
      "step": 43
    },
    {
      "epoch": 0.476319350473613,
      "grad_norm": 1.4237449938878564,
      "learning_rate": 7.918119765009979e-05,
      "loss": 0.511,
      "step": 44
    },
    {
      "epoch": 0.4871447902571042,
      "grad_norm": 1.3134667589020321,
      "learning_rate": 7.907605702873948e-05,
      "loss": 0.5101,
      "step": 45
    },
    {
      "epoch": 0.4979702300405954,
      "grad_norm": 1.0853939092704676,
      "learning_rate": 7.896464591981549e-05,
      "loss": 0.5086,
      "step": 46
    },
    {
      "epoch": 0.5087956698240866,
      "grad_norm": 1.6016710689451,
      "learning_rate": 7.884698220133357e-05,
      "loss": 0.5163,
      "step": 47
    },
    {
      "epoch": 0.5196211096075778,
      "grad_norm": 1.203978171530331,
      "learning_rate": 7.872308475464818e-05,
      "loss": 0.5086,
      "step": 48
    },
    {
      "epoch": 0.530446549391069,
      "grad_norm": 1.3897099566312143,
      "learning_rate": 7.859297346143258e-05,
      "loss": 0.5051,
      "step": 49
    },
    {
      "epoch": 0.5412719891745602,
      "grad_norm": 1.2903904022789845,
      "learning_rate": 7.84566692004885e-05,
      "loss": 0.5036,
      "step": 50
    },
    {
      "epoch": 0.5520974289580515,
      "grad_norm": 0.9821064573534952,
      "learning_rate": 7.831419384439565e-05,
      "loss": 0.4979,
      "step": 51
    },
    {
      "epoch": 0.5629228687415426,
      "grad_norm": 1.1967263169199975,
      "learning_rate": 7.816557025600196e-05,
      "loss": 0.4954,
      "step": 52
    },
    {
      "epoch": 0.5737483085250338,
      "grad_norm": 1.5118785206523027,
      "learning_rate": 7.80108222847547e-05,
      "loss": 0.4935,
      "step": 53
    },
    {
      "epoch": 0.584573748308525,
      "grad_norm": 1.328515593795635,
      "learning_rate": 7.784997476287349e-05,
      "loss": 0.5029,
      "step": 54
    },
    {
      "epoch": 0.5953991880920162,
      "grad_norm": 1.085170937371965,
      "learning_rate": 7.76830535013654e-05,
      "loss": 0.4922,
      "step": 55
    },
    {
      "epoch": 0.6062246278755075,
      "grad_norm": 1.1543845766613074,
      "learning_rate": 7.751008528588322e-05,
      "loss": 0.4954,
      "step": 56
    },
    {
      "epoch": 0.6170500676589986,
      "grad_norm": 1.2151109406542857,
      "learning_rate": 7.733109787242708e-05,
      "loss": 0.4981,
      "step": 57
    },
    {
      "epoch": 0.6278755074424899,
      "grad_norm": 1.6809116509174178,
      "learning_rate": 7.71461199828905e-05,
      "loss": 0.4992,
      "step": 58
    },
    {
      "epoch": 0.638700947225981,
      "grad_norm": 1.1929403575426991,
      "learning_rate": 7.695518130045147e-05,
      "loss": 0.4804,
      "step": 59
    },
    {
      "epoch": 0.6495263870094723,
      "grad_norm": 1.3373806033082491,
      "learning_rate": 7.675831246480923e-05,
      "loss": 0.4918,
      "step": 60
    },
    {
      "epoch": 0.6603518267929634,
      "grad_norm": 1.1370461825495959,
      "learning_rate": 7.655554506726747e-05,
      "loss": 0.5003,
      "step": 61
    },
    {
      "epoch": 0.6711772665764547,
      "grad_norm": 1.5541020268182568,
      "learning_rate": 7.6346911645665e-05,
      "loss": 0.4872,
      "step": 62
    },
    {
      "epoch": 0.6820027063599459,
      "grad_norm": 1.0196032384903368,
      "learning_rate": 7.61324456791544e-05,
      "loss": 0.4971,
      "step": 63
    },
    {
      "epoch": 0.6928281461434371,
      "grad_norm": 1.355258802840393,
      "learning_rate": 7.591218158282968e-05,
      "loss": 0.4961,
      "step": 64
    },
    {
      "epoch": 0.7036535859269283,
      "grad_norm": 1.5317006352800375,
      "learning_rate": 7.568615470220369e-05,
      "loss": 0.4836,
      "step": 65
    },
    {
      "epoch": 0.7144790257104194,
      "grad_norm": 0.9602544899213347,
      "learning_rate": 7.545440130753634e-05,
      "loss": 0.4934,
      "step": 66
    },
    {
      "epoch": 0.7253044654939107,
      "grad_norm": 1.3937893813198252,
      "learning_rate": 7.52169585880143e-05,
      "loss": 0.492,
      "step": 67
    },
    {
      "epoch": 0.7361299052774019,
      "grad_norm": 1.2170568243121482,
      "learning_rate": 7.497386464578329e-05,
      "loss": 0.5004,
      "step": 68
    },
    {
      "epoch": 0.7469553450608931,
      "grad_norm": 1.215088332835355,
      "learning_rate": 7.472515848983394e-05,
      "loss": 0.4927,
      "step": 69
    },
    {
      "epoch": 0.7577807848443843,
      "grad_norm": 1.068568427145697,
      "learning_rate": 7.447088002974199e-05,
      "loss": 0.4735,
      "step": 70
    },
    {
      "epoch": 0.7686062246278755,
      "grad_norm": 1.2666434730845704,
      "learning_rate": 7.421107006926408e-05,
      "loss": 0.4776,
      "step": 71
    },
    {
      "epoch": 0.7794316644113667,
      "grad_norm": 1.0693035758098317,
      "learning_rate": 7.394577029979004e-05,
      "loss": 0.4878,
      "step": 72
    },
    {
      "epoch": 0.790257104194858,
      "grad_norm": 1.5502809637898465,
      "learning_rate": 7.367502329365268e-05,
      "loss": 0.4949,
      "step": 73
    },
    {
      "epoch": 0.8010825439783491,
      "grad_norm": 0.8873154529201425,
      "learning_rate": 7.33988724972963e-05,
      "loss": 0.4889,
      "step": 74
    },
    {
      "epoch": 0.8119079837618404,
      "grad_norm": 1.27058950939853,
      "learning_rate": 7.311736222430487e-05,
      "loss": 0.4861,
      "step": 75
    },
    {
      "epoch": 0.8227334235453315,
      "grad_norm": 0.7866888980990746,
      "learning_rate": 7.283053764829106e-05,
      "loss": 0.4783,
      "step": 76
    },
    {
      "epoch": 0.8335588633288228,
      "grad_norm": 1.0938731157860102,
      "learning_rate": 7.253844479564737e-05,
      "loss": 0.4764,
      "step": 77
    },
    {
      "epoch": 0.8443843031123139,
      "grad_norm": 1.1807281415252031,
      "learning_rate": 7.224113053816021e-05,
      "loss": 0.4817,
      "step": 78
    },
    {
      "epoch": 0.8552097428958051,
      "grad_norm": 1.322179202967281,
      "learning_rate": 7.193864258548855e-05,
      "loss": 0.4853,
      "step": 79
    },
    {
      "epoch": 0.8660351826792964,
      "grad_norm": 1.2155369734676524,
      "learning_rate": 7.163102947750794e-05,
      "loss": 0.4796,
      "step": 80
    },
    {
      "epoch": 0.8768606224627875,
      "grad_norm": 0.588289504862737,
      "learning_rate": 7.131834057652142e-05,
      "loss": 0.4863,
      "step": 81
    },
    {
      "epoch": 0.8876860622462788,
      "grad_norm": 1.0985408806535708,
      "learning_rate": 7.100062605933835e-05,
      "loss": 0.4762,
      "step": 82
    },
    {
      "epoch": 0.8985115020297699,
      "grad_norm": 1.74082517425262,
      "learning_rate": 7.067793690922268e-05,
      "loss": 0.4853,
      "step": 83
    },
    {
      "epoch": 0.9093369418132612,
      "grad_norm": 0.7638072853814706,
      "learning_rate": 7.035032490771165e-05,
      "loss": 0.4794,
      "step": 84
    },
    {
      "epoch": 0.9201623815967523,
      "grad_norm": 1.2164285440187212,
      "learning_rate": 7.001784262630652e-05,
      "loss": 0.48,
      "step": 85
    },
    {
      "epoch": 0.9309878213802436,
      "grad_norm": 1.2499859479078839,
      "learning_rate": 6.968054341803644e-05,
      "loss": 0.4853,
      "step": 86
    },
    {
      "epoch": 0.9418132611637348,
      "grad_norm": 1.1495652891036365,
      "learning_rate": 6.933848140889705e-05,
      "loss": 0.476,
      "step": 87
    },
    {
      "epoch": 0.952638700947226,
      "grad_norm": 1.1590588610177872,
      "learning_rate": 6.89917114891648e-05,
      "loss": 0.4653,
      "step": 88
    },
    {
      "epoch": 0.9634641407307172,
      "grad_norm": 0.8043870196059654,
      "learning_rate": 6.864028930458892e-05,
      "loss": 0.4788,
      "step": 89
    },
    {
      "epoch": 0.9742895805142084,
      "grad_norm": 0.8931105851691834,
      "learning_rate": 6.828427124746191e-05,
      "loss": 0.4776,
      "step": 90
    },
    {
      "epoch": 0.9851150202976996,
      "grad_norm": 1.2474558559687694,
      "learning_rate": 6.792371444757037e-05,
      "loss": 0.4803,
      "step": 91
    },
    {
      "epoch": 0.9959404600811907,
      "grad_norm": 0.8560601402041761,
      "learning_rate": 6.755867676302747e-05,
      "loss": 0.4798,
      "step": 92
    },
    {
      "epoch": 1.006765899864682,
      "grad_norm": 2.0148609986627783,
      "learning_rate": 6.718921677098853e-05,
      "loss": 0.7491,
      "step": 93
    },
    {
      "epoch": 1.0175913396481733,
      "grad_norm": 0.6667855075813555,
      "learning_rate": 6.681539375825115e-05,
      "loss": 0.4535,
      "step": 94
    },
    {
      "epoch": 1.0284167794316643,
      "grad_norm": 0.9162841549268478,
      "learning_rate": 6.643726771174164e-05,
      "loss": 0.455,
      "step": 95
    },
    {
      "epoch": 1.0392422192151556,
      "grad_norm": 1.0155165692532704,
      "learning_rate": 6.60548993088889e-05,
      "loss": 0.454,
      "step": 96
    },
    {
      "epoch": 1.0500676589986468,
      "grad_norm": 1.078599089633044,
      "learning_rate": 6.56683499078876e-05,
      "loss": 0.4455,
      "step": 97
    },
    {
      "epoch": 1.060893098782138,
      "grad_norm": 1.1351986372223781,
      "learning_rate": 6.527768153785216e-05,
      "loss": 0.4541,
      "step": 98
    },
    {
      "epoch": 1.0717185385656292,
      "grad_norm": 0.7315670031697065,
      "learning_rate": 6.488295688886295e-05,
      "loss": 0.4479,
      "step": 99
    },
    {
      "epoch": 1.0825439783491204,
      "grad_norm": 1.3582463875830837,
      "learning_rate": 6.448423930190653e-05,
      "loss": 0.4486,
      "step": 100
    },
    {
      "epoch": 1.0933694181326117,
      "grad_norm": 0.8080345177396737,
      "learning_rate": 6.408159275871132e-05,
      "loss": 0.446,
      "step": 101
    },
    {
      "epoch": 1.104194857916103,
      "grad_norm": 1.160068769869394,
      "learning_rate": 6.36750818714807e-05,
      "loss": 0.4471,
      "step": 102
    },
    {
      "epoch": 1.115020297699594,
      "grad_norm": 0.8897562278529675,
      "learning_rate": 6.326477187252455e-05,
      "loss": 0.449,
      "step": 103
    },
    {
      "epoch": 1.1258457374830853,
      "grad_norm": 1.0199996165037868,
      "learning_rate": 6.28507286037917e-05,
      "loss": 0.4427,
      "step": 104
    },
    {
      "epoch": 1.1366711772665765,
      "grad_norm": 0.7249873315321067,
      "learning_rate": 6.243301850630419e-05,
      "loss": 0.4453,
      "step": 105
    },
    {
      "epoch": 1.1474966170500678,
      "grad_norm": 1.189445348682187,
      "learning_rate": 6.201170860949565e-05,
      "loss": 0.4515,
      "step": 106
    },
    {
      "epoch": 1.1583220568335588,
      "grad_norm": 1.1424188944548885,
      "learning_rate": 6.15868665204552e-05,
      "loss": 0.4452,
      "step": 107
    },
    {
      "epoch": 1.16914749661705,
      "grad_norm": 0.8857107534333528,
      "learning_rate": 6.11585604130785e-05,
      "loss": 0.4393,
      "step": 108
    },
    {
      "epoch": 1.1799729364005414,
      "grad_norm": 0.7495223764013124,
      "learning_rate": 6.072685901712808e-05,
      "loss": 0.4446,
      "step": 109
    },
    {
      "epoch": 1.1907983761840324,
      "grad_norm": 0.6658761551232241,
      "learning_rate": 6.02918316072043e-05,
      "loss": 0.4499,
      "step": 110
    },
    {
      "epoch": 1.2016238159675237,
      "grad_norm": 0.6905937205369403,
      "learning_rate": 5.9853547991628967e-05,
      "loss": 0.4477,
      "step": 111
    },
    {
      "epoch": 1.212449255751015,
      "grad_norm": 0.6381399917783538,
      "learning_rate": 5.941207850124325e-05,
      "loss": 0.4421,
      "step": 112
    },
    {
      "epoch": 1.2232746955345062,
      "grad_norm": 0.7541941618755881,
      "learning_rate": 5.896749397812181e-05,
      "loss": 0.4382,
      "step": 113
    },
    {
      "epoch": 1.2341001353179972,
      "grad_norm": 0.5923805845803851,
      "learning_rate": 5.8519865764204834e-05,
      "loss": 0.4459,
      "step": 114
    },
    {
      "epoch": 1.2449255751014885,
      "grad_norm": 0.6109388260249382,
      "learning_rate": 5.8069265689849884e-05,
      "loss": 0.4513,
      "step": 115
    },
    {
      "epoch": 1.2557510148849798,
      "grad_norm": 0.533980708840664,
      "learning_rate": 5.761576606230538e-05,
      "loss": 0.443,
      "step": 116
    },
    {
      "epoch": 1.266576454668471,
      "grad_norm": 0.7052629150640184,
      "learning_rate": 5.7159439654107506e-05,
      "loss": 0.4404,
      "step": 117
    },
    {
      "epoch": 1.277401894451962,
      "grad_norm": 0.6111064530168573,
      "learning_rate": 5.6700359691402533e-05,
      "loss": 0.4402,
      "step": 118
    },
    {
      "epoch": 1.2882273342354533,
      "grad_norm": 0.5508566705964758,
      "learning_rate": 5.6238599842196285e-05,
      "loss": 0.4449,
      "step": 119
    },
    {
      "epoch": 1.2990527740189446,
      "grad_norm": 0.6028745516198664,
      "learning_rate": 5.5774234204532746e-05,
      "loss": 0.4365,
      "step": 120
    },
    {
      "epoch": 1.3098782138024356,
      "grad_norm": 0.721324745203013,
      "learning_rate": 5.5307337294603595e-05,
      "loss": 0.4442,
      "step": 121
    },
    {
      "epoch": 1.320703653585927,
      "grad_norm": 0.9341313496474372,
      "learning_rate": 5.483798403479072e-05,
      "loss": 0.4403,
      "step": 122
    },
    {
      "epoch": 1.3315290933694182,
      "grad_norm": 1.0476784062313165,
      "learning_rate": 5.436624974164349e-05,
      "loss": 0.4434,
      "step": 123
    },
    {
      "epoch": 1.3423545331529092,
      "grad_norm": 0.917250294000449,
      "learning_rate": 5.389221011379281e-05,
      "loss": 0.4409,
      "step": 124
    },
    {
      "epoch": 1.3531799729364005,
      "grad_norm": 0.7818649176171457,
      "learning_rate": 5.3415941219803895e-05,
      "loss": 0.4405,
      "step": 125
    },
    {
      "epoch": 1.3640054127198917,
      "grad_norm": 0.5772276458708397,
      "learning_rate": 5.2937519485969525e-05,
      "loss": 0.4397,
      "step": 126
    },
    {
      "epoch": 1.374830852503383,
      "grad_norm": 0.4347968698038521,
      "learning_rate": 5.245702168404616e-05,
      "loss": 0.4325,
      "step": 127
    },
    {
      "epoch": 1.3856562922868743,
      "grad_norm": 0.49507391797694716,
      "learning_rate": 5.1974524918934336e-05,
      "loss": 0.4353,
      "step": 128
    },
    {
      "epoch": 1.3964817320703653,
      "grad_norm": 0.49182716841159074,
      "learning_rate": 5.14901066163058e-05,
      "loss": 0.4341,
      "step": 129
    },
    {
      "epoch": 1.4073071718538566,
      "grad_norm": 0.5209355239170891,
      "learning_rate": 5.1003844510179126e-05,
      "loss": 0.4391,
      "step": 130
    },
    {
      "epoch": 1.4181326116373478,
      "grad_norm": 0.555660282513782,
      "learning_rate": 5.0515816630445795e-05,
      "loss": 0.4408,
      "step": 131
    },
    {
      "epoch": 1.4289580514208389,
      "grad_norm": 0.4906371694319118,
      "learning_rate": 5.002610129034883e-05,
      "loss": 0.442,
      "step": 132
    },
    {
      "epoch": 1.4397834912043301,
      "grad_norm": 0.39564350538031545,
      "learning_rate": 4.953477707391597e-05,
      "loss": 0.4359,
      "step": 133
    },
    {
      "epoch": 1.4506089309878214,
      "grad_norm": 0.36420248532962185,
      "learning_rate": 4.90419228233494e-05,
      "loss": 0.4371,
      "step": 134
    },
    {
      "epoch": 1.4614343707713127,
      "grad_norm": 0.3693636826961334,
      "learning_rate": 4.854761762637403e-05,
      "loss": 0.4375,
      "step": 135
    },
    {
      "epoch": 1.472259810554804,
      "grad_norm": 0.37103016749538226,
      "learning_rate": 4.805194080354641e-05,
      "loss": 0.4367,
      "step": 136
    },
    {
      "epoch": 1.483085250338295,
      "grad_norm": 0.4634070182199332,
      "learning_rate": 4.7554971895526175e-05,
      "loss": 0.434,
      "step": 137
    },
    {
      "epoch": 1.4939106901217862,
      "grad_norm": 0.39637839232896416,
      "learning_rate": 4.705679065031235e-05,
      "loss": 0.4405,
      "step": 138
    },
    {
      "epoch": 1.5047361299052775,
      "grad_norm": 0.4167931396685767,
      "learning_rate": 4.6557477010446206e-05,
      "loss": 0.439,
      "step": 139
    },
    {
      "epoch": 1.5155615696887685,
      "grad_norm": 0.34078264652362333,
      "learning_rate": 4.605711110018307e-05,
      "loss": 0.4383,
      "step": 140
    },
    {
      "epoch": 1.5263870094722598,
      "grad_norm": 0.326633599627453,
      "learning_rate": 4.555577321263477e-05,
      "loss": 0.4344,
      "step": 141
    },
    {
      "epoch": 1.537212449255751,
      "grad_norm": 0.3976277159519867,
      "learning_rate": 4.505354379688518e-05,
      "loss": 0.4354,
      "step": 142
    },
    {
      "epoch": 1.548037889039242,
      "grad_norm": 0.4357572785065164,
      "learning_rate": 4.4550503445080606e-05,
      "loss": 0.4417,
      "step": 143
    },
    {
      "epoch": 1.5588633288227334,
      "grad_norm": 0.3916962827302459,
      "learning_rate": 4.4046732879497295e-05,
      "loss": 0.4341,
      "step": 144
    },
    {
      "epoch": 1.5696887686062246,
      "grad_norm": 0.30012283877463697,
      "learning_rate": 4.354231293958801e-05,
      "loss": 0.4322,
      "step": 145
    },
    {
      "epoch": 1.5805142083897157,
      "grad_norm": 0.2885107465168404,
      "learning_rate": 4.3037324569009854e-05,
      "loss": 0.4323,
      "step": 146
    },
    {
      "epoch": 1.5913396481732072,
      "grad_norm": 0.26624048862887445,
      "learning_rate": 4.2531848802635264e-05,
      "loss": 0.4278,
      "step": 147
    },
    {
      "epoch": 1.6021650879566982,
      "grad_norm": 0.3506386895842403,
      "learning_rate": 4.202596675354851e-05,
      "loss": 0.4479,
      "step": 148
    },
    {
      "epoch": 1.6129905277401895,
      "grad_norm": 0.40487113874803043,
      "learning_rate": 4.151975960002958e-05,
      "loss": 0.4302,
      "step": 149
    },
    {
      "epoch": 1.6238159675236807,
      "grad_norm": 0.23297174487170752,
      "learning_rate": 4.101330857252752e-05,
      "loss": 0.4367,
      "step": 150
    },
    {
      "epoch": 1.6346414073071718,
      "grad_norm": 0.30369826324347926,
      "learning_rate": 4.050669494062561e-05,
      "loss": 0.4372,
      "step": 151
    },
    {
      "epoch": 1.645466847090663,
      "grad_norm": 0.29979293086538217,
      "learning_rate": 4e-05,
      "loss": 0.4308,
      "step": 152
    },
    {
      "epoch": 1.6562922868741543,
      "grad_norm": 0.21998478305579725,
      "learning_rate": 3.9493305059374405e-05,
      "loss": 0.444,
      "step": 153
    },
    {
      "epoch": 1.6671177266576453,
      "grad_norm": 0.2822111397888722,
      "learning_rate": 3.8986691427472496e-05,
      "loss": 0.4362,
      "step": 154
    },
    {
      "epoch": 1.6779431664411368,
      "grad_norm": 0.19298090698642797,
      "learning_rate": 3.8480240399970436e-05,
      "loss": 0.4291,
      "step": 155
    },
    {
      "epoch": 1.6887686062246279,
      "grad_norm": 0.24201705060461937,
      "learning_rate": 3.7974033246451496e-05,
      "loss": 0.4374,
      "step": 156
    },
    {
      "epoch": 1.699594046008119,
      "grad_norm": 0.2926866609659229,
      "learning_rate": 3.746815119736475e-05,
      "loss": 0.4237,
      "step": 157
    },
    {
      "epoch": 1.7104194857916104,
      "grad_norm": 0.22986191257503155,
      "learning_rate": 3.696267543099016e-05,
      "loss": 0.4349,
      "step": 158
    },
    {
      "epoch": 1.7212449255751014,
      "grad_norm": 0.21762455575624456,
      "learning_rate": 3.6457687060412e-05,
      "loss": 0.4411,
      "step": 159
    },
    {
      "epoch": 1.7320703653585927,
      "grad_norm": 0.23006031819487235,
      "learning_rate": 3.595326712050272e-05,
      "loss": 0.4352,
      "step": 160
    },
    {
      "epoch": 1.742895805142084,
      "grad_norm": 0.23480749792102354,
      "learning_rate": 3.5449496554919414e-05,
      "loss": 0.432,
      "step": 161
    },
    {
      "epoch": 1.753721244925575,
      "grad_norm": 0.20514925914229476,
      "learning_rate": 3.494645620311484e-05,
      "loss": 0.4374,
      "step": 162
    },
    {
      "epoch": 1.7645466847090663,
      "grad_norm": 0.21723286316559756,
      "learning_rate": 3.444422678736525e-05,
      "loss": 0.4309,
      "step": 163
    },
    {
      "epoch": 1.7753721244925575,
      "grad_norm": 0.19908292246838455,
      "learning_rate": 3.394288889981695e-05,
      "loss": 0.4285,
      "step": 164
    },
    {
      "epoch": 1.7861975642760486,
      "grad_norm": 0.21848824155386468,
      "learning_rate": 3.34425229895538e-05,
      "loss": 0.4358,
      "step": 165
    },
    {
      "epoch": 1.79702300405954,
      "grad_norm": 0.19073388047330828,
      "learning_rate": 3.294320934968768e-05,
      "loss": 0.4285,
      "step": 166
    },
    {
      "epoch": 1.8078484438430311,
      "grad_norm": 0.20593306151831323,
      "learning_rate": 3.2445028104473845e-05,
      "loss": 0.4347,
      "step": 167
    },
    {
      "epoch": 1.8186738836265224,
      "grad_norm": 0.21441680855669412,
      "learning_rate": 3.194805919645359e-05,
      "loss": 0.4308,
      "step": 168
    },
    {
      "epoch": 1.8294993234100136,
      "grad_norm": 0.18002633127557865,
      "learning_rate": 3.145238237362596e-05,
      "loss": 0.4392,
      "step": 169
    },
    {
      "epoch": 1.8403247631935047,
      "grad_norm": 0.18961769159187056,
      "learning_rate": 3.0958077176650606e-05,
      "loss": 0.43,
      "step": 170
    },
    {
      "epoch": 1.851150202976996,
      "grad_norm": 0.16827003419417288,
      "learning_rate": 3.0465222926084036e-05,
      "loss": 0.426,
      "step": 171
    },
    {
      "epoch": 1.8619756427604872,
      "grad_norm": 0.1766724359697219,
      "learning_rate": 2.997389870965118e-05,
      "loss": 0.4273,
      "step": 172
    },
    {
      "epoch": 1.8728010825439783,
      "grad_norm": 0.1937265928321859,
      "learning_rate": 2.948418336955421e-05,
      "loss": 0.4216,
      "step": 173
    },
    {
      "epoch": 1.8836265223274695,
      "grad_norm": 0.1733839063792691,
      "learning_rate": 2.899615548982088e-05,
      "loss": 0.4342,
      "step": 174
    },
    {
      "epoch": 1.8944519621109608,
      "grad_norm": 0.1691834042941349,
      "learning_rate": 2.8509893383694213e-05,
      "loss": 0.428,
      "step": 175
    },
    {
      "epoch": 1.9052774018944518,
      "grad_norm": 0.17556950089441328,
      "learning_rate": 2.8025475081065684e-05,
      "loss": 0.4235,
      "step": 176
    },
    {
      "epoch": 1.9161028416779433,
      "grad_norm": 0.17292630455576483,
      "learning_rate": 2.754297831595385e-05,
      "loss": 0.4242,
      "step": 177
    },
    {
      "epoch": 1.9269282814614344,
      "grad_norm": 0.203395158544133,
      "learning_rate": 2.7062480514030478e-05,
      "loss": 0.4346,
      "step": 178
    },
    {
      "epoch": 1.9377537212449256,
      "grad_norm": 0.17761313930923656,
      "learning_rate": 2.658405878019612e-05,
      "loss": 0.4284,
      "step": 179
    },
    {
      "epoch": 1.9485791610284169,
      "grad_norm": 0.1857053642993675,
      "learning_rate": 2.6107789886207195e-05,
      "loss": 0.4311,
      "step": 180
    },
    {
      "epoch": 1.959404600811908,
      "grad_norm": 0.168623231091507,
      "learning_rate": 2.563375025835652e-05,
      "loss": 0.4296,
      "step": 181
    },
    {
      "epoch": 1.9702300405953992,
      "grad_norm": 0.16395731003065203,
      "learning_rate": 2.5162015965209295e-05,
      "loss": 0.4263,
      "step": 182
    },
    {
      "epoch": 1.9810554803788905,
      "grad_norm": 0.17813784128811913,
      "learning_rate": 2.4692662705396412e-05,
      "loss": 0.4337,
      "step": 183
    },
    {
      "epoch": 1.9918809201623815,
      "grad_norm": 0.1692020847630486,
      "learning_rate": 2.4225765795467267e-05,
      "loss": 0.4345,
      "step": 184
    },
    {
      "epoch": 2.002706359945873,
      "grad_norm": 0.29642981794747025,
      "learning_rate": 2.376140015780372e-05,
      "loss": 0.6828,
      "step": 185
    },
    {
      "epoch": 2.013531799729364,
      "grad_norm": 0.22282508513365898,
      "learning_rate": 2.3299640308597487e-05,
      "loss": 0.4033,
      "step": 186
    },
    {
      "epoch": 2.024357239512855,
      "grad_norm": 0.19906807218157607,
      "learning_rate": 2.2840560345892518e-05,
      "loss": 0.4028,
      "step": 187
    },
    {
      "epoch": 2.0351826792963466,
      "grad_norm": 0.21438465882623423,
      "learning_rate": 2.2384233937694626e-05,
      "loss": 0.4036,
      "step": 188
    },
    {
      "epoch": 2.0460081190798376,
      "grad_norm": 0.20912970610685283,
      "learning_rate": 2.1930734310150116e-05,
      "loss": 0.3975,
      "step": 189
    },
    {
      "epoch": 2.0568335588633286,
      "grad_norm": 0.2005177294111497,
      "learning_rate": 2.1480134235795173e-05,
      "loss": 0.4022,
      "step": 190
    },
    {
      "epoch": 2.06765899864682,
      "grad_norm": 0.19782562683985525,
      "learning_rate": 2.10325060218782e-05,
      "loss": 0.4073,
      "step": 191
    },
    {
      "epoch": 2.078484438430311,
      "grad_norm": 0.23611436091693633,
      "learning_rate": 2.0587921498756768e-05,
      "loss": 0.4109,
      "step": 192
    },
    {
      "epoch": 2.089309878213802,
      "grad_norm": 0.20974660868469683,
      "learning_rate": 2.014645200837105e-05,
      "loss": 0.4034,
      "step": 193
    },
    {
      "epoch": 2.1001353179972937,
      "grad_norm": 0.19267649854044877,
      "learning_rate": 1.9708168392795718e-05,
      "loss": 0.3953,
      "step": 194
    },
    {
      "epoch": 2.1109607577807847,
      "grad_norm": 0.22011389396639292,
      "learning_rate": 1.9273140982871936e-05,
      "loss": 0.4047,
      "step": 195
    },
    {
      "epoch": 2.121786197564276,
      "grad_norm": 0.18180878911713266,
      "learning_rate": 1.8841439586921515e-05,
      "loss": 0.4039,
      "step": 196
    },
    {
      "epoch": 2.1326116373477673,
      "grad_norm": 0.21012423766700125,
      "learning_rate": 1.841313347954482e-05,
      "loss": 0.3986,
      "step": 197
    },
    {
      "epoch": 2.1434370771312583,
      "grad_norm": 0.15463640007730903,
      "learning_rate": 1.7988291390504348e-05,
      "loss": 0.3975,
      "step": 198
    },
    {
      "epoch": 2.15426251691475,
      "grad_norm": 0.193505600575409,
      "learning_rate": 1.7566981493695828e-05,
      "loss": 0.4043,
      "step": 199
    },
    {
      "epoch": 2.165087956698241,
      "grad_norm": 0.16774935171850072,
      "learning_rate": 1.71492713962083e-05,
      "loss": 0.4068,
      "step": 200
    },
    {
      "epoch": 2.175913396481732,
      "grad_norm": 0.15543728413541416,
      "learning_rate": 1.673522812747544e-05,
      "loss": 0.4025,
      "step": 201
    },
    {
      "epoch": 2.1867388362652234,
      "grad_norm": 0.1581064576242491,
      "learning_rate": 1.6324918128519306e-05,
      "loss": 0.3995,
      "step": 202
    },
    {
      "epoch": 2.1975642760487144,
      "grad_norm": 0.1433403956609895,
      "learning_rate": 1.5918407241288678e-05,
      "loss": 0.406,
      "step": 203
    },
    {
      "epoch": 2.208389715832206,
      "grad_norm": 0.1444921358019384,
      "learning_rate": 1.5515760698093485e-05,
      "loss": 0.4043,
      "step": 204
    },
    {
      "epoch": 2.219215155615697,
      "grad_norm": 0.14781491117947443,
      "learning_rate": 1.511704311113705e-05,
      "loss": 0.4024,
      "step": 205
    },
    {
      "epoch": 2.230040595399188,
      "grad_norm": 0.1394579893993049,
      "learning_rate": 1.4722318462147844e-05,
      "loss": 0.4029,
      "step": 206
    },
    {
      "epoch": 2.2408660351826795,
      "grad_norm": 0.16023643196582224,
      "learning_rate": 1.4331650092112406e-05,
      "loss": 0.4066,
      "step": 207
    },
    {
      "epoch": 2.2516914749661705,
      "grad_norm": 0.1348210060414343,
      "learning_rate": 1.394510069111112e-05,
      "loss": 0.3973,
      "step": 208
    },
    {
      "epoch": 2.2625169147496615,
      "grad_norm": 0.13391825126021842,
      "learning_rate": 1.3562732288258377e-05,
      "loss": 0.4025,
      "step": 209
    },
    {
      "epoch": 2.273342354533153,
      "grad_norm": 0.12844012269731564,
      "learning_rate": 1.3184606241748857e-05,
      "loss": 0.3978,
      "step": 210
    },
    {
      "epoch": 2.284167794316644,
      "grad_norm": 0.13584285882845534,
      "learning_rate": 1.2810783229011486e-05,
      "loss": 0.3991,
      "step": 211
    },
    {
      "epoch": 2.2949932341001356,
      "grad_norm": 0.13168511677095765,
      "learning_rate": 1.2441323236972536e-05,
      "loss": 0.3999,
      "step": 212
    },
    {
      "epoch": 2.3058186738836266,
      "grad_norm": 0.12398647003241701,
      "learning_rate": 1.2076285552429642e-05,
      "loss": 0.4077,
      "step": 213
    },
    {
      "epoch": 2.3166441136671176,
      "grad_norm": 0.13229640626265612,
      "learning_rate": 1.1715728752538103e-05,
      "loss": 0.4019,
      "step": 214
    },
    {
      "epoch": 2.3274695534506087,
      "grad_norm": 0.12385054713595073,
      "learning_rate": 1.1359710695411086e-05,
      "loss": 0.4029,
      "step": 215
    },
    {
      "epoch": 2.3382949932341,
      "grad_norm": 0.12591732770348577,
      "learning_rate": 1.100828851083521e-05,
      "loss": 0.4084,
      "step": 216
    },
    {
      "epoch": 2.349120433017591,
      "grad_norm": 0.13251886738270005,
      "learning_rate": 1.0661518591102973e-05,
      "loss": 0.4019,
      "step": 217
    },
    {
      "epoch": 2.3599458728010827,
      "grad_norm": 0.1231996291813589,
      "learning_rate": 1.0319456581963578e-05,
      "loss": 0.4097,
      "step": 218
    },
    {
      "epoch": 2.3707713125845737,
      "grad_norm": 0.12007287837624477,
      "learning_rate": 9.982157373693502e-06,
      "loss": 0.407,
      "step": 219
    },
    {
      "epoch": 2.381596752368065,
      "grad_norm": 0.11828968531594988,
      "learning_rate": 9.649675092288366e-06,
      "loss": 0.3942,
      "step": 220
    },
    {
      "epoch": 2.3924221921515563,
      "grad_norm": 0.1266852929005505,
      "learning_rate": 9.322063090777331e-06,
      "loss": 0.402,
      "step": 221
    },
    {
      "epoch": 2.4032476319350473,
      "grad_norm": 0.13131525920519638,
      "learning_rate": 8.99937394066165e-06,
      "loss": 0.4079,
      "step": 222
    },
    {
      "epoch": 2.414073071718539,
      "grad_norm": 0.1276300518950864,
      "learning_rate": 8.681659423478587e-06,
      "loss": 0.4075,
      "step": 223
    },
    {
      "epoch": 2.42489851150203,
      "grad_norm": 0.1284596882094962,
      "learning_rate": 8.368970522492064e-06,
      "loss": 0.4041,
      "step": 224
    },
    {
      "epoch": 2.435723951285521,
      "grad_norm": 0.11600707488242912,
      "learning_rate": 8.06135741451146e-06,
      "loss": 0.4024,
      "step": 225
    },
    {
      "epoch": 2.4465493910690124,
      "grad_norm": 0.1369413509292135,
      "learning_rate": 7.758869461839808e-06,
      "loss": 0.4054,
      "step": 226
    },
    {
      "epoch": 2.4573748308525034,
      "grad_norm": 0.11693298201832159,
      "learning_rate": 7.461555204352655e-06,
      "loss": 0.3988,
      "step": 227
    },
    {
      "epoch": 2.4682002706359945,
      "grad_norm": 0.11962758908233623,
      "learning_rate": 7.169462351708958e-06,
      "loss": 0.4007,
      "step": 228
    },
    {
      "epoch": 2.479025710419486,
      "grad_norm": 0.12081712635114038,
      "learning_rate": 6.882637775695147e-06,
      "loss": 0.3973,
      "step": 229
    },
    {
      "epoch": 2.489851150202977,
      "grad_norm": 0.12426870098020382,
      "learning_rate": 6.60112750270371e-06,
      "loss": 0.3985,
      "step": 230
    },
    {
      "epoch": 2.500676589986468,
      "grad_norm": 0.11586754894597827,
      "learning_rate": 6.324976706347317e-06,
      "loss": 0.4078,
      "step": 231
    },
    {
      "epoch": 2.5115020297699595,
      "grad_norm": 0.10751212068030966,
      "learning_rate": 6.054229700209959e-06,
      "loss": 0.4054,
      "step": 232
    },
    {
      "epoch": 2.5223274695534506,
      "grad_norm": 0.12026857943267745,
      "learning_rate": 5.788929930735916e-06,
      "loss": 0.4,
      "step": 233
    },
    {
      "epoch": 2.533152909336942,
      "grad_norm": 0.10350592956190559,
      "learning_rate": 5.529119970258014e-06,
      "loss": 0.4032,
      "step": 234
    },
    {
      "epoch": 2.543978349120433,
      "grad_norm": 0.09850020851661254,
      "learning_rate": 5.274841510166062e-06,
      "loss": 0.3961,
      "step": 235
    },
    {
      "epoch": 2.554803788903924,
      "grad_norm": 0.10692798146184619,
      "learning_rate": 5.026135354216717e-06,
      "loss": 0.3953,
      "step": 236
    },
    {
      "epoch": 2.565629228687415,
      "grad_norm": 0.1026156051139519,
      "learning_rate": 4.783041411985716e-06,
      "loss": 0.4026,
      "step": 237
    },
    {
      "epoch": 2.5764546684709067,
      "grad_norm": 0.0984956835805336,
      "learning_rate": 4.545598692463675e-06,
      "loss": 0.4012,
      "step": 238
    },
    {
      "epoch": 2.5872801082543977,
      "grad_norm": 0.09517297299245513,
      "learning_rate": 4.3138452977963266e-06,
      "loss": 0.3986,
      "step": 239
    },
    {
      "epoch": 2.598105548037889,
      "grad_norm": 0.1017472403907421,
      "learning_rate": 4.087818417170337e-06,
      "loss": 0.4054,
      "step": 240
    },
    {
      "epoch": 2.60893098782138,
      "grad_norm": 0.10517463959416935,
      "learning_rate": 3.867554320845601e-06,
      "loss": 0.4081,
      "step": 241
    },
    {
      "epoch": 2.6197564276048713,
      "grad_norm": 0.10283676657115687,
      "learning_rate": 3.6530883543350038e-06,
      "loss": 0.4078,
      "step": 242
    },
    {
      "epoch": 2.6305818673883627,
      "grad_norm": 0.09668185967839843,
      "learning_rate": 3.4444549327325325e-06,
      "loss": 0.3986,
      "step": 243
    },
    {
      "epoch": 2.641407307171854,
      "grad_norm": 0.10655772797882479,
      "learning_rate": 3.241687535190776e-06,
      "loss": 0.3971,
      "step": 244
    },
    {
      "epoch": 2.6522327469553453,
      "grad_norm": 0.10412740990228317,
      "learning_rate": 3.0448186995485307e-06,
      "loss": 0.398,
      "step": 245
    },
    {
      "epoch": 2.6630581867388363,
      "grad_norm": 0.10294803068291954,
      "learning_rate": 2.853880017109516e-06,
      "loss": 0.3984,
      "step": 246
    },
    {
      "epoch": 2.6738836265223274,
      "grad_norm": 0.09605814302194737,
      "learning_rate": 2.6689021275729366e-06,
      "loss": 0.4045,
      "step": 247
    },
    {
      "epoch": 2.6847090663058184,
      "grad_norm": 0.10355709761757277,
      "learning_rate": 2.489914714116788e-06,
      "loss": 0.3993,
      "step": 248
    },
    {
      "epoch": 2.69553450608931,
      "grad_norm": 0.10534682127661901,
      "learning_rate": 2.316946498634605e-06,
      "loss": 0.396,
      "step": 249
    },
    {
      "epoch": 2.706359945872801,
      "grad_norm": 0.09850383904197658,
      "learning_rate": 2.1500252371265253e-06,
      "loss": 0.4071,
      "step": 250
    },
    {
      "epoch": 2.7171853856562924,
      "grad_norm": 0.09796274629692603,
      "learning_rate": 1.989177715245307e-06,
      "loss": 0.4029,
      "step": 251
    },
    {
      "epoch": 2.7280108254397835,
      "grad_norm": 0.09270281667503455,
      "learning_rate": 1.8344297439980475e-06,
      "loss": 0.4067,
      "step": 252
    },
    {
      "epoch": 2.7388362652232745,
      "grad_norm": 0.09318564757274785,
      "learning_rate": 1.685806155604346e-06,
      "loss": 0.4055,
      "step": 253
    },
    {
      "epoch": 2.749661705006766,
      "grad_norm": 0.09361381405899726,
      "learning_rate": 1.5433307995115043e-06,
      "loss": 0.3995,
      "step": 254
    },
    {
      "epoch": 2.760487144790257,
      "grad_norm": 0.08807467149477492,
      "learning_rate": 1.4070265385674176e-06,
      "loss": 0.3949,
      "step": 255
    },
    {
      "epoch": 2.7713125845737485,
      "grad_norm": 0.09403613096147612,
      "learning_rate": 1.276915245351833e-06,
      "loss": 0.3936,
      "step": 256
    },
    {
      "epoch": 2.7821380243572396,
      "grad_norm": 0.0872925419115015,
      "learning_rate": 1.1530177986664425e-06,
      "loss": 0.4077,
      "step": 257
    },
    {
      "epoch": 2.7929634641407306,
      "grad_norm": 0.0850031850974251,
      "learning_rate": 1.0353540801845229e-06,
      "loss": 0.4003,
      "step": 258
    },
    {
      "epoch": 2.803788903924222,
      "grad_norm": 0.08692284082323248,
      "learning_rate": 9.239429712605274e-07,
      "loss": 0.3991,
      "step": 259
    },
    {
      "epoch": 2.814614343707713,
      "grad_norm": 0.08257477054022927,
      "learning_rate": 8.188023499002206e-07,
      "loss": 0.4046,
      "step": 260
    },
    {
      "epoch": 2.825439783491204,
      "grad_norm": 0.08278787367986447,
      "learning_rate": 7.199490878918314e-07,
      "loss": 0.3939,
      "step": 261
    },
    {
      "epoch": 2.8362652232746957,
      "grad_norm": 0.08297388622794061,
      "learning_rate": 6.273990480986314e-07,
      "loss": 0.3993,
      "step": 262
    },
    {
      "epoch": 2.8470906630581867,
      "grad_norm": 0.08598446782763781,
      "learning_rate": 5.411670819134651e-07,
      "loss": 0.4034,
      "step": 263
    },
    {
      "epoch": 2.8579161028416777,
      "grad_norm": 0.08779257361395829,
      "learning_rate": 4.6126702687554483e-07,
      "loss": 0.399,
      "step": 264
    },
    {
      "epoch": 2.8687415426251692,
      "grad_norm": 0.08649244863382963,
      "learning_rate": 3.8771170444996895e-07,
      "loss": 0.4036,
      "step": 265
    },
    {
      "epoch": 2.8795669824086603,
      "grad_norm": 0.08480882436189274,
      "learning_rate": 3.2051291797027925e-07,
      "loss": 0.402,
      "step": 266
    },
    {
      "epoch": 2.8903924221921518,
      "grad_norm": 0.08449641777806831,
      "learning_rate": 2.59681450744389e-07,
      "loss": 0.3975,
      "step": 267
    },
    {
      "epoch": 2.901217861975643,
      "grad_norm": 0.08224385199164784,
      "learning_rate": 2.0522706432419382e-07,
      "loss": 0.3989,
      "step": 268
    },
    {
      "epoch": 2.912043301759134,
      "grad_norm": 0.08340221321505677,
      "learning_rate": 1.5715849693916264e-07,
      "loss": 0.4009,
      "step": 269
    },
    {
      "epoch": 2.9228687415426253,
      "grad_norm": 0.08271264876685368,
      "learning_rate": 1.1548346209410366e-07,
      "loss": 0.4023,
      "step": 270
    },
    {
      "epoch": 2.9336941813261164,
      "grad_norm": 0.08260005569786469,
      "learning_rate": 8.020864733140343e-08,
      "loss": 0.3993,
      "step": 271
    },
    {
      "epoch": 2.944519621109608,
      "grad_norm": 0.08278469542802738,
      "learning_rate": 5.133971315788966e-08,
      "loss": 0.4053,
      "step": 272
    },
    {
      "epoch": 2.955345060893099,
      "grad_norm": 0.08624728365001678,
      "learning_rate": 2.8881292136468952e-08,
      "loss": 0.4024,
      "step": 273
    },
    {
      "epoch": 2.96617050067659,
      "grad_norm": 0.08795666384161806,
      "learning_rate": 1.2836988142779228e-08,
      "loss": 0.4015,
      "step": 274
    },
    {
      "epoch": 2.976995940460081,
      "grad_norm": 0.08192972934685559,
      "learning_rate": 3.209375786856761e-09,
      "loss": 0.4094,
      "step": 275
    },
    {
      "epoch": 2.9878213802435725,
      "grad_norm": 0.07989052292080247,
      "learning_rate": 0.0,
      "loss": 0.3967,
      "step": 276
    },
    {
      "epoch": 2.9878213802435725,
      "step": 276,
      "total_flos": 6.629121904995205e+18,
      "train_loss": 0.463554704102917,
      "train_runtime": 27205.7226,
      "train_samples_per_second": 5.214,
      "train_steps_per_second": 0.01
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 276,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.629121904995205e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}