{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9904153354632586,
  "eval_steps": 500,
  "global_step": 312,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009584664536741214,
      "grad_norm": 5.816953152720709,
      "learning_rate": 1.25e-06,
      "loss": 0.8809,
      "step": 1
    },
    {
      "epoch": 0.019169329073482427,
      "grad_norm": 5.964463857967548,
      "learning_rate": 2.5e-06,
      "loss": 0.8476,
      "step": 2
    },
    {
      "epoch": 0.02875399361022364,
      "grad_norm": 5.85622862207096,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.8749,
      "step": 3
    },
    {
      "epoch": 0.038338658146964855,
      "grad_norm": 5.365349431954843,
      "learning_rate": 5e-06,
      "loss": 0.8422,
      "step": 4
    },
    {
      "epoch": 0.04792332268370607,
      "grad_norm": 4.006201698306361,
      "learning_rate": 6.25e-06,
      "loss": 0.8005,
      "step": 5
    },
    {
      "epoch": 0.05750798722044728,
      "grad_norm": 2.122270635740227,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.7687,
      "step": 6
    },
    {
      "epoch": 0.0670926517571885,
      "grad_norm": 3.975962985256971,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.7686,
      "step": 7
    },
    {
      "epoch": 0.07667731629392971,
      "grad_norm": 4.3633197716567595,
      "learning_rate": 1e-05,
      "loss": 0.761,
      "step": 8
    },
    {
      "epoch": 0.08626198083067092,
      "grad_norm": 4.812484653337378,
      "learning_rate": 1.125e-05,
      "loss": 0.7451,
      "step": 9
    },
    {
      "epoch": 0.09584664536741214,
      "grad_norm": 4.996167212561038,
      "learning_rate": 1.25e-05,
      "loss": 0.7657,
      "step": 10
    },
    {
      "epoch": 0.10543130990415335,
      "grad_norm": 3.391593689539008,
      "learning_rate": 1.375e-05,
      "loss": 0.7293,
      "step": 11
    },
    {
      "epoch": 0.11501597444089456,
      "grad_norm": 2.024258637124186,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.6753,
      "step": 12
    },
    {
      "epoch": 0.12460063897763578,
      "grad_norm": 2.368340767997839,
      "learning_rate": 1.6250000000000002e-05,
      "loss": 0.675,
      "step": 13
    },
    {
      "epoch": 0.134185303514377,
      "grad_norm": 1.9034533687196296,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.6346,
      "step": 14
    },
    {
      "epoch": 0.14376996805111822,
      "grad_norm": 1.511274358606332,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.6395,
      "step": 15
    },
    {
      "epoch": 0.15335463258785942,
      "grad_norm": 1.5258008797446443,
      "learning_rate": 2e-05,
      "loss": 0.6605,
      "step": 16
    },
    {
      "epoch": 0.16293929712460065,
      "grad_norm": 1.4701453015485093,
      "learning_rate": 2.125e-05,
      "loss": 0.6345,
      "step": 17
    },
    {
      "epoch": 0.17252396166134185,
      "grad_norm": 1.139297657192997,
      "learning_rate": 2.25e-05,
      "loss": 0.6313,
      "step": 18
    },
    {
      "epoch": 0.18210862619808307,
      "grad_norm": 1.0628460283930738,
      "learning_rate": 2.375e-05,
      "loss": 0.6093,
      "step": 19
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 1.0582580403121424,
      "learning_rate": 2.5e-05,
      "loss": 0.5867,
      "step": 20
    },
    {
      "epoch": 0.2012779552715655,
      "grad_norm": 0.8977556913012904,
      "learning_rate": 2.625e-05,
      "loss": 0.5876,
      "step": 21
    },
    {
      "epoch": 0.2108626198083067,
      "grad_norm": 1.3235677545982394,
      "learning_rate": 2.75e-05,
      "loss": 0.5909,
      "step": 22
    },
    {
      "epoch": 0.22044728434504793,
      "grad_norm": 0.8869847698482135,
      "learning_rate": 2.875e-05,
      "loss": 0.5739,
      "step": 23
    },
    {
      "epoch": 0.23003194888178913,
      "grad_norm": 1.1759129421868764,
      "learning_rate": 3.0000000000000004e-05,
      "loss": 0.588,
      "step": 24
    },
    {
      "epoch": 0.23961661341853036,
      "grad_norm": 0.9909103854124853,
      "learning_rate": 3.125e-05,
      "loss": 0.5675,
      "step": 25
    },
    {
      "epoch": 0.24920127795527156,
      "grad_norm": 1.0713203550156574,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.5492,
      "step": 26
    },
    {
      "epoch": 0.25878594249201275,
      "grad_norm": 0.9586606498055252,
      "learning_rate": 3.375e-05,
      "loss": 0.5539,
      "step": 27
    },
    {
      "epoch": 0.268370607028754,
      "grad_norm": 0.9213875490205139,
      "learning_rate": 3.5000000000000004e-05,
      "loss": 0.5319,
      "step": 28
    },
    {
      "epoch": 0.2779552715654952,
      "grad_norm": 1.0198534952792746,
      "learning_rate": 3.625e-05,
      "loss": 0.5722,
      "step": 29
    },
    {
      "epoch": 0.28753993610223644,
      "grad_norm": 0.9853545050808614,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.5514,
      "step": 30
    },
    {
      "epoch": 0.2971246006389776,
      "grad_norm": 0.9771316358011265,
      "learning_rate": 3.875e-05,
      "loss": 0.548,
      "step": 31
    },
    {
      "epoch": 0.30670926517571884,
      "grad_norm": 1.070148811782629,
      "learning_rate": 4e-05,
      "loss": 0.5776,
      "step": 32
    },
    {
      "epoch": 0.31629392971246006,
      "grad_norm": 0.8966401903811619,
      "learning_rate": 3.9998741135094016e-05,
      "loss": 0.5444,
      "step": 33
    },
    {
      "epoch": 0.3258785942492013,
      "grad_norm": 0.893875729745141,
      "learning_rate": 3.999496469885014e-05,
      "loss": 0.5704,
      "step": 34
    },
    {
      "epoch": 0.3354632587859425,
      "grad_norm": 1.072867766600089,
      "learning_rate": 3.998867116667067e-05,
      "loss": 0.5477,
      "step": 35
    },
    {
      "epoch": 0.3450479233226837,
      "grad_norm": 0.9635968083722172,
      "learning_rate": 3.9979861330826295e-05,
      "loss": 0.5534,
      "step": 36
    },
    {
      "epoch": 0.3546325878594249,
      "grad_norm": 0.9899242418062558,
      "learning_rate": 3.996853630035634e-05,
      "loss": 0.5655,
      "step": 37
    },
    {
      "epoch": 0.36421725239616615,
      "grad_norm": 0.9130531711455715,
      "learning_rate": 3.995469750092912e-05,
      "loss": 0.5817,
      "step": 38
    },
    {
      "epoch": 0.3738019169329074,
      "grad_norm": 0.9387442528138782,
      "learning_rate": 3.9938346674662565e-05,
      "loss": 0.5044,
      "step": 39
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 1.0142174519180858,
      "learning_rate": 3.991948587990479e-05,
      "loss": 0.5359,
      "step": 40
    },
    {
      "epoch": 0.3929712460063898,
      "grad_norm": 0.8959196832437484,
      "learning_rate": 3.989811749097505e-05,
      "loss": 0.5537,
      "step": 41
    },
    {
      "epoch": 0.402555910543131,
      "grad_norm": 0.9987977928627828,
      "learning_rate": 3.9874244197864856e-05,
      "loss": 0.5782,
      "step": 42
    },
    {
      "epoch": 0.41214057507987223,
      "grad_norm": 0.8845338538227933,
      "learning_rate": 3.984786900589929e-05,
      "loss": 0.5415,
      "step": 43
    },
    {
      "epoch": 0.4217252396166134,
      "grad_norm": 1.0047800925961548,
      "learning_rate": 3.98189952353587e-05,
      "loss": 0.5542,
      "step": 44
    },
    {
      "epoch": 0.43130990415335463,
      "grad_norm": 0.8022173762000696,
      "learning_rate": 3.9787626521060736e-05,
      "loss": 0.5483,
      "step": 45
    },
    {
      "epoch": 0.44089456869009586,
      "grad_norm": 0.7883000207976826,
      "learning_rate": 3.9753766811902756e-05,
      "loss": 0.5473,
      "step": 46
    },
    {
      "epoch": 0.4504792332268371,
      "grad_norm": 0.8434281567738428,
      "learning_rate": 3.971742037036472e-05,
      "loss": 0.5634,
      "step": 47
    },
    {
      "epoch": 0.46006389776357826,
      "grad_norm": 1.0277496566918978,
      "learning_rate": 3.96785917719726e-05,
      "loss": 0.558,
      "step": 48
    },
    {
      "epoch": 0.4696485623003195,
      "grad_norm": 0.99005810315587,
      "learning_rate": 3.9637285904722376e-05,
      "loss": 0.5508,
      "step": 49
    },
    {
      "epoch": 0.4792332268370607,
      "grad_norm": 1.0782113201945098,
      "learning_rate": 3.9593507968464714e-05,
      "loss": 0.5478,
      "step": 50
    },
    {
      "epoch": 0.48881789137380194,
      "grad_norm": 0.721872938484921,
      "learning_rate": 3.9547263474250385e-05,
      "loss": 0.5366,
      "step": 51
    },
    {
      "epoch": 0.4984025559105431,
      "grad_norm": 1.2142224090633027,
      "learning_rate": 3.949855824363647e-05,
      "loss": 0.5544,
      "step": 52
    },
    {
      "epoch": 0.5079872204472844,
      "grad_norm": 0.9185255274236219,
      "learning_rate": 3.9447398407953536e-05,
      "loss": 0.4976,
      "step": 53
    },
    {
      "epoch": 0.5175718849840255,
      "grad_norm": 1.0876007980497202,
      "learning_rate": 3.939379040753374e-05,
      "loss": 0.5374,
      "step": 54
    },
    {
      "epoch": 0.5271565495207667,
      "grad_norm": 1.2327514959394714,
      "learning_rate": 3.933774099090013e-05,
      "loss": 0.5517,
      "step": 55
    },
    {
      "epoch": 0.536741214057508,
      "grad_norm": 0.6905831120274069,
      "learning_rate": 3.927925721391707e-05,
      "loss": 0.5114,
      "step": 56
    },
    {
      "epoch": 0.5463258785942492,
      "grad_norm": 1.3327650152930188,
      "learning_rate": 3.9218346438901996e-05,
      "loss": 0.5556,
      "step": 57
    },
    {
      "epoch": 0.5559105431309904,
      "grad_norm": 0.7267801369686653,
      "learning_rate": 3.9155016333698615e-05,
      "loss": 0.5446,
      "step": 58
    },
    {
      "epoch": 0.5654952076677316,
      "grad_norm": 1.1332074133198213,
      "learning_rate": 3.908927487071162e-05,
      "loss": 0.5297,
      "step": 59
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 1.1446700448302367,
      "learning_rate": 3.9021130325903076e-05,
      "loss": 0.5283,
      "step": 60
    },
    {
      "epoch": 0.5846645367412141,
      "grad_norm": 0.9428641512869639,
      "learning_rate": 3.895059127775058e-05,
      "loss": 0.5276,
      "step": 61
    },
    {
      "epoch": 0.5942492012779552,
      "grad_norm": 0.984662057057416,
      "learning_rate": 3.8877666606167354e-05,
      "loss": 0.5053,
      "step": 62
    },
    {
      "epoch": 0.6038338658146964,
      "grad_norm": 0.9200152881682313,
      "learning_rate": 3.880236549138438e-05,
      "loss": 0.5019,
      "step": 63
    },
    {
      "epoch": 0.6134185303514377,
      "grad_norm": 0.9735224575622512,
      "learning_rate": 3.872469741279475e-05,
      "loss": 0.5027,
      "step": 64
    },
    {
      "epoch": 0.6230031948881789,
      "grad_norm": 0.8601738218870937,
      "learning_rate": 3.8644672147760286e-05,
      "loss": 0.5373,
      "step": 65
    },
    {
      "epoch": 0.6325878594249201,
      "grad_norm": 1.0690694771471052,
      "learning_rate": 3.856229977038078e-05,
      "loss": 0.5141,
      "step": 66
    },
    {
      "epoch": 0.6421725239616614,
      "grad_norm": 0.639345682465469,
      "learning_rate": 3.8477590650225735e-05,
      "loss": 0.5237,
      "step": 67
    },
    {
      "epoch": 0.6517571884984026,
      "grad_norm": 0.9825658908653957,
      "learning_rate": 3.839055545102902e-05,
      "loss": 0.5218,
      "step": 68
    },
    {
      "epoch": 0.6613418530351438,
      "grad_norm": 0.6640592278410514,
      "learning_rate": 3.83012051293464e-05,
      "loss": 0.5106,
      "step": 69
    },
    {
      "epoch": 0.670926517571885,
      "grad_norm": 0.9942824925976919,
      "learning_rate": 3.8209550933176324e-05,
      "loss": 0.5371,
      "step": 70
    },
    {
      "epoch": 0.6805111821086262,
      "grad_norm": 0.8321580496371637,
      "learning_rate": 3.8115604400543885e-05,
      "loss": 0.5285,
      "step": 71
    },
    {
      "epoch": 0.6900958466453674,
      "grad_norm": 0.6720920559775677,
      "learning_rate": 3.801937735804838e-05,
      "loss": 0.5226,
      "step": 72
    },
    {
      "epoch": 0.6996805111821086,
      "grad_norm": 0.869173718218821,
      "learning_rate": 3.792088191937451e-05,
      "loss": 0.5575,
      "step": 73
    },
    {
      "epoch": 0.7092651757188498,
      "grad_norm": 0.9252932692964674,
      "learning_rate": 3.782013048376736e-05,
      "loss": 0.5331,
      "step": 74
    },
    {
      "epoch": 0.7188498402555911,
      "grad_norm": 0.7617188628088486,
      "learning_rate": 3.77171357344716e-05,
      "loss": 0.4955,
      "step": 75
    },
    {
      "epoch": 0.7284345047923323,
      "grad_norm": 0.8768762141880961,
      "learning_rate": 3.761191063713476e-05,
      "loss": 0.5319,
      "step": 76
    },
    {
      "epoch": 0.7380191693290735,
      "grad_norm": 0.5856671731846674,
      "learning_rate": 3.7504468438175076e-05,
      "loss": 0.5286,
      "step": 77
    },
    {
      "epoch": 0.7476038338658147,
      "grad_norm": 0.7357152016449541,
      "learning_rate": 3.7394822663113915e-05,
      "loss": 0.5202,
      "step": 78
    },
    {
      "epoch": 0.7571884984025559,
      "grad_norm": 0.8421603022757149,
      "learning_rate": 3.72829871148731e-05,
      "loss": 0.5094,
      "step": 79
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 0.7711330057528273,
      "learning_rate": 3.716897587203733e-05,
      "loss": 0.5067,
      "step": 80
    },
    {
      "epoch": 0.7763578274760383,
      "grad_norm": 0.8061539763692872,
      "learning_rate": 3.705280328708185e-05,
      "loss": 0.5052,
      "step": 81
    },
    {
      "epoch": 0.7859424920127795,
      "grad_norm": 0.9052464984501254,
      "learning_rate": 3.6934483984565684e-05,
      "loss": 0.5396,
      "step": 82
    },
    {
      "epoch": 0.7955271565495208,
      "grad_norm": 0.7613495063825277,
      "learning_rate": 3.681403285929061e-05,
      "loss": 0.5372,
      "step": 83
    },
    {
      "epoch": 0.805111821086262,
      "grad_norm": 0.9112267984327458,
      "learning_rate": 3.669146507442606e-05,
      "loss": 0.5022,
      "step": 84
    },
    {
      "epoch": 0.8146964856230032,
      "grad_norm": 0.776862619807027,
      "learning_rate": 3.6566796059600334e-05,
      "loss": 0.4943,
      "step": 85
    },
    {
      "epoch": 0.8242811501597445,
      "grad_norm": 0.6857530550183694,
      "learning_rate": 3.644004150895821e-05,
      "loss": 0.5137,
      "step": 86
    },
    {
      "epoch": 0.8338658146964856,
      "grad_norm": 1.0001214073388247,
      "learning_rate": 3.631121737918521e-05,
      "loss": 0.5211,
      "step": 87
    },
    {
      "epoch": 0.8434504792332268,
      "grad_norm": 0.614100200634229,
      "learning_rate": 3.6180339887498953e-05,
      "loss": 0.5426,
      "step": 88
    },
    {
      "epoch": 0.853035143769968,
      "grad_norm": 0.7416467757559503,
      "learning_rate": 3.6047425509607566e-05,
      "loss": 0.4952,
      "step": 89
    },
    {
      "epoch": 0.8626198083067093,
      "grad_norm": 0.8216759125475441,
      "learning_rate": 3.591249097763562e-05,
      "loss": 0.5101,
      "step": 90
    },
    {
      "epoch": 0.8722044728434505,
      "grad_norm": 0.8783521934817947,
      "learning_rate": 3.5775553278017824e-05,
      "loss": 0.5298,
      "step": 91
    },
    {
      "epoch": 0.8817891373801917,
      "grad_norm": 1.0305280265606676,
      "learning_rate": 3.56366296493606e-05,
      "loss": 0.509,
      "step": 92
    },
    {
      "epoch": 0.8913738019169329,
      "grad_norm": 0.9394240012360305,
      "learning_rate": 3.5495737580272024e-05,
      "loss": 0.5078,
      "step": 93
    },
    {
      "epoch": 0.9009584664536742,
      "grad_norm": 0.9309348675119059,
      "learning_rate": 3.535289480716023e-05,
      "loss": 0.5207,
      "step": 94
    },
    {
      "epoch": 0.9105431309904153,
      "grad_norm": 0.9046584824077331,
      "learning_rate": 3.520811931200063e-05,
      "loss": 0.5141,
      "step": 95
    },
    {
      "epoch": 0.9201277955271565,
      "grad_norm": 0.9752162620950977,
      "learning_rate": 3.5061429320072225e-05,
      "loss": 0.509,
      "step": 96
    },
    {
      "epoch": 0.9297124600638977,
      "grad_norm": 0.6999864028990369,
      "learning_rate": 3.4912843297663315e-05,
      "loss": 0.5009,
      "step": 97
    },
    {
      "epoch": 0.939297124600639,
      "grad_norm": 0.8545396872964545,
      "learning_rate": 3.476237994974682e-05,
      "loss": 0.5094,
      "step": 98
    },
    {
      "epoch": 0.9488817891373802,
      "grad_norm": 0.7634132561961192,
      "learning_rate": 3.4610058217625554e-05,
      "loss": 0.5254,
      "step": 99
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 0.7968935260475134,
      "learning_rate": 3.4455897276547836e-05,
      "loss": 0.5103,
      "step": 100
    },
    {
      "epoch": 0.9680511182108626,
      "grad_norm": 0.8779383622683993,
      "learning_rate": 3.429991653329351e-05,
      "loss": 0.4962,
      "step": 101
    },
    {
      "epoch": 0.9776357827476039,
      "grad_norm": 0.8873245389104287,
      "learning_rate": 3.4142135623730954e-05,
      "loss": 0.4886,
      "step": 102
    },
    {
      "epoch": 0.987220447284345,
      "grad_norm": 0.6369979342430059,
      "learning_rate": 3.398257441034515e-05,
      "loss": 0.5153,
      "step": 103
    },
    {
      "epoch": 0.9968051118210862,
      "grad_norm": 0.7723810447745028,
      "learning_rate": 3.38212529797373e-05,
      "loss": 0.5153,
      "step": 104
    },
    {
      "epoch": 1.0063897763578276,
      "grad_norm": 1.3732822336657626,
      "learning_rate": 3.365819164009614e-05,
      "loss": 0.8678,
      "step": 105
    },
    {
      "epoch": 1.0159744408945688,
      "grad_norm": 0.6491143685994553,
      "learning_rate": 3.349341091864149e-05,
      "loss": 0.3831,
      "step": 106
    },
    {
      "epoch": 1.0255591054313098,
      "grad_norm": 0.7068223705318214,
      "learning_rate": 3.3326931559040084e-05,
      "loss": 0.4093,
      "step": 107
    },
    {
      "epoch": 1.035143769968051,
      "grad_norm": 0.7448924108254381,
      "learning_rate": 3.315877451879426e-05,
      "loss": 0.4151,
      "step": 108
    },
    {
      "epoch": 1.0447284345047922,
      "grad_norm": 0.6408577512276367,
      "learning_rate": 3.298896096660367e-05,
      "loss": 0.4591,
      "step": 109
    },
    {
      "epoch": 1.0543130990415335,
      "grad_norm": 0.7245321395554075,
      "learning_rate": 3.2817512279700486e-05,
      "loss": 0.3995,
      "step": 110
    },
    {
      "epoch": 1.0638977635782747,
      "grad_norm": 0.8050047056671462,
      "learning_rate": 3.26444500411582e-05,
      "loss": 0.4517,
      "step": 111
    },
    {
      "epoch": 1.073482428115016,
      "grad_norm": 0.5692172276488292,
      "learning_rate": 3.246979603717467e-05,
      "loss": 0.3451,
      "step": 112
    },
    {
      "epoch": 1.0830670926517572,
      "grad_norm": 0.7143343812753972,
      "learning_rate": 3.2293572254329546e-05,
      "loss": 0.4343,
      "step": 113
    },
    {
      "epoch": 1.0926517571884984,
      "grad_norm": 0.621683888292599,
      "learning_rate": 3.21158008768164e-05,
      "loss": 0.4192,
      "step": 114
    },
    {
      "epoch": 1.1022364217252396,
      "grad_norm": 0.7045370668340967,
      "learning_rate": 3.1936504283650076e-05,
      "loss": 0.4369,
      "step": 115
    },
    {
      "epoch": 1.1118210862619808,
      "grad_norm": 0.689049088523779,
      "learning_rate": 3.1755705045849465e-05,
      "loss": 0.3995,
      "step": 116
    },
    {
      "epoch": 1.121405750798722,
      "grad_norm": 0.6154467041297189,
      "learning_rate": 3.157342592359612e-05,
      "loss": 0.4309,
      "step": 117
    },
    {
      "epoch": 1.1309904153354633,
      "grad_norm": 0.6721872707368349,
      "learning_rate": 3.138968986336904e-05,
      "loss": 0.3839,
      "step": 118
    },
    {
      "epoch": 1.1405750798722045,
      "grad_norm": 0.6919770496482566,
      "learning_rate": 3.1204519995056056e-05,
      "loss": 0.4536,
      "step": 119
    },
    {
      "epoch": 1.1501597444089458,
      "grad_norm": 0.548860031455983,
      "learning_rate": 3.101793962904205e-05,
      "loss": 0.3669,
      "step": 120
    },
    {
      "epoch": 1.159744408945687,
      "grad_norm": 0.662490642467861,
      "learning_rate": 3.082997225327452e-05,
      "loss": 0.4413,
      "step": 121
    },
    {
      "epoch": 1.1693290734824282,
      "grad_norm": 0.6083191080476268,
      "learning_rate": 3.064064153030673e-05,
      "loss": 0.3873,
      "step": 122
    },
    {
      "epoch": 1.1789137380191694,
      "grad_norm": 0.6537310951440493,
      "learning_rate": 3.0449971294318977e-05,
      "loss": 0.4173,
      "step": 123
    },
    {
      "epoch": 1.1884984025559104,
      "grad_norm": 0.6521406167992244,
      "learning_rate": 3.0257985548118127e-05,
      "loss": 0.4252,
      "step": 124
    },
    {
      "epoch": 1.1980830670926517,
      "grad_norm": 0.6149682224088401,
      "learning_rate": 3.0064708460116007e-05,
      "loss": 0.4375,
      "step": 125
    },
    {
      "epoch": 1.207667731629393,
      "grad_norm": 0.5660735007468374,
      "learning_rate": 2.987016436128694e-05,
      "loss": 0.4002,
      "step": 126
    },
    {
      "epoch": 1.2172523961661341,
      "grad_norm": 0.6860797663501914,
      "learning_rate": 2.9674377742104798e-05,
      "loss": 0.4501,
      "step": 127
    },
    {
      "epoch": 1.2268370607028753,
      "grad_norm": 0.5579864217961026,
      "learning_rate": 2.9477373249459974e-05,
      "loss": 0.3881,
      "step": 128
    },
    {
      "epoch": 1.2364217252396166,
      "grad_norm": 0.7520213328572914,
      "learning_rate": 2.9279175683556684e-05,
      "loss": 0.4309,
      "step": 129
    },
    {
      "epoch": 1.2460063897763578,
      "grad_norm": 0.5836183639678406,
      "learning_rate": 2.9079809994790937e-05,
      "loss": 0.4493,
      "step": 130
    },
    {
      "epoch": 1.255591054313099,
      "grad_norm": 0.5454961635732396,
      "learning_rate": 2.8879301280609645e-05,
      "loss": 0.4248,
      "step": 131
    },
    {
      "epoch": 1.2651757188498403,
      "grad_norm": 0.5814578805359331,
      "learning_rate": 2.8677674782351164e-05,
      "loss": 0.4133,
      "step": 132
    },
    {
      "epoch": 1.2747603833865815,
      "grad_norm": 0.529954247172676,
      "learning_rate": 2.8474955882067776e-05,
      "loss": 0.4165,
      "step": 133
    },
    {
      "epoch": 1.2843450479233227,
      "grad_norm": 0.5428271229133367,
      "learning_rate": 2.8271170099330415e-05,
      "loss": 0.4439,
      "step": 134
    },
    {
      "epoch": 1.293929712460064,
      "grad_norm": 0.6847831612932209,
      "learning_rate": 2.8066343088016105e-05,
      "loss": 0.441,
      "step": 135
    },
    {
      "epoch": 1.3035143769968052,
      "grad_norm": 0.49210190834226275,
      "learning_rate": 2.7860500633078475e-05,
      "loss": 0.4165,
      "step": 136
    },
    {
      "epoch": 1.3130990415335464,
      "grad_norm": 0.6253827417860478,
      "learning_rate": 2.7653668647301797e-05,
      "loss": 0.4177,
      "step": 137
    },
    {
      "epoch": 1.3226837060702876,
      "grad_norm": 0.5902351365403176,
      "learning_rate": 2.7445873168038906e-05,
      "loss": 0.418,
      "step": 138
    },
    {
      "epoch": 1.3322683706070286,
      "grad_norm": 0.5490945288774098,
      "learning_rate": 2.7237140353933445e-05,
      "loss": 0.4352,
      "step": 139
    },
    {
      "epoch": 1.34185303514377,
      "grad_norm": 0.6497645956557252,
      "learning_rate": 2.7027496481626858e-05,
      "loss": 0.442,
      "step": 140
    },
    {
      "epoch": 1.351437699680511,
      "grad_norm": 0.542052646410243,
      "learning_rate": 2.68169679424505e-05,
      "loss": 0.3954,
      "step": 141
    },
    {
      "epoch": 1.3610223642172525,
      "grad_norm": 0.6469294947936336,
      "learning_rate": 2.6605581239103347e-05,
      "loss": 0.4324,
      "step": 142
    },
    {
      "epoch": 1.3706070287539935,
      "grad_norm": 0.6114843722995497,
      "learning_rate": 2.6393362982315632e-05,
      "loss": 0.4086,
      "step": 143
    },
    {
      "epoch": 1.3801916932907348,
      "grad_norm": 0.5302481010581994,
      "learning_rate": 2.618033988749895e-05,
      "loss": 0.3797,
      "step": 144
    },
    {
      "epoch": 1.389776357827476,
      "grad_norm": 0.6290020988684742,
      "learning_rate": 2.5966538771383124e-05,
      "loss": 0.4616,
      "step": 145
    },
    {
      "epoch": 1.3993610223642172,
      "grad_norm": 0.4520856565609043,
      "learning_rate": 2.5751986548640345e-05,
      "loss": 0.3691,
      "step": 146
    },
    {
      "epoch": 1.4089456869009584,
      "grad_norm": 0.6229237189143199,
      "learning_rate": 2.5536710228496986e-05,
      "loss": 0.4428,
      "step": 147
    },
    {
      "epoch": 1.4185303514376997,
      "grad_norm": 0.5985663011286853,
      "learning_rate": 2.5320736911333503e-05,
      "loss": 0.4101,
      "step": 148
    },
    {
      "epoch": 1.428115015974441,
      "grad_norm": 0.5602348014424229,
      "learning_rate": 2.5104093785272854e-05,
      "loss": 0.3799,
      "step": 149
    },
    {
      "epoch": 1.4376996805111821,
      "grad_norm": 0.5489525751898331,
      "learning_rate": 2.4886808122757882e-05,
      "loss": 0.4414,
      "step": 150
    },
    {
      "epoch": 1.4472843450479234,
      "grad_norm": 0.624773105379871,
      "learning_rate": 2.4668907277118114e-05,
      "loss": 0.381,
      "step": 151
    },
    {
      "epoch": 1.4568690095846646,
      "grad_norm": 0.5426555057346395,
      "learning_rate": 2.445041867912629e-05,
      "loss": 0.4539,
      "step": 152
    },
    {
      "epoch": 1.4664536741214058,
      "grad_norm": 0.55536633578182,
      "learning_rate": 2.423136983354526e-05,
      "loss": 0.3802,
      "step": 153
    },
    {
      "epoch": 1.476038338658147,
      "grad_norm": 0.4416725668124843,
      "learning_rate": 2.401178831566546e-05,
      "loss": 0.3748,
      "step": 154
    },
    {
      "epoch": 1.4856230031948883,
      "grad_norm": 0.6230029207177754,
      "learning_rate": 2.379170176783357e-05,
      "loss": 0.4455,
      "step": 155
    },
    {
      "epoch": 1.4952076677316293,
      "grad_norm": 0.42530667418850676,
      "learning_rate": 2.3571137895972735e-05,
      "loss": 0.3917,
      "step": 156
    },
    {
      "epoch": 1.5047923322683707,
      "grad_norm": 0.5190777897196971,
      "learning_rate": 2.335012446609473e-05,
      "loss": 0.3944,
      "step": 157
    },
    {
      "epoch": 1.5143769968051117,
      "grad_norm": 0.46663489738662295,
      "learning_rate": 2.312868930080462e-05,
      "loss": 0.4337,
      "step": 158
    },
    {
      "epoch": 1.5239616613418532,
      "grad_norm": 0.4797855409160743,
      "learning_rate": 2.2906860275798257e-05,
      "loss": 0.3447,
      "step": 159
    },
    {
      "epoch": 1.5335463258785942,
      "grad_norm": 0.48961810463783534,
      "learning_rate": 2.2684665316353112e-05,
      "loss": 0.4036,
      "step": 160
    },
    {
      "epoch": 1.5431309904153354,
      "grad_norm": 0.5015403228340246,
      "learning_rate": 2.246213239381286e-05,
      "loss": 0.4494,
      "step": 161
    },
    {
      "epoch": 1.5527156549520766,
      "grad_norm": 0.49826419920046305,
      "learning_rate": 2.2239289522066157e-05,
      "loss": 0.3662,
      "step": 162
    },
    {
      "epoch": 1.5623003194888179,
      "grad_norm": 0.5982999217769431,
      "learning_rate": 2.201616475402009e-05,
      "loss": 0.4716,
      "step": 163
    },
    {
      "epoch": 1.571884984025559,
      "grad_norm": 0.4282708816156415,
      "learning_rate": 2.179278617806867e-05,
      "loss": 0.3435,
      "step": 164
    },
    {
      "epoch": 1.5814696485623003,
      "grad_norm": 0.4734658108408669,
      "learning_rate": 2.1569181914556904e-05,
      "loss": 0.4212,
      "step": 165
    },
    {
      "epoch": 1.5910543130990416,
      "grad_norm": 0.49859752482578384,
      "learning_rate": 2.1345380112240796e-05,
      "loss": 0.425,
      "step": 166
    },
    {
      "epoch": 1.6006389776357828,
      "grad_norm": 0.47157462374026493,
      "learning_rate": 2.1121408944743838e-05,
      "loss": 0.4507,
      "step": 167
    },
    {
      "epoch": 1.610223642172524,
      "grad_norm": 0.4248405831748351,
      "learning_rate": 2.08972966070103e-05,
      "loss": 0.3602,
      "step": 168
    },
    {
      "epoch": 1.619808306709265,
      "grad_norm": 0.5006569891631065,
      "learning_rate": 2.0673071311755885e-05,
      "loss": 0.4462,
      "step": 169
    },
    {
      "epoch": 1.6293929712460065,
      "grad_norm": 0.43452414054342864,
      "learning_rate": 2.0448761285916103e-05,
      "loss": 0.3398,
      "step": 170
    },
    {
      "epoch": 1.6389776357827475,
      "grad_norm": 0.5593870865885154,
      "learning_rate": 2.022439476709292e-05,
      "loss": 0.4578,
      "step": 171
    },
    {
      "epoch": 1.648562300319489,
      "grad_norm": 0.48660923787373483,
      "learning_rate": 2e-05,
      "loss": 0.402,
      "step": 172
    },
    {
      "epoch": 1.65814696485623,
      "grad_norm": 0.47042892474320597,
      "learning_rate": 1.9775605232907085e-05,
      "loss": 0.3732,
      "step": 173
    },
    {
      "epoch": 1.6677316293929714,
      "grad_norm": 0.4637788148192955,
      "learning_rate": 1.9551238714083903e-05,
      "loss": 0.4164,
      "step": 174
    },
    {
      "epoch": 1.6773162939297124,
      "grad_norm": 0.5061846620195124,
      "learning_rate": 1.932692868824413e-05,
      "loss": 0.375,
      "step": 175
    },
    {
      "epoch": 1.6869009584664538,
      "grad_norm": 0.4389361045425349,
      "learning_rate": 1.910270339298971e-05,
      "loss": 0.4262,
      "step": 176
    },
    {
      "epoch": 1.6964856230031948,
      "grad_norm": 0.4781379704434354,
      "learning_rate": 1.8878591055256165e-05,
      "loss": 0.4094,
      "step": 177
    },
    {
      "epoch": 1.706070287539936,
      "grad_norm": 0.4719818469093975,
      "learning_rate": 1.8654619887759207e-05,
      "loss": 0.4398,
      "step": 178
    },
    {
      "epoch": 1.7156549520766773,
      "grad_norm": 0.4401976778891567,
      "learning_rate": 1.8430818085443106e-05,
      "loss": 0.3973,
      "step": 179
    },
    {
      "epoch": 1.7252396166134185,
      "grad_norm": 0.5070123943848236,
      "learning_rate": 1.8207213821931332e-05,
      "loss": 0.4068,
      "step": 180
    },
    {
      "epoch": 1.7348242811501597,
      "grad_norm": 0.4165356789504617,
      "learning_rate": 1.7983835245979914e-05,
      "loss": 0.4368,
      "step": 181
    },
    {
      "epoch": 1.744408945686901,
      "grad_norm": 0.5166587940461813,
      "learning_rate": 1.7760710477933846e-05,
      "loss": 0.4043,
      "step": 182
    },
    {
      "epoch": 1.7539936102236422,
      "grad_norm": 0.48269705466364915,
      "learning_rate": 1.7537867606187145e-05,
      "loss": 0.4118,
      "step": 183
    },
    {
      "epoch": 1.7635782747603834,
      "grad_norm": 0.47756180477333937,
      "learning_rate": 1.7315334683646898e-05,
      "loss": 0.42,
      "step": 184
    },
    {
      "epoch": 1.7731629392971247,
      "grad_norm": 0.4765695567575237,
      "learning_rate": 1.7093139724201753e-05,
      "loss": 0.4012,
      "step": 185
    },
    {
      "epoch": 1.7827476038338657,
      "grad_norm": 0.4681108878215689,
      "learning_rate": 1.687131069919538e-05,
      "loss": 0.41,
      "step": 186
    },
    {
      "epoch": 1.792332268370607,
      "grad_norm": 0.46037561395697485,
      "learning_rate": 1.6649875533905276e-05,
      "loss": 0.4043,
      "step": 187
    },
    {
      "epoch": 1.8019169329073481,
      "grad_norm": 0.419707165268871,
      "learning_rate": 1.642886210402727e-05,
      "loss": 0.3914,
      "step": 188
    },
    {
      "epoch": 1.8115015974440896,
      "grad_norm": 0.4619831244311854,
      "learning_rate": 1.620829823216643e-05,
      "loss": 0.4228,
      "step": 189
    },
    {
      "epoch": 1.8210862619808306,
      "grad_norm": 0.4361667434307363,
      "learning_rate": 1.5988211684334548e-05,
      "loss": 0.4047,
      "step": 190
    },
    {
      "epoch": 1.830670926517572,
      "grad_norm": 0.46079446208057,
      "learning_rate": 1.5768630166454746e-05,
      "loss": 0.4362,
      "step": 191
    },
    {
      "epoch": 1.840255591054313,
      "grad_norm": 0.4550050932615957,
      "learning_rate": 1.5549581320873715e-05,
      "loss": 0.4118,
      "step": 192
    },
    {
      "epoch": 1.8498402555910545,
      "grad_norm": 0.39293704953570213,
      "learning_rate": 1.53310927228819e-05,
      "loss": 0.407,
      "step": 193
    },
    {
      "epoch": 1.8594249201277955,
      "grad_norm": 0.4352517300933002,
      "learning_rate": 1.5113191877242116e-05,
      "loss": 0.3869,
      "step": 194
    },
    {
      "epoch": 1.8690095846645367,
      "grad_norm": 0.41513949179810855,
      "learning_rate": 1.4895906214727149e-05,
      "loss": 0.3755,
      "step": 195
    },
    {
      "epoch": 1.878594249201278,
      "grad_norm": 0.422581641804126,
      "learning_rate": 1.46792630886665e-05,
      "loss": 0.3707,
      "step": 196
    },
    {
      "epoch": 1.8881789137380192,
      "grad_norm": 0.4315158978443658,
      "learning_rate": 1.4463289771503015e-05,
      "loss": 0.4135,
      "step": 197
    },
    {
      "epoch": 1.8977635782747604,
      "grad_norm": 0.4439035469190941,
      "learning_rate": 1.4248013451359657e-05,
      "loss": 0.4299,
      "step": 198
    },
    {
      "epoch": 1.9073482428115016,
      "grad_norm": 0.38114332649556365,
      "learning_rate": 1.403346122861688e-05,
      "loss": 0.3986,
      "step": 199
    },
    {
      "epoch": 1.9169329073482428,
      "grad_norm": 0.4562893320122979,
      "learning_rate": 1.3819660112501054e-05,
      "loss": 0.4447,
      "step": 200
    },
    {
      "epoch": 1.926517571884984,
      "grad_norm": 0.41109636995050963,
      "learning_rate": 1.3606637017684375e-05,
      "loss": 0.4231,
      "step": 201
    },
    {
      "epoch": 1.9361022364217253,
      "grad_norm": 0.4108514256627659,
      "learning_rate": 1.3394418760896665e-05,
      "loss": 0.3919,
      "step": 202
    },
    {
      "epoch": 1.9456869009584663,
      "grad_norm": 0.4403482836797348,
      "learning_rate": 1.31830320575495e-05,
      "loss": 0.4004,
      "step": 203
    },
    {
      "epoch": 1.9552715654952078,
      "grad_norm": 0.46058376777126453,
      "learning_rate": 1.2972503518373145e-05,
      "loss": 0.4271,
      "step": 204
    },
    {
      "epoch": 1.9648562300319488,
      "grad_norm": 0.4549563896458377,
      "learning_rate": 1.2762859646066561e-05,
      "loss": 0.4155,
      "step": 205
    },
    {
      "epoch": 1.9744408945686902,
      "grad_norm": 0.47102947639065273,
      "learning_rate": 1.2554126831961097e-05,
      "loss": 0.3947,
      "step": 206
    },
    {
      "epoch": 1.9840255591054312,
      "grad_norm": 0.43255866772480006,
      "learning_rate": 1.2346331352698206e-05,
      "loss": 0.3965,
      "step": 207
    },
    {
      "epoch": 1.9936102236421727,
      "grad_norm": 0.6055961125329625,
      "learning_rate": 1.213949936692153e-05,
      "loss": 0.4308,
      "step": 208
    },
    {
      "epoch": 2.0031948881789137,
      "grad_norm": 0.8201258360584602,
      "learning_rate": 1.1933656911983901e-05,
      "loss": 0.5536,
      "step": 209
    },
    {
      "epoch": 2.012779552715655,
      "grad_norm": 0.5116622937539458,
      "learning_rate": 1.1728829900669592e-05,
      "loss": 0.3256,
      "step": 210
    },
    {
      "epoch": 2.022364217252396,
      "grad_norm": 0.6977084341791524,
      "learning_rate": 1.1525044117932227e-05,
      "loss": 0.3275,
      "step": 211
    },
    {
      "epoch": 2.0319488817891376,
      "grad_norm": 0.639393679219998,
      "learning_rate": 1.132232521764884e-05,
      "loss": 0.3009,
      "step": 212
    },
    {
      "epoch": 2.0415335463258786,
      "grad_norm": 0.4890609363134351,
      "learning_rate": 1.1120698719390362e-05,
      "loss": 0.3387,
      "step": 213
    },
    {
      "epoch": 2.0511182108626196,
      "grad_norm": 0.5588000853620018,
      "learning_rate": 1.0920190005209066e-05,
      "loss": 0.3105,
      "step": 214
    },
    {
      "epoch": 2.060702875399361,
      "grad_norm": 0.5036006567654044,
      "learning_rate": 1.0720824316443321e-05,
      "loss": 0.2901,
      "step": 215
    },
    {
      "epoch": 2.070287539936102,
      "grad_norm": 0.5915822002117069,
      "learning_rate": 1.0522626750540029e-05,
      "loss": 0.3493,
      "step": 216
    },
    {
      "epoch": 2.0798722044728435,
      "grad_norm": 0.4214007118584772,
      "learning_rate": 1.0325622257895205e-05,
      "loss": 0.2985,
      "step": 217
    },
    {
      "epoch": 2.0894568690095845,
      "grad_norm": 0.5295101270711986,
      "learning_rate": 1.0129835638713064e-05,
      "loss": 0.3166,
      "step": 218
    },
    {
      "epoch": 2.099041533546326,
      "grad_norm": 0.4906090405908181,
      "learning_rate": 9.935291539884e-06,
      "loss": 0.2843,
      "step": 219
    },
    {
      "epoch": 2.108626198083067,
      "grad_norm": 0.3938053903767187,
      "learning_rate": 9.74201445188188e-06,
      "loss": 0.3163,
      "step": 220
    },
    {
      "epoch": 2.1182108626198084,
      "grad_norm": 0.5012038895473048,
      "learning_rate": 9.550028705681024e-06,
      "loss": 0.3222,
      "step": 221
    },
    {
      "epoch": 2.1277955271565494,
      "grad_norm": 0.4321669775385389,
      "learning_rate": 9.359358469693272e-06,
      "loss": 0.2909,
      "step": 222
    },
    {
      "epoch": 2.137380191693291,
      "grad_norm": 0.4117178329226722,
      "learning_rate": 9.170027746725487e-06,
      "loss": 0.3146,
      "step": 223
    },
    {
      "epoch": 2.146964856230032,
      "grad_norm": 0.3793388916643685,
      "learning_rate": 8.982060370957953e-06,
      "loss": 0.3017,
      "step": 224
    },
    {
      "epoch": 2.1565495207667733,
      "grad_norm": 0.4535610427107228,
      "learning_rate": 8.795480004943946e-06,
      "loss": 0.3456,
      "step": 225
    },
    {
      "epoch": 2.1661341853035143,
      "grad_norm": 0.37244536401873557,
      "learning_rate": 8.610310136630962e-06,
      "loss": 0.2761,
      "step": 226
    },
    {
      "epoch": 2.1757188498402558,
      "grad_norm": 0.3905670548801518,
      "learning_rate": 8.426574076403887e-06,
      "loss": 0.302,
      "step": 227
    },
    {
      "epoch": 2.1853035143769968,
      "grad_norm": 0.4359210822644894,
      "learning_rate": 8.24429495415054e-06,
      "loss": 0.338,
      "step": 228
    },
    {
      "epoch": 2.194888178913738,
      "grad_norm": 0.35920435170508774,
      "learning_rate": 8.063495716349929e-06,
      "loss": 0.3081,
      "step": 229
    },
    {
      "epoch": 2.2044728434504792,
      "grad_norm": 0.37621369347346134,
      "learning_rate": 7.884199123183604e-06,
      "loss": 0.3419,
      "step": 230
    },
    {
      "epoch": 2.2140575079872207,
      "grad_norm": 0.3458946393058899,
      "learning_rate": 7.706427745670458e-06,
      "loss": 0.2749,
      "step": 231
    },
    {
      "epoch": 2.2236421725239617,
      "grad_norm": 0.4279444858711213,
      "learning_rate": 7.530203962825331e-06,
      "loss": 0.3365,
      "step": 232
    },
    {
      "epoch": 2.2332268370607027,
      "grad_norm": 0.3940413928415214,
      "learning_rate": 7.355549958841808e-06,
      "loss": 0.3282,
      "step": 233
    },
    {
      "epoch": 2.242811501597444,
      "grad_norm": 0.39687410854343524,
      "learning_rate": 7.182487720299518e-06,
      "loss": 0.2943,
      "step": 234
    },
    {
      "epoch": 2.252396166134185,
      "grad_norm": 0.42285032823633234,
      "learning_rate": 7.01103903339633e-06,
      "loss": 0.3252,
      "step": 235
    },
    {
      "epoch": 2.2619808306709266,
      "grad_norm": 0.3730435843457758,
      "learning_rate": 6.841225481205749e-06,
      "loss": 0.2755,
      "step": 236
    },
    {
      "epoch": 2.2715654952076676,
      "grad_norm": 0.35400532023912695,
      "learning_rate": 6.6730684409599225e-06,
      "loss": 0.2962,
      "step": 237
    },
    {
      "epoch": 2.281150159744409,
      "grad_norm": 0.39975466800088416,
      "learning_rate": 6.5065890813585145e-06,
      "loss": 0.2951,
      "step": 238
    },
    {
      "epoch": 2.29073482428115,
      "grad_norm": 0.4291346122689877,
      "learning_rate": 6.3418083599038624e-06,
      "loss": 0.317,
      "step": 239
    },
    {
      "epoch": 2.3003194888178915,
      "grad_norm": 0.3209273222551871,
      "learning_rate": 6.178747020262708e-06,
      "loss": 0.2881,
      "step": 240
    },
    {
      "epoch": 2.3099041533546325,
      "grad_norm": 0.3305467433463152,
      "learning_rate": 6.017425589654853e-06,
      "loss": 0.2595,
      "step": 241
    },
    {
      "epoch": 2.319488817891374,
      "grad_norm": 0.38111354545103865,
      "learning_rate": 5.857864376269051e-06,
      "loss": 0.326,
      "step": 242
    },
    {
      "epoch": 2.329073482428115,
      "grad_norm": 0.30011533883033215,
      "learning_rate": 5.700083466706494e-06,
      "loss": 0.2744,
      "step": 243
    },
    {
      "epoch": 2.3386581469648564,
      "grad_norm": 0.3813259748607558,
      "learning_rate": 5.544102723452171e-06,
      "loss": 0.3105,
      "step": 244
    },
    {
      "epoch": 2.3482428115015974,
      "grad_norm": 0.35123991131982374,
      "learning_rate": 5.38994178237445e-06,
      "loss": 0.316,
      "step": 245
    },
    {
      "epoch": 2.357827476038339,
      "grad_norm": 0.32456708820534624,
      "learning_rate": 5.237620050253189e-06,
      "loss": 0.3226,
      "step": 246
    },
    {
      "epoch": 2.36741214057508,
      "grad_norm": 0.2976838087469898,
      "learning_rate": 5.087156702336689e-06,
      "loss": 0.2614,
      "step": 247
    },
    {
      "epoch": 2.376996805111821,
      "grad_norm": 0.36552843906148524,
      "learning_rate": 4.938570679927784e-06,
      "loss": 0.3123,
      "step": 248
    },
    {
      "epoch": 2.3865814696485623,
      "grad_norm": 0.35095503173458703,
      "learning_rate": 4.791880687999382e-06,
      "loss": 0.3123,
      "step": 249
    },
    {
      "epoch": 2.3961661341853033,
      "grad_norm": 0.30625640312787195,
      "learning_rate": 4.647105192839778e-06,
      "loss": 0.3176,
      "step": 250
    },
    {
      "epoch": 2.405750798722045,
      "grad_norm": 0.341761533375559,
      "learning_rate": 4.504262419727983e-06,
      "loss": 0.3144,
      "step": 251
    },
    {
      "epoch": 2.415335463258786,
      "grad_norm": 0.3533342249857026,
      "learning_rate": 4.363370350639405e-06,
      "loss": 0.34,
      "step": 252
    },
    {
      "epoch": 2.4249201277955272,
      "grad_norm": 0.34883281084308354,
      "learning_rate": 4.2244467219821806e-06,
      "loss": 0.2955,
      "step": 253
    },
    {
      "epoch": 2.4345047923322682,
      "grad_norm": 0.32463619347820316,
      "learning_rate": 4.087509022364382e-06,
      "loss": 0.283,
      "step": 254
    },
    {
      "epoch": 2.4440894568690097,
      "grad_norm": 0.33770338798840704,
      "learning_rate": 3.952574490392443e-06,
      "loss": 0.3177,
      "step": 255
    },
    {
      "epoch": 2.4536741214057507,
      "grad_norm": 0.3110029435092563,
      "learning_rate": 3.819660112501053e-06,
      "loss": 0.302,
      "step": 256
    },
    {
      "epoch": 2.463258785942492,
      "grad_norm": 0.30800590517974463,
      "learning_rate": 3.6887826208147968e-06,
      "loss": 0.302,
      "step": 257
    },
    {
      "epoch": 2.472843450479233,
      "grad_norm": 0.3232681352037157,
      "learning_rate": 3.5599584910418037e-06,
      "loss": 0.2646,
      "step": 258
    },
    {
      "epoch": 2.4824281150159746,
      "grad_norm": 0.36421296660907726,
      "learning_rate": 3.433203940399672e-06,
      "loss": 0.3084,
      "step": 259
    },
    {
      "epoch": 2.4920127795527156,
      "grad_norm": 0.3831030312811903,
      "learning_rate": 3.3085349255739475e-06,
      "loss": 0.3469,
      "step": 260
    },
    {
      "epoch": 2.501597444089457,
      "grad_norm": 0.3325867758576896,
      "learning_rate": 3.1859671407093984e-06,
      "loss": 0.3093,
      "step": 261
    },
    {
      "epoch": 2.511182108626198,
      "grad_norm": 0.337706654419708,
      "learning_rate": 3.0655160154343177e-06,
      "loss": 0.2979,
      "step": 262
    },
    {
      "epoch": 2.520766773162939,
      "grad_norm": 0.32967438884893885,
      "learning_rate": 2.947196712918157e-06,
      "loss": 0.3047,
      "step": 263
    },
    {
      "epoch": 2.5303514376996805,
      "grad_norm": 0.3233433151169985,
      "learning_rate": 2.8310241279626784e-06,
      "loss": 0.2969,
      "step": 264
    },
    {
      "epoch": 2.539936102236422,
      "grad_norm": 0.31301087906685726,
      "learning_rate": 2.7170128851269084e-06,
      "loss": 0.2981,
      "step": 265
    },
    {
      "epoch": 2.549520766773163,
      "grad_norm": 0.32786149056284464,
      "learning_rate": 2.6051773368860935e-06,
      "loss": 0.3308,
      "step": 266
    },
    {
      "epoch": 2.559105431309904,
      "grad_norm": 0.31611139290200363,
      "learning_rate": 2.4955315618249263e-06,
      "loss": 0.3134,
      "step": 267
    },
    {
      "epoch": 2.5686900958466454,
      "grad_norm": 0.3202278737140233,
      "learning_rate": 2.38808936286524e-06,
      "loss": 0.2988,
      "step": 268
    },
    {
      "epoch": 2.5782747603833864,
      "grad_norm": 0.3187169467092785,
      "learning_rate": 2.2828642655284038e-06,
      "loss": 0.275,
      "step": 269
    },
    {
      "epoch": 2.587859424920128,
      "grad_norm": 0.31150312167717303,
      "learning_rate": 2.1798695162326444e-06,
      "loss": 0.2913,
      "step": 270
    },
    {
      "epoch": 2.597444089456869,
      "grad_norm": 0.3199602869029221,
      "learning_rate": 2.0791180806254975e-06,
      "loss": 0.3142,
      "step": 271
    },
    {
      "epoch": 2.6070287539936103,
      "grad_norm": 0.2859711870729773,
      "learning_rate": 1.9806226419516195e-06,
      "loss": 0.2994,
      "step": 272
    },
    {
      "epoch": 2.6166134185303513,
      "grad_norm": 0.31034616192733694,
      "learning_rate": 1.8843955994561191e-06,
      "loss": 0.3301,
      "step": 273
    },
    {
      "epoch": 2.626198083067093,
      "grad_norm": 0.2911594390655468,
      "learning_rate": 1.790449066823683e-06,
      "loss": 0.3136,
      "step": 274
    },
    {
      "epoch": 2.635782747603834,
      "grad_norm": 0.31846174945632705,
      "learning_rate": 1.6987948706536038e-06,
      "loss": 0.3265,
      "step": 275
    },
    {
      "epoch": 2.6453674121405752,
      "grad_norm": 0.3091371745510634,
      "learning_rate": 1.6094445489709886e-06,
      "loss": 0.3116,
      "step": 276
    },
    {
      "epoch": 2.6549520766773163,
      "grad_norm": 0.28744483162457096,
      "learning_rate": 1.5224093497742654e-06,
      "loss": 0.3081,
      "step": 277
    },
    {
      "epoch": 2.6645367412140573,
      "grad_norm": 0.3010104381158104,
      "learning_rate": 1.4377002296192233e-06,
      "loss": 0.2914,
      "step": 278
    },
    {
      "epoch": 2.6741214057507987,
      "grad_norm": 0.3184703269379148,
      "learning_rate": 1.3553278522397162e-06,
      "loss": 0.3398,
      "step": 279
    },
    {
      "epoch": 2.68370607028754,
      "grad_norm": 0.3012745996881748,
      "learning_rate": 1.275302587205256e-06,
      "loss": 0.2895,
      "step": 280
    },
    {
      "epoch": 2.693290734824281,
      "grad_norm": 0.32944896697723586,
      "learning_rate": 1.1976345086156193e-06,
      "loss": 0.3205,
      "step": 281
    },
    {
      "epoch": 2.702875399361022,
      "grad_norm": 0.30329250730790236,
      "learning_rate": 1.1223333938326486e-06,
      "loss": 0.307,
      "step": 282
    },
    {
      "epoch": 2.7124600638977636,
      "grad_norm": 0.31477266554231703,
      "learning_rate": 1.0494087222494253e-06,
      "loss": 0.2977,
      "step": 283
    },
    {
      "epoch": 2.722044728434505,
      "grad_norm": 0.27218751575717026,
      "learning_rate": 9.788696740969295e-07,
      "loss": 0.2727,
      "step": 284
    },
    {
      "epoch": 2.731629392971246,
      "grad_norm": 0.2957347661302385,
      "learning_rate": 9.107251292883856e-07,
      "loss": 0.3241,
      "step": 285
    },
    {
      "epoch": 2.741214057507987,
      "grad_norm": 0.2754457767022288,
      "learning_rate": 8.44983666301391e-07,
      "loss": 0.2888,
      "step": 286
    },
    {
      "epoch": 2.7507987220447285,
      "grad_norm": 0.3009061607925657,
      "learning_rate": 7.816535610980103e-07,
      "loss": 0.3236,
      "step": 287
    },
    {
      "epoch": 2.7603833865814695,
      "grad_norm": 0.3042692164321764,
      "learning_rate": 7.207427860829352e-07,
      "loss": 0.3293,
      "step": 288
    },
    {
      "epoch": 2.769968051118211,
      "grad_norm": 0.3084699080559463,
      "learning_rate": 6.622590090998727e-07,
      "loss": 0.2896,
      "step": 289
    },
    {
      "epoch": 2.779552715654952,
      "grad_norm": 0.2874746995613232,
      "learning_rate": 6.062095924662625e-07,
      "loss": 0.2977,
      "step": 290
    },
    {
      "epoch": 2.7891373801916934,
      "grad_norm": 0.30432287110598966,
      "learning_rate": 5.526015920464689e-07,
      "loss": 0.3527,
      "step": 291
    },
    {
      "epoch": 2.7987220447284344,
      "grad_norm": 0.2752619163287045,
      "learning_rate": 5.014417563635276e-07,
      "loss": 0.2575,
      "step": 292
    },
    {
      "epoch": 2.8083067092651754,
      "grad_norm": 0.3000832642958543,
      "learning_rate": 4.5273652574961745e-07,
      "loss": 0.3026,
      "step": 293
    },
    {
      "epoch": 2.817891373801917,
      "grad_norm": 0.28459336557371145,
      "learning_rate": 4.064920315352905e-07,
      "loss": 0.2953,
      "step": 294
    },
    {
      "epoch": 2.8274760383386583,
      "grad_norm": 0.3403049604495616,
      "learning_rate": 3.62714095277632e-07,
      "loss": 0.3698,
      "step": 295
    },
    {
      "epoch": 2.8370607028753994,
      "grad_norm": 0.28503876136131184,
      "learning_rate": 3.214082280274067e-07,
      "loss": 0.2709,
      "step": 296
    },
    {
      "epoch": 2.8466453674121404,
      "grad_norm": 0.299085461383954,
      "learning_rate": 2.825796296352823e-07,
      "loss": 0.3268,
      "step": 297
    },
    {
      "epoch": 2.856230031948882,
      "grad_norm": 0.3016256794104278,
      "learning_rate": 2.462331880972468e-07,
      "loss": 0.3032,
      "step": 298
    },
    {
      "epoch": 2.8658146964856233,
      "grad_norm": 0.45115419737664014,
      "learning_rate": 2.123734789392673e-07,
      "loss": 0.3539,
      "step": 299
    },
    {
      "epoch": 2.8753993610223643,
      "grad_norm": 0.2661772363746791,
      "learning_rate": 1.81004764641306e-07,
      "loss": 0.2696,
      "step": 300
    },
    {
      "epoch": 2.8849840255591053,
      "grad_norm": 0.2901415907606709,
      "learning_rate": 1.5213099410071873e-07,
      "loss": 0.3199,
      "step": 301
    },
    {
      "epoch": 2.8945686900958467,
      "grad_norm": 0.29953435371037257,
      "learning_rate": 1.2575580213514792e-07,
      "loss": 0.3403,
      "step": 302
    },
    {
      "epoch": 2.9041533546325877,
      "grad_norm": 0.28555862760239925,
      "learning_rate": 1.0188250902495312e-07,
      "loss": 0.3211,
      "step": 303
    },
    {
      "epoch": 2.913738019169329,
      "grad_norm": 0.3026928327893977,
      "learning_rate": 8.051412009521864e-08,
      "loss": 0.3422,
      "step": 304
    },
    {
      "epoch": 2.92332268370607,
      "grad_norm": 0.27586171376383506,
      "learning_rate": 6.165332533744072e-08,
      "loss": 0.2635,
      "step": 305
    },
    {
      "epoch": 2.9329073482428116,
      "grad_norm": 0.2797989784483068,
      "learning_rate": 4.530249907087836e-08,
      "loss": 0.2942,
      "step": 306
    },
    {
      "epoch": 2.9424920127795526,
      "grad_norm": 0.32210333279055925,
      "learning_rate": 3.146369964366791e-08,
      "loss": 0.3829,
      "step": 307
    },
    {
      "epoch": 2.952076677316294,
      "grad_norm": 0.27055679198155164,
      "learning_rate": 2.0138669173708213e-08,
      "loss": 0.2838,
      "step": 308
    },
    {
      "epoch": 2.961661341853035,
      "grad_norm": 0.28816307097851795,
      "learning_rate": 1.1328833329333767e-08,
      "loss": 0.3031,
      "step": 309
    },
    {
      "epoch": 2.9712460063897765,
      "grad_norm": 0.28899005094993563,
      "learning_rate": 5.0353011498693875e-09,
      "loss": 0.3185,
      "step": 310
    },
    {
      "epoch": 2.9808306709265175,
      "grad_norm": 0.26574618706072956,
      "learning_rate": 1.2588649059885883e-09,
      "loss": 0.2968,
      "step": 311
    },
    {
      "epoch": 2.9904153354632586,
      "grad_norm": 0.28718668323400676,
      "learning_rate": 0.0,
      "loss": 0.3009,
      "step": 312
    },
    {
      "epoch": 2.9904153354632586,
      "step": 312,
      "total_flos": 4.4115474684667494e+17,
      "train_loss": 0.43267787887881964,
      "train_runtime": 10181.5896,
      "train_samples_per_second": 2.946,
      "train_steps_per_second": 0.031
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 312,
  "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": 4.4115474684667494e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}