{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 4287,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005831583858175881,
      "grad_norm": 1.3819491863250732,
      "learning_rate": 5e-05,
      "loss": 8.7045,
      "step": 25
    },
    {
      "epoch": 0.011663167716351761,
      "grad_norm": 0.21646396815776825,
      "learning_rate": 0.0001,
      "loss": 0.0321,
      "step": 50
    },
    {
      "epoch": 0.01749475157452764,
      "grad_norm": 0.11650694161653519,
      "learning_rate": 9.940995987727166e-05,
      "loss": 0.014,
      "step": 75
    },
    {
      "epoch": 0.023326335432703522,
      "grad_norm": 0.12974420189857483,
      "learning_rate": 9.881991975454331e-05,
      "loss": 0.0099,
      "step": 100
    },
    {
      "epoch": 0.0291579192908794,
      "grad_norm": 0.16743576526641846,
      "learning_rate": 9.822987963181498e-05,
      "loss": 0.0101,
      "step": 125
    },
    {
      "epoch": 0.03498950314905528,
      "grad_norm": 0.2192072570323944,
      "learning_rate": 9.763983950908663e-05,
      "loss": 0.0095,
      "step": 150
    },
    {
      "epoch": 0.040821087007231166,
      "grad_norm": 0.1741335690021515,
      "learning_rate": 9.704979938635828e-05,
      "loss": 0.0096,
      "step": 175
    },
    {
      "epoch": 0.046652670865407045,
      "grad_norm": 0.11039887368679047,
      "learning_rate": 9.645975926362993e-05,
      "loss": 0.0072,
      "step": 200
    },
    {
      "epoch": 0.052484254723582924,
      "grad_norm": 0.16088618338108063,
      "learning_rate": 9.586971914090159e-05,
      "loss": 0.009,
      "step": 225
    },
    {
      "epoch": 0.0583158385817588,
      "grad_norm": 0.07560613751411438,
      "learning_rate": 9.527967901817324e-05,
      "loss": 0.0074,
      "step": 250
    },
    {
      "epoch": 0.06414742243993468,
      "grad_norm": 0.0758112445473671,
      "learning_rate": 9.46896388954449e-05,
      "loss": 0.0092,
      "step": 275
    },
    {
      "epoch": 0.06997900629811056,
      "grad_norm": 0.10088574141263962,
      "learning_rate": 9.409959877271654e-05,
      "loss": 0.008,
      "step": 300
    },
    {
      "epoch": 0.07581059015628645,
      "grad_norm": 0.09721095860004425,
      "learning_rate": 9.350955864998821e-05,
      "loss": 0.0072,
      "step": 325
    },
    {
      "epoch": 0.08164217401446233,
      "grad_norm": 0.11294808238744736,
      "learning_rate": 9.291951852725986e-05,
      "loss": 0.0069,
      "step": 350
    },
    {
      "epoch": 0.08747375787263821,
      "grad_norm": 0.1476767361164093,
      "learning_rate": 9.232947840453151e-05,
      "loss": 0.009,
      "step": 375
    },
    {
      "epoch": 0.09330534173081409,
      "grad_norm": 0.11318453401327133,
      "learning_rate": 9.173943828180317e-05,
      "loss": 0.0081,
      "step": 400
    },
    {
      "epoch": 0.09913692558898997,
      "grad_norm": 0.10059420764446259,
      "learning_rate": 9.114939815907482e-05,
      "loss": 0.0073,
      "step": 425
    },
    {
      "epoch": 0.10496850944716585,
      "grad_norm": 0.0688878670334816,
      "learning_rate": 9.055935803634647e-05,
      "loss": 0.0071,
      "step": 450
    },
    {
      "epoch": 0.11080009330534173,
      "grad_norm": 0.2254699021577835,
      "learning_rate": 8.996931791361814e-05,
      "loss": 0.0084,
      "step": 475
    },
    {
      "epoch": 0.1166316771635176,
      "grad_norm": 0.09738397598266602,
      "learning_rate": 8.937927779088978e-05,
      "loss": 0.0074,
      "step": 500
    },
    {
      "epoch": 0.1224632610216935,
      "grad_norm": 0.10569097846746445,
      "learning_rate": 8.878923766816144e-05,
      "loss": 0.0076,
      "step": 525
    },
    {
      "epoch": 0.12829484487986936,
      "grad_norm": 0.09912018477916718,
      "learning_rate": 8.81991975454331e-05,
      "loss": 0.0082,
      "step": 550
    },
    {
      "epoch": 0.13412642873804526,
      "grad_norm": 0.06858157366514206,
      "learning_rate": 8.760915742270475e-05,
      "loss": 0.008,
      "step": 575
    },
    {
      "epoch": 0.13995801259622112,
      "grad_norm": 0.09230218082666397,
      "learning_rate": 8.70191172999764e-05,
      "loss": 0.008,
      "step": 600
    },
    {
      "epoch": 0.145789596454397,
      "grad_norm": 0.09949750453233719,
      "learning_rate": 8.642907717724805e-05,
      "loss": 0.008,
      "step": 625
    },
    {
      "epoch": 0.1516211803125729,
      "grad_norm": 0.1029815599322319,
      "learning_rate": 8.58390370545197e-05,
      "loss": 0.008,
      "step": 650
    },
    {
      "epoch": 0.15745276417074877,
      "grad_norm": 0.05565624684095383,
      "learning_rate": 8.524899693179137e-05,
      "loss": 0.0076,
      "step": 675
    },
    {
      "epoch": 0.16328434802892466,
      "grad_norm": 0.07810457050800323,
      "learning_rate": 8.465895680906301e-05,
      "loss": 0.0082,
      "step": 700
    },
    {
      "epoch": 0.16911593188710053,
      "grad_norm": 0.16746339201927185,
      "learning_rate": 8.406891668633468e-05,
      "loss": 0.0076,
      "step": 725
    },
    {
      "epoch": 0.17494751574527642,
      "grad_norm": 0.2016323357820511,
      "learning_rate": 8.347887656360633e-05,
      "loss": 0.007,
      "step": 750
    },
    {
      "epoch": 0.1807790996034523,
      "grad_norm": 0.11883407831192017,
      "learning_rate": 8.288883644087798e-05,
      "loss": 0.0074,
      "step": 775
    },
    {
      "epoch": 0.18661068346162818,
      "grad_norm": 0.08053091168403625,
      "learning_rate": 8.229879631814964e-05,
      "loss": 0.0079,
      "step": 800
    },
    {
      "epoch": 0.19244226731980407,
      "grad_norm": 0.05473201349377632,
      "learning_rate": 8.170875619542129e-05,
      "loss": 0.0072,
      "step": 825
    },
    {
      "epoch": 0.19827385117797994,
      "grad_norm": 0.06678241491317749,
      "learning_rate": 8.111871607269294e-05,
      "loss": 0.0068,
      "step": 850
    },
    {
      "epoch": 0.20410543503615583,
      "grad_norm": 0.04147233068943024,
      "learning_rate": 8.05286759499646e-05,
      "loss": 0.0071,
      "step": 875
    },
    {
      "epoch": 0.2099370188943317,
      "grad_norm": 0.14302900433540344,
      "learning_rate": 7.993863582723626e-05,
      "loss": 0.0069,
      "step": 900
    },
    {
      "epoch": 0.2157686027525076,
      "grad_norm": 0.07474330812692642,
      "learning_rate": 7.934859570450791e-05,
      "loss": 0.0081,
      "step": 925
    },
    {
      "epoch": 0.22160018661068345,
      "grad_norm": 0.053421154618263245,
      "learning_rate": 7.875855558177956e-05,
      "loss": 0.0074,
      "step": 950
    },
    {
      "epoch": 0.22743177046885935,
      "grad_norm": 0.10754521191120148,
      "learning_rate": 7.816851545905122e-05,
      "loss": 0.0068,
      "step": 975
    },
    {
      "epoch": 0.2332633543270352,
      "grad_norm": 0.028845086693763733,
      "learning_rate": 7.757847533632287e-05,
      "loss": 0.0064,
      "step": 1000
    },
    {
      "epoch": 0.2390949381852111,
      "grad_norm": 0.08012504875659943,
      "learning_rate": 7.698843521359452e-05,
      "loss": 0.007,
      "step": 1025
    },
    {
      "epoch": 0.244926522043387,
      "grad_norm": 0.12295565754175186,
      "learning_rate": 7.639839509086619e-05,
      "loss": 0.0064,
      "step": 1050
    },
    {
      "epoch": 0.2507581059015629,
      "grad_norm": 0.07045774906873703,
      "learning_rate": 7.580835496813784e-05,
      "loss": 0.0077,
      "step": 1075
    },
    {
      "epoch": 0.2565896897597387,
      "grad_norm": 0.07725486904382706,
      "learning_rate": 7.521831484540949e-05,
      "loss": 0.0058,
      "step": 1100
    },
    {
      "epoch": 0.2624212736179146,
      "grad_norm": 0.062186215072870255,
      "learning_rate": 7.462827472268115e-05,
      "loss": 0.0082,
      "step": 1125
    },
    {
      "epoch": 0.2682528574760905,
      "grad_norm": 0.04699672758579254,
      "learning_rate": 7.403823459995281e-05,
      "loss": 0.0064,
      "step": 1150
    },
    {
      "epoch": 0.2740844413342664,
      "grad_norm": 0.0840052142739296,
      "learning_rate": 7.344819447722445e-05,
      "loss": 0.0083,
      "step": 1175
    },
    {
      "epoch": 0.27991602519244224,
      "grad_norm": 0.08768948912620544,
      "learning_rate": 7.285815435449612e-05,
      "loss": 0.0067,
      "step": 1200
    },
    {
      "epoch": 0.28574760905061813,
      "grad_norm": 0.11442771553993225,
      "learning_rate": 7.226811423176776e-05,
      "loss": 0.0068,
      "step": 1225
    },
    {
      "epoch": 0.291579192908794,
      "grad_norm": 0.04983456805348396,
      "learning_rate": 7.167807410903942e-05,
      "loss": 0.0079,
      "step": 1250
    },
    {
      "epoch": 0.2974107767669699,
      "grad_norm": 0.0609428770840168,
      "learning_rate": 7.108803398631107e-05,
      "loss": 0.0074,
      "step": 1275
    },
    {
      "epoch": 0.3032423606251458,
      "grad_norm": 0.04944339394569397,
      "learning_rate": 7.049799386358273e-05,
      "loss": 0.008,
      "step": 1300
    },
    {
      "epoch": 0.30907394448332165,
      "grad_norm": 0.07327156513929367,
      "learning_rate": 6.990795374085438e-05,
      "loss": 0.0079,
      "step": 1325
    },
    {
      "epoch": 0.31490552834149754,
      "grad_norm": 0.04584205895662308,
      "learning_rate": 6.931791361812605e-05,
      "loss": 0.0084,
      "step": 1350
    },
    {
      "epoch": 0.32073711219967344,
      "grad_norm": 0.02472819946706295,
      "learning_rate": 6.872787349539768e-05,
      "loss": 0.0068,
      "step": 1375
    },
    {
      "epoch": 0.32656869605784933,
      "grad_norm": 0.04464813694357872,
      "learning_rate": 6.813783337266935e-05,
      "loss": 0.0068,
      "step": 1400
    },
    {
      "epoch": 0.33240027991602517,
      "grad_norm": 0.1550447791814804,
      "learning_rate": 6.754779324994099e-05,
      "loss": 0.0085,
      "step": 1425
    },
    {
      "epoch": 0.33823186377420106,
      "grad_norm": 0.07317336648702621,
      "learning_rate": 6.695775312721266e-05,
      "loss": 0.0078,
      "step": 1450
    },
    {
      "epoch": 0.34406344763237695,
      "grad_norm": 0.08646785467863083,
      "learning_rate": 6.636771300448431e-05,
      "loss": 0.0089,
      "step": 1475
    },
    {
      "epoch": 0.34989503149055284,
      "grad_norm": 0.05935555696487427,
      "learning_rate": 6.577767288175596e-05,
      "loss": 0.0079,
      "step": 1500
    },
    {
      "epoch": 0.35572661534872874,
      "grad_norm": 0.045770954340696335,
      "learning_rate": 6.518763275902761e-05,
      "loss": 0.0075,
      "step": 1525
    },
    {
      "epoch": 0.3615581992069046,
      "grad_norm": 0.060902174562215805,
      "learning_rate": 6.459759263629928e-05,
      "loss": 0.0063,
      "step": 1550
    },
    {
      "epoch": 0.36738978306508047,
      "grad_norm": 0.11909040063619614,
      "learning_rate": 6.400755251357092e-05,
      "loss": 0.0068,
      "step": 1575
    },
    {
      "epoch": 0.37322136692325636,
      "grad_norm": 0.10495050251483917,
      "learning_rate": 6.341751239084258e-05,
      "loss": 0.0074,
      "step": 1600
    },
    {
      "epoch": 0.37905295078143225,
      "grad_norm": 0.08483859896659851,
      "learning_rate": 6.282747226811424e-05,
      "loss": 0.0066,
      "step": 1625
    },
    {
      "epoch": 0.38488453463960814,
      "grad_norm": 0.05936511233448982,
      "learning_rate": 6.223743214538589e-05,
      "loss": 0.0068,
      "step": 1650
    },
    {
      "epoch": 0.390716118497784,
      "grad_norm": 0.03783701732754707,
      "learning_rate": 6.164739202265754e-05,
      "loss": 0.0067,
      "step": 1675
    },
    {
      "epoch": 0.3965477023559599,
      "grad_norm": 0.03322950750589371,
      "learning_rate": 6.10573518999292e-05,
      "loss": 0.0067,
      "step": 1700
    },
    {
      "epoch": 0.40237928621413577,
      "grad_norm": 0.06542570888996124,
      "learning_rate": 6.0467311777200854e-05,
      "loss": 0.007,
      "step": 1725
    },
    {
      "epoch": 0.40821087007231166,
      "grad_norm": 0.05515117570757866,
      "learning_rate": 5.987727165447251e-05,
      "loss": 0.0069,
      "step": 1750
    },
    {
      "epoch": 0.4140424539304875,
      "grad_norm": 0.06597180664539337,
      "learning_rate": 5.928723153174416e-05,
      "loss": 0.0066,
      "step": 1775
    },
    {
      "epoch": 0.4198740377886634,
      "grad_norm": 0.14175090193748474,
      "learning_rate": 5.869719140901582e-05,
      "loss": 0.0072,
      "step": 1800
    },
    {
      "epoch": 0.4257056216468393,
      "grad_norm": 0.08393794298171997,
      "learning_rate": 5.810715128628748e-05,
      "loss": 0.007,
      "step": 1825
    },
    {
      "epoch": 0.4315372055050152,
      "grad_norm": 0.05573498085141182,
      "learning_rate": 5.751711116355912e-05,
      "loss": 0.0079,
      "step": 1850
    },
    {
      "epoch": 0.43736878936319107,
      "grad_norm": 0.08545993268489838,
      "learning_rate": 5.692707104083078e-05,
      "loss": 0.0073,
      "step": 1875
    },
    {
      "epoch": 0.4432003732213669,
      "grad_norm": 0.05015942454338074,
      "learning_rate": 5.633703091810243e-05,
      "loss": 0.0078,
      "step": 1900
    },
    {
      "epoch": 0.4490319570795428,
      "grad_norm": 0.08075055480003357,
      "learning_rate": 5.574699079537409e-05,
      "loss": 0.0078,
      "step": 1925
    },
    {
      "epoch": 0.4548635409377187,
      "grad_norm": 0.027930203825235367,
      "learning_rate": 5.515695067264575e-05,
      "loss": 0.0068,
      "step": 1950
    },
    {
      "epoch": 0.4606951247958946,
      "grad_norm": 0.12208261340856552,
      "learning_rate": 5.456691054991739e-05,
      "loss": 0.0076,
      "step": 1975
    },
    {
      "epoch": 0.4665267086540704,
      "grad_norm": 0.07081254571676254,
      "learning_rate": 5.397687042718905e-05,
      "loss": 0.0073,
      "step": 2000
    },
    {
      "epoch": 0.4723582925122463,
      "grad_norm": 0.06605388969182968,
      "learning_rate": 5.338683030446071e-05,
      "loss": 0.0067,
      "step": 2025
    },
    {
      "epoch": 0.4781898763704222,
      "grad_norm": 0.07699355483055115,
      "learning_rate": 5.279679018173236e-05,
      "loss": 0.0074,
      "step": 2050
    },
    {
      "epoch": 0.4840214602285981,
      "grad_norm": 0.07484355568885803,
      "learning_rate": 5.2206750059004016e-05,
      "loss": 0.008,
      "step": 2075
    },
    {
      "epoch": 0.489853044086774,
      "grad_norm": 0.05196300894021988,
      "learning_rate": 5.161670993627566e-05,
      "loss": 0.0086,
      "step": 2100
    },
    {
      "epoch": 0.49568462794494983,
      "grad_norm": 0.07191181927919388,
      "learning_rate": 5.102666981354732e-05,
      "loss": 0.0063,
      "step": 2125
    },
    {
      "epoch": 0.5015162118031258,
      "grad_norm": 0.12823079526424408,
      "learning_rate": 5.043662969081898e-05,
      "loss": 0.0065,
      "step": 2150
    },
    {
      "epoch": 0.5073477956613016,
      "grad_norm": 0.05935099348425865,
      "learning_rate": 4.9846589568090633e-05,
      "loss": 0.0072,
      "step": 2175
    },
    {
      "epoch": 0.5131793795194775,
      "grad_norm": 0.07369107007980347,
      "learning_rate": 4.9256549445362286e-05,
      "loss": 0.0071,
      "step": 2200
    },
    {
      "epoch": 0.5190109633776534,
      "grad_norm": 0.05355490744113922,
      "learning_rate": 4.866650932263394e-05,
      "loss": 0.007,
      "step": 2225
    },
    {
      "epoch": 0.5248425472358292,
      "grad_norm": 0.0854572206735611,
      "learning_rate": 4.807646919990559e-05,
      "loss": 0.0069,
      "step": 2250
    },
    {
      "epoch": 0.5306741310940052,
      "grad_norm": 0.058111775666475296,
      "learning_rate": 4.748642907717725e-05,
      "loss": 0.0074,
      "step": 2275
    },
    {
      "epoch": 0.536505714952181,
      "grad_norm": 0.06656444817781448,
      "learning_rate": 4.68963889544489e-05,
      "loss": 0.0064,
      "step": 2300
    },
    {
      "epoch": 0.5423372988103569,
      "grad_norm": 0.08224783837795258,
      "learning_rate": 4.6306348831720555e-05,
      "loss": 0.0082,
      "step": 2325
    },
    {
      "epoch": 0.5481688826685328,
      "grad_norm": 0.06290777772665024,
      "learning_rate": 4.5716308708992215e-05,
      "loss": 0.0066,
      "step": 2350
    },
    {
      "epoch": 0.5540004665267086,
      "grad_norm": 0.046868737787008286,
      "learning_rate": 4.512626858626387e-05,
      "loss": 0.0064,
      "step": 2375
    },
    {
      "epoch": 0.5598320503848845,
      "grad_norm": 0.08161591738462448,
      "learning_rate": 4.453622846353552e-05,
      "loss": 0.0086,
      "step": 2400
    },
    {
      "epoch": 0.5656636342430604,
      "grad_norm": 0.06194046139717102,
      "learning_rate": 4.394618834080718e-05,
      "loss": 0.0071,
      "step": 2425
    },
    {
      "epoch": 0.5714952181012363,
      "grad_norm": 0.06799814105033875,
      "learning_rate": 4.335614821807883e-05,
      "loss": 0.0065,
      "step": 2450
    },
    {
      "epoch": 0.5773268019594122,
      "grad_norm": 0.06985451281070709,
      "learning_rate": 4.276610809535049e-05,
      "loss": 0.0096,
      "step": 2475
    },
    {
      "epoch": 0.583158385817588,
      "grad_norm": 0.045887541025877,
      "learning_rate": 4.2176067972622144e-05,
      "loss": 0.0066,
      "step": 2500
    },
    {
      "epoch": 0.5889899696757639,
      "grad_norm": 0.10378813743591309,
      "learning_rate": 4.1586027849893796e-05,
      "loss": 0.0066,
      "step": 2525
    },
    {
      "epoch": 0.5948215535339398,
      "grad_norm": 0.037475138902664185,
      "learning_rate": 4.099598772716545e-05,
      "loss": 0.0077,
      "step": 2550
    },
    {
      "epoch": 0.6006531373921157,
      "grad_norm": 0.13552173972129822,
      "learning_rate": 4.040594760443711e-05,
      "loss": 0.0082,
      "step": 2575
    },
    {
      "epoch": 0.6064847212502916,
      "grad_norm": 0.049305260181427,
      "learning_rate": 3.981590748170876e-05,
      "loss": 0.0063,
      "step": 2600
    },
    {
      "epoch": 0.6123163051084675,
      "grad_norm": 0.0553220734000206,
      "learning_rate": 3.922586735898041e-05,
      "loss": 0.0079,
      "step": 2625
    },
    {
      "epoch": 0.6181478889666433,
      "grad_norm": 0.043462011963129044,
      "learning_rate": 3.8635827236252066e-05,
      "loss": 0.0067,
      "step": 2650
    },
    {
      "epoch": 0.6239794728248192,
      "grad_norm": 0.331167995929718,
      "learning_rate": 3.8045787113523725e-05,
      "loss": 0.008,
      "step": 2675
    },
    {
      "epoch": 0.6298110566829951,
      "grad_norm": 0.06077446788549423,
      "learning_rate": 3.745574699079538e-05,
      "loss": 0.0067,
      "step": 2700
    },
    {
      "epoch": 0.635642640541171,
      "grad_norm": 0.06422346830368042,
      "learning_rate": 3.686570686806703e-05,
      "loss": 0.0072,
      "step": 2725
    },
    {
      "epoch": 0.6414742243993469,
      "grad_norm": 0.053520068526268005,
      "learning_rate": 3.627566674533868e-05,
      "loss": 0.0057,
      "step": 2750
    },
    {
      "epoch": 0.6473058082575227,
      "grad_norm": 0.08369456231594086,
      "learning_rate": 3.568562662261034e-05,
      "loss": 0.0071,
      "step": 2775
    },
    {
      "epoch": 0.6531373921156987,
      "grad_norm": 0.07306842505931854,
      "learning_rate": 3.5095586499881994e-05,
      "loss": 0.0065,
      "step": 2800
    },
    {
      "epoch": 0.6589689759738745,
      "grad_norm": 0.04199352487921715,
      "learning_rate": 3.450554637715365e-05,
      "loss": 0.007,
      "step": 2825
    },
    {
      "epoch": 0.6648005598320503,
      "grad_norm": 0.04420465603470802,
      "learning_rate": 3.39155062544253e-05,
      "loss": 0.0064,
      "step": 2850
    },
    {
      "epoch": 0.6706321436902263,
      "grad_norm": 0.09053666144609451,
      "learning_rate": 3.332546613169696e-05,
      "loss": 0.0057,
      "step": 2875
    },
    {
      "epoch": 0.6764637275484021,
      "grad_norm": 0.04140497371554375,
      "learning_rate": 3.273542600896861e-05,
      "loss": 0.0067,
      "step": 2900
    },
    {
      "epoch": 0.6822953114065781,
      "grad_norm": 0.038969893008470535,
      "learning_rate": 3.2145385886240264e-05,
      "loss": 0.0059,
      "step": 2925
    },
    {
      "epoch": 0.6881268952647539,
      "grad_norm": 0.05273483693599701,
      "learning_rate": 3.1555345763511916e-05,
      "loss": 0.0065,
      "step": 2950
    },
    {
      "epoch": 0.6939584791229297,
      "grad_norm": 0.0726546123623848,
      "learning_rate": 3.0965305640783576e-05,
      "loss": 0.0061,
      "step": 2975
    },
    {
      "epoch": 0.6997900629811057,
      "grad_norm": 0.07580599933862686,
      "learning_rate": 3.0375265518055228e-05,
      "loss": 0.0068,
      "step": 3000
    },
    {
      "epoch": 0.7056216468392815,
      "grad_norm": 0.036154165863990784,
      "learning_rate": 2.978522539532688e-05,
      "loss": 0.0078,
      "step": 3025
    },
    {
      "epoch": 0.7114532306974575,
      "grad_norm": 0.07397032529115677,
      "learning_rate": 2.919518527259854e-05,
      "loss": 0.006,
      "step": 3050
    },
    {
      "epoch": 0.7172848145556333,
      "grad_norm": 0.11983916163444519,
      "learning_rate": 2.8605145149870193e-05,
      "loss": 0.0067,
      "step": 3075
    },
    {
      "epoch": 0.7231163984138091,
      "grad_norm": 0.0615057498216629,
      "learning_rate": 2.801510502714185e-05,
      "loss": 0.0079,
      "step": 3100
    },
    {
      "epoch": 0.7289479822719851,
      "grad_norm": 0.03690054267644882,
      "learning_rate": 2.74250649044135e-05,
      "loss": 0.0067,
      "step": 3125
    },
    {
      "epoch": 0.7347795661301609,
      "grad_norm": 0.07274270802736282,
      "learning_rate": 2.6835024781685157e-05,
      "loss": 0.0073,
      "step": 3150
    },
    {
      "epoch": 0.7406111499883369,
      "grad_norm": 0.0663859024643898,
      "learning_rate": 2.6244984658956813e-05,
      "loss": 0.0071,
      "step": 3175
    },
    {
      "epoch": 0.7464427338465127,
      "grad_norm": 0.05553248152136803,
      "learning_rate": 2.5654944536228466e-05,
      "loss": 0.0075,
      "step": 3200
    },
    {
      "epoch": 0.7522743177046886,
      "grad_norm": 0.04285744950175285,
      "learning_rate": 2.5064904413500118e-05,
      "loss": 0.0066,
      "step": 3225
    },
    {
      "epoch": 0.7581059015628645,
      "grad_norm": 0.09008394181728363,
      "learning_rate": 2.4474864290771774e-05,
      "loss": 0.0076,
      "step": 3250
    },
    {
      "epoch": 0.7639374854210403,
      "grad_norm": 0.058279018849134445,
      "learning_rate": 2.388482416804343e-05,
      "loss": 0.0075,
      "step": 3275
    },
    {
      "epoch": 0.7697690692792163,
      "grad_norm": 0.05219094827771187,
      "learning_rate": 2.3294784045315082e-05,
      "loss": 0.0071,
      "step": 3300
    },
    {
      "epoch": 0.7756006531373921,
      "grad_norm": 0.05435522645711899,
      "learning_rate": 2.270474392258674e-05,
      "loss": 0.006,
      "step": 3325
    },
    {
      "epoch": 0.781432236995568,
      "grad_norm": 0.05999380722641945,
      "learning_rate": 2.211470379985839e-05,
      "loss": 0.0084,
      "step": 3350
    },
    {
      "epoch": 0.7872638208537439,
      "grad_norm": 0.029552578926086426,
      "learning_rate": 2.1524663677130047e-05,
      "loss": 0.0061,
      "step": 3375
    },
    {
      "epoch": 0.7930954047119197,
      "grad_norm": 0.039268866181373596,
      "learning_rate": 2.09346235544017e-05,
      "loss": 0.0069,
      "step": 3400
    },
    {
      "epoch": 0.7989269885700956,
      "grad_norm": 0.05782666802406311,
      "learning_rate": 2.0344583431673355e-05,
      "loss": 0.0067,
      "step": 3425
    },
    {
      "epoch": 0.8047585724282715,
      "grad_norm": 0.03869437053799629,
      "learning_rate": 1.9754543308945008e-05,
      "loss": 0.0066,
      "step": 3450
    },
    {
      "epoch": 0.8105901562864474,
      "grad_norm": 0.05132381245493889,
      "learning_rate": 1.9164503186216664e-05,
      "loss": 0.0064,
      "step": 3475
    },
    {
      "epoch": 0.8164217401446233,
      "grad_norm": 0.08721393346786499,
      "learning_rate": 1.8574463063488316e-05,
      "loss": 0.0084,
      "step": 3500
    },
    {
      "epoch": 0.8222533240027992,
      "grad_norm": 0.08003130555152893,
      "learning_rate": 1.7984422940759972e-05,
      "loss": 0.0058,
      "step": 3525
    },
    {
      "epoch": 0.828084907860975,
      "grad_norm": 0.05190620571374893,
      "learning_rate": 1.7394382818031625e-05,
      "loss": 0.0066,
      "step": 3550
    },
    {
      "epoch": 0.8339164917191509,
      "grad_norm": 0.06803402304649353,
      "learning_rate": 1.680434269530328e-05,
      "loss": 0.0069,
      "step": 3575
    },
    {
      "epoch": 0.8397480755773268,
      "grad_norm": 0.05517745763063431,
      "learning_rate": 1.6214302572574937e-05,
      "loss": 0.0086,
      "step": 3600
    },
    {
      "epoch": 0.8455796594355027,
      "grad_norm": 0.05708625167608261,
      "learning_rate": 1.5624262449846593e-05,
      "loss": 0.0063,
      "step": 3625
    },
    {
      "epoch": 0.8514112432936786,
      "grad_norm": 0.05021214112639427,
      "learning_rate": 1.5034222327118243e-05,
      "loss": 0.0056,
      "step": 3650
    },
    {
      "epoch": 0.8572428271518544,
      "grad_norm": 0.06552550941705704,
      "learning_rate": 1.44441822043899e-05,
      "loss": 0.0065,
      "step": 3675
    },
    {
      "epoch": 0.8630744110100304,
      "grad_norm": 0.05650312826037407,
      "learning_rate": 1.3854142081661554e-05,
      "loss": 0.0071,
      "step": 3700
    },
    {
      "epoch": 0.8689059948682062,
      "grad_norm": 0.021646451205015182,
      "learning_rate": 1.326410195893321e-05,
      "loss": 0.0056,
      "step": 3725
    },
    {
      "epoch": 0.8747375787263821,
      "grad_norm": 0.04382350668311119,
      "learning_rate": 1.2674061836204862e-05,
      "loss": 0.0068,
      "step": 3750
    },
    {
      "epoch": 0.880569162584558,
      "grad_norm": 0.06425108760595322,
      "learning_rate": 1.2084021713476516e-05,
      "loss": 0.008,
      "step": 3775
    },
    {
      "epoch": 0.8864007464427338,
      "grad_norm": 0.07273637503385544,
      "learning_rate": 1.1493981590748172e-05,
      "loss": 0.0065,
      "step": 3800
    },
    {
      "epoch": 0.8922323303009098,
      "grad_norm": 0.0765676200389862,
      "learning_rate": 1.0903941468019826e-05,
      "loss": 0.006,
      "step": 3825
    },
    {
      "epoch": 0.8980639141590856,
      "grad_norm": 0.05742628127336502,
      "learning_rate": 1.031390134529148e-05,
      "loss": 0.007,
      "step": 3850
    },
    {
      "epoch": 0.9038954980172614,
      "grad_norm": 0.059533748775720596,
      "learning_rate": 9.723861222563135e-06,
      "loss": 0.0064,
      "step": 3875
    },
    {
      "epoch": 0.9097270818754374,
      "grad_norm": 0.02601277269423008,
      "learning_rate": 9.13382109983479e-06,
      "loss": 0.0061,
      "step": 3900
    },
    {
      "epoch": 0.9155586657336132,
      "grad_norm": 0.04568032547831535,
      "learning_rate": 8.543780977106443e-06,
      "loss": 0.0064,
      "step": 3925
    },
    {
      "epoch": 0.9213902495917892,
      "grad_norm": 0.04665420576930046,
      "learning_rate": 7.9537408543781e-06,
      "loss": 0.0072,
      "step": 3950
    },
    {
      "epoch": 0.927221833449965,
      "grad_norm": 0.04735620319843292,
      "learning_rate": 7.363700731649753e-06,
      "loss": 0.0072,
      "step": 3975
    },
    {
      "epoch": 0.9330534173081408,
      "grad_norm": 0.09656576067209244,
      "learning_rate": 6.773660608921408e-06,
      "loss": 0.0068,
      "step": 4000
    },
    {
      "epoch": 0.9388850011663168,
      "grad_norm": 0.05522262305021286,
      "learning_rate": 6.183620486193061e-06,
      "loss": 0.0056,
      "step": 4025
    },
    {
      "epoch": 0.9447165850244926,
      "grad_norm": 0.05257203057408333,
      "learning_rate": 5.5935803634647154e-06,
      "loss": 0.0056,
      "step": 4050
    },
    {
      "epoch": 0.9505481688826686,
      "grad_norm": 0.04314320161938667,
      "learning_rate": 5.00354024073637e-06,
      "loss": 0.0063,
      "step": 4075
    },
    {
      "epoch": 0.9563797527408444,
      "grad_norm": 0.04623764008283615,
      "learning_rate": 4.413500118008025e-06,
      "loss": 0.0067,
      "step": 4100
    },
    {
      "epoch": 0.9622113365990203,
      "grad_norm": 0.05779367685317993,
      "learning_rate": 3.823459995279679e-06,
      "loss": 0.007,
      "step": 4125
    },
    {
      "epoch": 0.9680429204571962,
      "grad_norm": 0.07973254472017288,
      "learning_rate": 3.2334198725513332e-06,
      "loss": 0.0068,
      "step": 4150
    },
    {
      "epoch": 0.973874504315372,
      "grad_norm": 0.056475065648555756,
      "learning_rate": 2.643379749822988e-06,
      "loss": 0.0056,
      "step": 4175
    },
    {
      "epoch": 0.979706088173548,
      "grad_norm": 0.05260898917913437,
      "learning_rate": 2.0533396270946426e-06,
      "loss": 0.0061,
      "step": 4200
    },
    {
      "epoch": 0.9855376720317238,
      "grad_norm": 0.07043284177780151,
      "learning_rate": 1.463299504366297e-06,
      "loss": 0.0061,
      "step": 4225
    },
    {
      "epoch": 0.9913692558898997,
      "grad_norm": 0.09899988770484924,
      "learning_rate": 8.732593816379515e-07,
      "loss": 0.0076,
      "step": 4250
    },
    {
      "epoch": 0.9972008397480756,
      "grad_norm": 0.03969600424170494,
      "learning_rate": 2.832192589096059e-07,
      "loss": 0.0067,
      "step": 4275
    }
  ],
  "logging_steps": 25,
  "max_steps": 4287,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 250,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.927104921604435e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}