{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 6935,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.021629416005767843,
      "grad_norm": 76.04898834228516,
      "learning_rate": 4.978370583994232e-05,
      "loss": 1.1957,
      "step": 30
    },
    {
      "epoch": 0.043258832011535686,
      "grad_norm": 15.82689094543457,
      "learning_rate": 4.956741167988465e-05,
      "loss": 0.1664,
      "step": 60
    },
    {
      "epoch": 0.06488824801730353,
      "grad_norm": 0.9152764678001404,
      "learning_rate": 4.935111751982697e-05,
      "loss": 0.0335,
      "step": 90
    },
    {
      "epoch": 0.08651766402307137,
      "grad_norm": 0.8367058634757996,
      "learning_rate": 4.913482335976929e-05,
      "loss": 0.0071,
      "step": 120
    },
    {
      "epoch": 0.10814708002883922,
      "grad_norm": 0.9554459452629089,
      "learning_rate": 4.891852919971161e-05,
      "loss": 0.0019,
      "step": 150
    },
    {
      "epoch": 0.12977649603460706,
      "grad_norm": 0.09180067479610443,
      "learning_rate": 4.8702235039653934e-05,
      "loss": 0.009,
      "step": 180
    },
    {
      "epoch": 0.1514059120403749,
      "grad_norm": 0.45300760865211487,
      "learning_rate": 4.8485940879596254e-05,
      "loss": 0.002,
      "step": 210
    },
    {
      "epoch": 0.17303532804614274,
      "grad_norm": 0.018595978617668152,
      "learning_rate": 4.826964671953857e-05,
      "loss": 0.0038,
      "step": 240
    },
    {
      "epoch": 0.1946647440519106,
      "grad_norm": 0.014309559017419815,
      "learning_rate": 4.80533525594809e-05,
      "loss": 0.0024,
      "step": 270
    },
    {
      "epoch": 0.21629416005767843,
      "grad_norm": 0.2507375180721283,
      "learning_rate": 4.783705839942322e-05,
      "loss": 0.009,
      "step": 300
    },
    {
      "epoch": 0.23792357606344627,
      "grad_norm": 0.031992267817258835,
      "learning_rate": 4.762076423936554e-05,
      "loss": 0.0095,
      "step": 330
    },
    {
      "epoch": 0.2595529920692141,
      "grad_norm": 0.04604390636086464,
      "learning_rate": 4.740447007930786e-05,
      "loss": 0.0024,
      "step": 360
    },
    {
      "epoch": 0.28118240807498196,
      "grad_norm": 0.03024614229798317,
      "learning_rate": 4.7188175919250185e-05,
      "loss": 0.0403,
      "step": 390
    },
    {
      "epoch": 0.3028118240807498,
      "grad_norm": 0.3150508403778076,
      "learning_rate": 4.6971881759192505e-05,
      "loss": 0.0028,
      "step": 420
    },
    {
      "epoch": 0.32444124008651765,
      "grad_norm": 0.11980323493480682,
      "learning_rate": 4.6755587599134824e-05,
      "loss": 0.0023,
      "step": 450
    },
    {
      "epoch": 0.3460706560922855,
      "grad_norm": 0.5558916926383972,
      "learning_rate": 4.6539293439077144e-05,
      "loss": 0.0117,
      "step": 480
    },
    {
      "epoch": 0.36770007209805333,
      "grad_norm": 0.1800965517759323,
      "learning_rate": 4.632299927901947e-05,
      "loss": 0.0029,
      "step": 510
    },
    {
      "epoch": 0.3893294881038212,
      "grad_norm": 0.05012692138552666,
      "learning_rate": 4.610670511896179e-05,
      "loss": 0.0104,
      "step": 540
    },
    {
      "epoch": 0.410958904109589,
      "grad_norm": 0.38384830951690674,
      "learning_rate": 4.589041095890411e-05,
      "loss": 0.0065,
      "step": 570
    },
    {
      "epoch": 0.43258832011535686,
      "grad_norm": 0.010520127601921558,
      "learning_rate": 4.5674116798846436e-05,
      "loss": 0.0014,
      "step": 600
    },
    {
      "epoch": 0.4542177361211247,
      "grad_norm": 0.5669275522232056,
      "learning_rate": 4.5457822638788756e-05,
      "loss": 0.0247,
      "step": 630
    },
    {
      "epoch": 0.47584715212689255,
      "grad_norm": 0.26488080620765686,
      "learning_rate": 4.5241528478731076e-05,
      "loss": 0.0065,
      "step": 660
    },
    {
      "epoch": 0.4974765681326604,
      "grad_norm": 0.6768619418144226,
      "learning_rate": 4.5025234318673395e-05,
      "loss": 0.005,
      "step": 690
    },
    {
      "epoch": 0.5191059841384282,
      "grad_norm": 0.007197729777544737,
      "learning_rate": 4.480894015861572e-05,
      "loss": 0.0013,
      "step": 720
    },
    {
      "epoch": 0.5407354001441961,
      "grad_norm": 0.10392127186059952,
      "learning_rate": 4.459264599855804e-05,
      "loss": 0.0014,
      "step": 750
    },
    {
      "epoch": 0.5623648161499639,
      "grad_norm": 0.01045119110494852,
      "learning_rate": 4.437635183850036e-05,
      "loss": 0.0194,
      "step": 780
    },
    {
      "epoch": 0.5839942321557318,
      "grad_norm": 0.004247739445418119,
      "learning_rate": 4.416005767844269e-05,
      "loss": 0.0001,
      "step": 810
    },
    {
      "epoch": 0.6056236481614996,
      "grad_norm": 0.0034305218141525984,
      "learning_rate": 4.394376351838501e-05,
      "loss": 0.003,
      "step": 840
    },
    {
      "epoch": 0.6272530641672674,
      "grad_norm": 0.0035491471644490957,
      "learning_rate": 4.372746935832733e-05,
      "loss": 0.0004,
      "step": 870
    },
    {
      "epoch": 0.6488824801730353,
      "grad_norm": 0.32892727851867676,
      "learning_rate": 4.3511175198269646e-05,
      "loss": 0.0128,
      "step": 900
    },
    {
      "epoch": 0.6705118961788031,
      "grad_norm": 0.0013508042320609093,
      "learning_rate": 4.329488103821197e-05,
      "loss": 0.0186,
      "step": 930
    },
    {
      "epoch": 0.692141312184571,
      "grad_norm": 0.7996568083763123,
      "learning_rate": 4.307858687815429e-05,
      "loss": 0.0033,
      "step": 960
    },
    {
      "epoch": 0.7137707281903388,
      "grad_norm": 0.49807533621788025,
      "learning_rate": 4.286229271809661e-05,
      "loss": 0.0131,
      "step": 990
    },
    {
      "epoch": 0.7354001441961067,
      "grad_norm": 0.002483088057488203,
      "learning_rate": 4.264599855803894e-05,
      "loss": 0.0006,
      "step": 1020
    },
    {
      "epoch": 0.7570295602018745,
      "grad_norm": 0.0030093444511294365,
      "learning_rate": 4.242970439798126e-05,
      "loss": 0.0013,
      "step": 1050
    },
    {
      "epoch": 0.7786589762076424,
      "grad_norm": 0.0018839107360690832,
      "learning_rate": 4.221341023792358e-05,
      "loss": 0.0123,
      "step": 1080
    },
    {
      "epoch": 0.8002883922134102,
      "grad_norm": 0.604235827922821,
      "learning_rate": 4.19971160778659e-05,
      "loss": 0.0006,
      "step": 1110
    },
    {
      "epoch": 0.821917808219178,
      "grad_norm": 0.009270986542105675,
      "learning_rate": 4.1780821917808224e-05,
      "loss": 0.0051,
      "step": 1140
    },
    {
      "epoch": 0.8435472242249459,
      "grad_norm": 0.168134868144989,
      "learning_rate": 4.1564527757750544e-05,
      "loss": 0.003,
      "step": 1170
    },
    {
      "epoch": 0.8651766402307137,
      "grad_norm": 0.3333891034126282,
      "learning_rate": 4.134823359769286e-05,
      "loss": 0.0247,
      "step": 1200
    },
    {
      "epoch": 0.8868060562364816,
      "grad_norm": 3.897454023361206,
      "learning_rate": 4.113193943763518e-05,
      "loss": 0.0005,
      "step": 1230
    },
    {
      "epoch": 0.9084354722422494,
      "grad_norm": 0.003549683140590787,
      "learning_rate": 4.091564527757751e-05,
      "loss": 0.0047,
      "step": 1260
    },
    {
      "epoch": 0.9300648882480173,
      "grad_norm": 0.01988086849451065,
      "learning_rate": 4.069935111751983e-05,
      "loss": 0.0025,
      "step": 1290
    },
    {
      "epoch": 0.9516943042537851,
      "grad_norm": 0.011269631795585155,
      "learning_rate": 4.048305695746215e-05,
      "loss": 0.0003,
      "step": 1320
    },
    {
      "epoch": 0.9733237202595529,
      "grad_norm": 0.0007853784481994808,
      "learning_rate": 4.0266762797404475e-05,
      "loss": 0.0003,
      "step": 1350
    },
    {
      "epoch": 0.9949531362653208,
      "grad_norm": 0.6295477747917175,
      "learning_rate": 4.0050468637346795e-05,
      "loss": 0.0018,
      "step": 1380
    },
    {
      "epoch": 1.0165825522710887,
      "grad_norm": 0.0005618122522719204,
      "learning_rate": 3.9834174477289115e-05,
      "loss": 0.0002,
      "step": 1410
    },
    {
      "epoch": 1.0382119682768565,
      "grad_norm": 0.00244883238337934,
      "learning_rate": 3.9617880317231434e-05,
      "loss": 0.0002,
      "step": 1440
    },
    {
      "epoch": 1.0598413842826244,
      "grad_norm": 0.0004200914700049907,
      "learning_rate": 3.940158615717376e-05,
      "loss": 0.0058,
      "step": 1470
    },
    {
      "epoch": 1.0814708002883922,
      "grad_norm": 0.0034315199591219425,
      "learning_rate": 3.918529199711608e-05,
      "loss": 0.0001,
      "step": 1500
    },
    {
      "epoch": 1.10310021629416,
      "grad_norm": 0.00489798141643405,
      "learning_rate": 3.89689978370584e-05,
      "loss": 0.0003,
      "step": 1530
    },
    {
      "epoch": 1.1247296322999278,
      "grad_norm": 1.112964391708374,
      "learning_rate": 3.8752703677000726e-05,
      "loss": 0.0001,
      "step": 1560
    },
    {
      "epoch": 1.1463590483056958,
      "grad_norm": 0.00021760116214863956,
      "learning_rate": 3.8536409516943046e-05,
      "loss": 0.0002,
      "step": 1590
    },
    {
      "epoch": 1.1679884643114635,
      "grad_norm": 0.03224096819758415,
      "learning_rate": 3.8320115356885366e-05,
      "loss": 0.0001,
      "step": 1620
    },
    {
      "epoch": 1.1896178803172315,
      "grad_norm": 0.0035550747998058796,
      "learning_rate": 3.8103821196827685e-05,
      "loss": 0.0002,
      "step": 1650
    },
    {
      "epoch": 1.2112472963229992,
      "grad_norm": 0.002545340685173869,
      "learning_rate": 3.788752703677001e-05,
      "loss": 0.0017,
      "step": 1680
    },
    {
      "epoch": 1.2328767123287672,
      "grad_norm": 0.0011141377035528421,
      "learning_rate": 3.767123287671233e-05,
      "loss": 0.0008,
      "step": 1710
    },
    {
      "epoch": 1.254506128334535,
      "grad_norm": 0.002276007318869233,
      "learning_rate": 3.745493871665465e-05,
      "loss": 0.0,
      "step": 1740
    },
    {
      "epoch": 1.2761355443403029,
      "grad_norm": 0.0016464549116790295,
      "learning_rate": 3.723864455659698e-05,
      "loss": 0.0001,
      "step": 1770
    },
    {
      "epoch": 1.2977649603460706,
      "grad_norm": 0.010479186661541462,
      "learning_rate": 3.70223503965393e-05,
      "loss": 0.0004,
      "step": 1800
    },
    {
      "epoch": 1.3193943763518385,
      "grad_norm": 0.0026020577643066645,
      "learning_rate": 3.680605623648162e-05,
      "loss": 0.0158,
      "step": 1830
    },
    {
      "epoch": 1.3410237923576063,
      "grad_norm": 0.007555871736258268,
      "learning_rate": 3.6589762076423937e-05,
      "loss": 0.0003,
      "step": 1860
    },
    {
      "epoch": 1.3626532083633742,
      "grad_norm": 0.05298294499516487,
      "learning_rate": 3.637346791636626e-05,
      "loss": 0.0001,
      "step": 1890
    },
    {
      "epoch": 1.384282624369142,
      "grad_norm": 0.10733333975076675,
      "learning_rate": 3.615717375630858e-05,
      "loss": 0.0001,
      "step": 1920
    },
    {
      "epoch": 1.40591204037491,
      "grad_norm": 0.0003156494931317866,
      "learning_rate": 3.59408795962509e-05,
      "loss": 0.0001,
      "step": 1950
    },
    {
      "epoch": 1.4275414563806776,
      "grad_norm": 0.0002769752754829824,
      "learning_rate": 3.572458543619322e-05,
      "loss": 0.0001,
      "step": 1980
    },
    {
      "epoch": 1.4491708723864456,
      "grad_norm": 0.0004361444734968245,
      "learning_rate": 3.550829127613555e-05,
      "loss": 0.0001,
      "step": 2010
    },
    {
      "epoch": 1.4708002883922133,
      "grad_norm": 0.012317474000155926,
      "learning_rate": 3.529199711607787e-05,
      "loss": 0.0001,
      "step": 2040
    },
    {
      "epoch": 1.4924297043979813,
      "grad_norm": 7.03033510944806e-05,
      "learning_rate": 3.507570295602019e-05,
      "loss": 0.0,
      "step": 2070
    },
    {
      "epoch": 1.5140591204037492,
      "grad_norm": 0.04190248250961304,
      "learning_rate": 3.4859408795962514e-05,
      "loss": 0.0001,
      "step": 2100
    },
    {
      "epoch": 1.535688536409517,
      "grad_norm": 0.017271051183342934,
      "learning_rate": 3.4643114635904834e-05,
      "loss": 0.0001,
      "step": 2130
    },
    {
      "epoch": 1.5573179524152847,
      "grad_norm": 0.023393860086798668,
      "learning_rate": 3.4426820475847154e-05,
      "loss": 0.0267,
      "step": 2160
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 0.0016879999311640859,
      "learning_rate": 3.421052631578947e-05,
      "loss": 0.0001,
      "step": 2190
    },
    {
      "epoch": 1.6005767844268206,
      "grad_norm": 0.00021976766583975405,
      "learning_rate": 3.39942321557318e-05,
      "loss": 0.0001,
      "step": 2220
    },
    {
      "epoch": 1.6222062004325883,
      "grad_norm": 0.04353722557425499,
      "learning_rate": 3.377793799567412e-05,
      "loss": 0.0017,
      "step": 2250
    },
    {
      "epoch": 1.643835616438356,
      "grad_norm": 0.10071003437042236,
      "learning_rate": 3.356164383561644e-05,
      "loss": 0.0002,
      "step": 2280
    },
    {
      "epoch": 1.665465032444124,
      "grad_norm": 0.0025140675716102123,
      "learning_rate": 3.3345349675558765e-05,
      "loss": 0.0001,
      "step": 2310
    },
    {
      "epoch": 1.687094448449892,
      "grad_norm": 0.0024733401369303465,
      "learning_rate": 3.3129055515501085e-05,
      "loss": 0.0001,
      "step": 2340
    },
    {
      "epoch": 1.7087238644556597,
      "grad_norm": 0.008491634391248226,
      "learning_rate": 3.2912761355443405e-05,
      "loss": 0.0002,
      "step": 2370
    },
    {
      "epoch": 1.7303532804614274,
      "grad_norm": 0.0007006598170846701,
      "learning_rate": 3.2696467195385724e-05,
      "loss": 0.0062,
      "step": 2400
    },
    {
      "epoch": 1.7519826964671954,
      "grad_norm": 0.007970300503075123,
      "learning_rate": 3.248017303532805e-05,
      "loss": 0.0003,
      "step": 2430
    },
    {
      "epoch": 1.7736121124729634,
      "grad_norm": 0.006255310960114002,
      "learning_rate": 3.226387887527037e-05,
      "loss": 0.0003,
      "step": 2460
    },
    {
      "epoch": 1.795241528478731,
      "grad_norm": 0.006729640997946262,
      "learning_rate": 3.204758471521269e-05,
      "loss": 0.0001,
      "step": 2490
    },
    {
      "epoch": 1.8168709444844988,
      "grad_norm": 0.002449074760079384,
      "learning_rate": 3.1831290555155017e-05,
      "loss": 0.0004,
      "step": 2520
    },
    {
      "epoch": 1.8385003604902668,
      "grad_norm": 0.005135863088071346,
      "learning_rate": 3.1614996395097336e-05,
      "loss": 0.0002,
      "step": 2550
    },
    {
      "epoch": 1.8601297764960347,
      "grad_norm": 0.006429166533052921,
      "learning_rate": 3.1398702235039656e-05,
      "loss": 0.0001,
      "step": 2580
    },
    {
      "epoch": 1.8817591925018025,
      "grad_norm": 0.1541679948568344,
      "learning_rate": 3.1182408074981976e-05,
      "loss": 0.0026,
      "step": 2610
    },
    {
      "epoch": 1.9033886085075702,
      "grad_norm": 0.005381646100431681,
      "learning_rate": 3.09661139149243e-05,
      "loss": 0.0007,
      "step": 2640
    },
    {
      "epoch": 1.9250180245133381,
      "grad_norm": 0.003323981538414955,
      "learning_rate": 3.074981975486662e-05,
      "loss": 0.0006,
      "step": 2670
    },
    {
      "epoch": 1.946647440519106,
      "grad_norm": 0.0015521268360316753,
      "learning_rate": 3.053352559480894e-05,
      "loss": 0.0086,
      "step": 2700
    },
    {
      "epoch": 1.9682768565248738,
      "grad_norm": 0.005004936829209328,
      "learning_rate": 3.0317231434751264e-05,
      "loss": 0.0001,
      "step": 2730
    },
    {
      "epoch": 1.9899062725306416,
      "grad_norm": 0.0032932290341705084,
      "learning_rate": 3.0100937274693587e-05,
      "loss": 0.0039,
      "step": 2760
    },
    {
      "epoch": 2.0115356885364095,
      "grad_norm": 0.014805459417402744,
      "learning_rate": 2.9884643114635907e-05,
      "loss": 0.0001,
      "step": 2790
    },
    {
      "epoch": 2.0331651045421775,
      "grad_norm": 0.15489494800567627,
      "learning_rate": 2.966834895457823e-05,
      "loss": 0.0006,
      "step": 2820
    },
    {
      "epoch": 2.0547945205479454,
      "grad_norm": 0.0016624957788735628,
      "learning_rate": 2.945205479452055e-05,
      "loss": 0.0051,
      "step": 2850
    },
    {
      "epoch": 2.076423936553713,
      "grad_norm": 0.0007174916681833565,
      "learning_rate": 2.9235760634462873e-05,
      "loss": 0.0,
      "step": 2880
    },
    {
      "epoch": 2.098053352559481,
      "grad_norm": 0.012850388884544373,
      "learning_rate": 2.9019466474405192e-05,
      "loss": 0.0007,
      "step": 2910
    },
    {
      "epoch": 2.119682768565249,
      "grad_norm": 0.0006323842681013048,
      "learning_rate": 2.8803172314347516e-05,
      "loss": 0.0002,
      "step": 2940
    },
    {
      "epoch": 2.1413121845710164,
      "grad_norm": 0.0014068299205973744,
      "learning_rate": 2.8586878154289835e-05,
      "loss": 0.0001,
      "step": 2970
    },
    {
      "epoch": 2.1629416005767843,
      "grad_norm": 0.01084825862199068,
      "learning_rate": 2.8370583994232158e-05,
      "loss": 0.0001,
      "step": 3000
    },
    {
      "epoch": 2.1845710165825523,
      "grad_norm": 0.0006276288186199963,
      "learning_rate": 2.815428983417448e-05,
      "loss": 0.0,
      "step": 3030
    },
    {
      "epoch": 2.20620043258832,
      "grad_norm": 0.002132813213393092,
      "learning_rate": 2.79379956741168e-05,
      "loss": 0.0001,
      "step": 3060
    },
    {
      "epoch": 2.227829848594088,
      "grad_norm": 6.484696388244629,
      "learning_rate": 2.7721701514059124e-05,
      "loss": 0.0003,
      "step": 3090
    },
    {
      "epoch": 2.2494592645998557,
      "grad_norm": 0.0018794754287227988,
      "learning_rate": 2.7505407354001444e-05,
      "loss": 0.0001,
      "step": 3120
    },
    {
      "epoch": 2.2710886806056236,
      "grad_norm": 0.006788654252886772,
      "learning_rate": 2.7289113193943767e-05,
      "loss": 0.0,
      "step": 3150
    },
    {
      "epoch": 2.2927180966113916,
      "grad_norm": 0.002043538261204958,
      "learning_rate": 2.7072819033886086e-05,
      "loss": 0.0,
      "step": 3180
    },
    {
      "epoch": 2.314347512617159,
      "grad_norm": 0.0010403773048892617,
      "learning_rate": 2.685652487382841e-05,
      "loss": 0.0,
      "step": 3210
    },
    {
      "epoch": 2.335976928622927,
      "grad_norm": 0.009223374538123608,
      "learning_rate": 2.664023071377073e-05,
      "loss": 0.0,
      "step": 3240
    },
    {
      "epoch": 2.357606344628695,
      "grad_norm": 0.030018163844943047,
      "learning_rate": 2.6423936553713052e-05,
      "loss": 0.0,
      "step": 3270
    },
    {
      "epoch": 2.379235760634463,
      "grad_norm": 0.11071939766407013,
      "learning_rate": 2.6207642393655375e-05,
      "loss": 0.0,
      "step": 3300
    },
    {
      "epoch": 2.400865176640231,
      "grad_norm": 0.00016689977201167494,
      "learning_rate": 2.5991348233597695e-05,
      "loss": 0.0,
      "step": 3330
    },
    {
      "epoch": 2.4224945926459984,
      "grad_norm": 0.00032225140603259206,
      "learning_rate": 2.5775054073540018e-05,
      "loss": 0.0,
      "step": 3360
    },
    {
      "epoch": 2.4441240086517664,
      "grad_norm": 0.0011028953595086932,
      "learning_rate": 2.5558759913482338e-05,
      "loss": 0.0001,
      "step": 3390
    },
    {
      "epoch": 2.4657534246575343,
      "grad_norm": 0.00032993065542541444,
      "learning_rate": 2.534246575342466e-05,
      "loss": 0.0,
      "step": 3420
    },
    {
      "epoch": 2.487382840663302,
      "grad_norm": 0.00264033698476851,
      "learning_rate": 2.512617159336698e-05,
      "loss": 0.0,
      "step": 3450
    },
    {
      "epoch": 2.50901225666907,
      "grad_norm": 0.0002550692006479949,
      "learning_rate": 2.49098774333093e-05,
      "loss": 0.0001,
      "step": 3480
    },
    {
      "epoch": 2.5306416726748377,
      "grad_norm": 0.007194190286099911,
      "learning_rate": 2.4693583273251623e-05,
      "loss": 0.0,
      "step": 3510
    },
    {
      "epoch": 2.5522710886806057,
      "grad_norm": 0.001481028157286346,
      "learning_rate": 2.4477289113193943e-05,
      "loss": 0.0002,
      "step": 3540
    },
    {
      "epoch": 2.5739005046863737,
      "grad_norm": 0.014366182498633862,
      "learning_rate": 2.4260994953136266e-05,
      "loss": 0.0001,
      "step": 3570
    },
    {
      "epoch": 2.595529920692141,
      "grad_norm": 6.567173113580793e-05,
      "learning_rate": 2.4044700793078585e-05,
      "loss": 0.0001,
      "step": 3600
    },
    {
      "epoch": 2.617159336697909,
      "grad_norm": 9.722941467771307e-05,
      "learning_rate": 2.382840663302091e-05,
      "loss": 0.0,
      "step": 3630
    },
    {
      "epoch": 2.638788752703677,
      "grad_norm": 1.6977719496935606e-05,
      "learning_rate": 2.361211247296323e-05,
      "loss": 0.0,
      "step": 3660
    },
    {
      "epoch": 2.6604181687094446,
      "grad_norm": 0.01529549341648817,
      "learning_rate": 2.339581831290555e-05,
      "loss": 0.0,
      "step": 3690
    },
    {
      "epoch": 2.6820475847152125,
      "grad_norm": 0.0031721957493573427,
      "learning_rate": 2.3179524152847874e-05,
      "loss": 0.0003,
      "step": 3720
    },
    {
      "epoch": 2.7036770007209805,
      "grad_norm": 0.006362732034176588,
      "learning_rate": 2.2963229992790194e-05,
      "loss": 0.0001,
      "step": 3750
    },
    {
      "epoch": 2.7253064167267484,
      "grad_norm": 0.002783670322969556,
      "learning_rate": 2.2746935832732517e-05,
      "loss": 0.0,
      "step": 3780
    },
    {
      "epoch": 2.7469358327325164,
      "grad_norm": 0.00039837489021010697,
      "learning_rate": 2.2530641672674837e-05,
      "loss": 0.0,
      "step": 3810
    },
    {
      "epoch": 2.768565248738284,
      "grad_norm": 0.0010957849444821477,
      "learning_rate": 2.231434751261716e-05,
      "loss": 0.0,
      "step": 3840
    },
    {
      "epoch": 2.790194664744052,
      "grad_norm": 0.02844543755054474,
      "learning_rate": 2.209805335255948e-05,
      "loss": 0.0001,
      "step": 3870
    },
    {
      "epoch": 2.81182408074982,
      "grad_norm": 6.718010263284668e-05,
      "learning_rate": 2.1881759192501802e-05,
      "loss": 0.0,
      "step": 3900
    },
    {
      "epoch": 2.8334534967555873,
      "grad_norm": 0.006276572123169899,
      "learning_rate": 2.1665465032444125e-05,
      "loss": 0.0004,
      "step": 3930
    },
    {
      "epoch": 2.8550829127613553,
      "grad_norm": 0.0002856945211533457,
      "learning_rate": 2.1449170872386445e-05,
      "loss": 0.0002,
      "step": 3960
    },
    {
      "epoch": 2.8767123287671232,
      "grad_norm": 0.0002423626574454829,
      "learning_rate": 2.1232876712328768e-05,
      "loss": 0.0001,
      "step": 3990
    },
    {
      "epoch": 2.898341744772891,
      "grad_norm": 0.007484122645109892,
      "learning_rate": 2.1016582552271088e-05,
      "loss": 0.0,
      "step": 4020
    },
    {
      "epoch": 2.919971160778659,
      "grad_norm": 0.0007640399853698909,
      "learning_rate": 2.080028839221341e-05,
      "loss": 0.0,
      "step": 4050
    },
    {
      "epoch": 2.9416005767844267,
      "grad_norm": 0.00018265791004523635,
      "learning_rate": 2.058399423215573e-05,
      "loss": 0.0,
      "step": 4080
    },
    {
      "epoch": 2.9632299927901946,
      "grad_norm": 0.0007543243118561804,
      "learning_rate": 2.0367700072098053e-05,
      "loss": 0.0,
      "step": 4110
    },
    {
      "epoch": 2.9848594087959626,
      "grad_norm": 0.00029646529583260417,
      "learning_rate": 2.0151405912040373e-05,
      "loss": 0.0,
      "step": 4140
    },
    {
      "epoch": 3.0064888248017305,
      "grad_norm": 0.00011471308243926615,
      "learning_rate": 1.9935111751982696e-05,
      "loss": 0.0,
      "step": 4170
    },
    {
      "epoch": 3.028118240807498,
      "grad_norm": 0.008354193530976772,
      "learning_rate": 1.971881759192502e-05,
      "loss": 0.0,
      "step": 4200
    },
    {
      "epoch": 3.049747656813266,
      "grad_norm": 0.9368739724159241,
      "learning_rate": 1.950252343186734e-05,
      "loss": 0.0001,
      "step": 4230
    },
    {
      "epoch": 3.071377072819034,
      "grad_norm": 0.006005523260682821,
      "learning_rate": 1.9286229271809662e-05,
      "loss": 0.0,
      "step": 4260
    },
    {
      "epoch": 3.093006488824802,
      "grad_norm": 0.006732122041285038,
      "learning_rate": 1.906993511175198e-05,
      "loss": 0.0001,
      "step": 4290
    },
    {
      "epoch": 3.1146359048305694,
      "grad_norm": 0.00011062090197810903,
      "learning_rate": 1.8853640951694305e-05,
      "loss": 0.0001,
      "step": 4320
    },
    {
      "epoch": 3.1362653208363374,
      "grad_norm": 0.0026187694165855646,
      "learning_rate": 1.8637346791636624e-05,
      "loss": 0.0,
      "step": 4350
    },
    {
      "epoch": 3.1578947368421053,
      "grad_norm": 0.0003347834281157702,
      "learning_rate": 1.8421052631578947e-05,
      "loss": 0.0,
      "step": 4380
    },
    {
      "epoch": 3.1795241528478733,
      "grad_norm": 0.0027828714810311794,
      "learning_rate": 1.820475847152127e-05,
      "loss": 0.0,
      "step": 4410
    },
    {
      "epoch": 3.2011535688536408,
      "grad_norm": 0.01742876134812832,
      "learning_rate": 1.798846431146359e-05,
      "loss": 0.0001,
      "step": 4440
    },
    {
      "epoch": 3.2227829848594087,
      "grad_norm": 0.001497482880949974,
      "learning_rate": 1.7772170151405913e-05,
      "loss": 0.0,
      "step": 4470
    },
    {
      "epoch": 3.2444124008651767,
      "grad_norm": 0.0012416379759088159,
      "learning_rate": 1.7555875991348233e-05,
      "loss": 0.0,
      "step": 4500
    },
    {
      "epoch": 3.2660418168709446,
      "grad_norm": 0.004536696709692478,
      "learning_rate": 1.7339581831290556e-05,
      "loss": 0.0001,
      "step": 4530
    },
    {
      "epoch": 3.287671232876712,
      "grad_norm": 0.000561400200240314,
      "learning_rate": 1.7123287671232875e-05,
      "loss": 0.0,
      "step": 4560
    },
    {
      "epoch": 3.30930064888248,
      "grad_norm": 0.002419061027467251,
      "learning_rate": 1.69069935111752e-05,
      "loss": 0.0,
      "step": 4590
    },
    {
      "epoch": 3.330930064888248,
      "grad_norm": 0.012188425287604332,
      "learning_rate": 1.6690699351117518e-05,
      "loss": 0.0,
      "step": 4620
    },
    {
      "epoch": 3.352559480894016,
      "grad_norm": 0.0004496476030908525,
      "learning_rate": 1.647440519105984e-05,
      "loss": 0.0004,
      "step": 4650
    },
    {
      "epoch": 3.3741888968997835,
      "grad_norm": 0.00017531379126012325,
      "learning_rate": 1.6258111031002164e-05,
      "loss": 0.0,
      "step": 4680
    },
    {
      "epoch": 3.3958183129055515,
      "grad_norm": 0.00026417901972308755,
      "learning_rate": 1.6041816870944484e-05,
      "loss": 0.0,
      "step": 4710
    },
    {
      "epoch": 3.4174477289113194,
      "grad_norm": 0.004818809684365988,
      "learning_rate": 1.5825522710886807e-05,
      "loss": 0.0,
      "step": 4740
    },
    {
      "epoch": 3.4390771449170874,
      "grad_norm": 0.001972823403775692,
      "learning_rate": 1.5609228550829127e-05,
      "loss": 0.0,
      "step": 4770
    },
    {
      "epoch": 3.460706560922855,
      "grad_norm": 0.0018862077267840505,
      "learning_rate": 1.539293439077145e-05,
      "loss": 0.0,
      "step": 4800
    },
    {
      "epoch": 3.482335976928623,
      "grad_norm": 0.013610747642815113,
      "learning_rate": 1.5176640230713771e-05,
      "loss": 0.0001,
      "step": 4830
    },
    {
      "epoch": 3.503965392934391,
      "grad_norm": 0.00016503581719007343,
      "learning_rate": 1.4960346070656092e-05,
      "loss": 0.0,
      "step": 4860
    },
    {
      "epoch": 3.5255948089401588,
      "grad_norm": 0.0001457014586776495,
      "learning_rate": 1.4744051910598414e-05,
      "loss": 0.0,
      "step": 4890
    },
    {
      "epoch": 3.5472242249459267,
      "grad_norm": 0.006511267740279436,
      "learning_rate": 1.4527757750540735e-05,
      "loss": 0.0,
      "step": 4920
    },
    {
      "epoch": 3.568853640951694,
      "grad_norm": 0.0005859262309968472,
      "learning_rate": 1.4311463590483056e-05,
      "loss": 0.0002,
      "step": 4950
    },
    {
      "epoch": 3.590483056957462,
      "grad_norm": 0.002227773889899254,
      "learning_rate": 1.4095169430425378e-05,
      "loss": 0.0,
      "step": 4980
    },
    {
      "epoch": 3.61211247296323,
      "grad_norm": 0.0008713772404007614,
      "learning_rate": 1.38788752703677e-05,
      "loss": 0.0,
      "step": 5010
    },
    {
      "epoch": 3.6337418889689976,
      "grad_norm": 0.0006510260864160955,
      "learning_rate": 1.3662581110310022e-05,
      "loss": 0.0,
      "step": 5040
    },
    {
      "epoch": 3.6553713049747656,
      "grad_norm": 0.004389900714159012,
      "learning_rate": 1.3446286950252344e-05,
      "loss": 0.0,
      "step": 5070
    },
    {
      "epoch": 3.6770007209805335,
      "grad_norm": 0.0011542979627847672,
      "learning_rate": 1.3229992790194665e-05,
      "loss": 0.0,
      "step": 5100
    },
    {
      "epoch": 3.6986301369863015,
      "grad_norm": 0.014204882085323334,
      "learning_rate": 1.3013698630136986e-05,
      "loss": 0.0,
      "step": 5130
    },
    {
      "epoch": 3.7202595529920695,
      "grad_norm": 0.0007670574705116451,
      "learning_rate": 1.2797404470079308e-05,
      "loss": 0.0,
      "step": 5160
    },
    {
      "epoch": 3.741888968997837,
      "grad_norm": 0.0014096363447606564,
      "learning_rate": 1.2581110310021629e-05,
      "loss": 0.0001,
      "step": 5190
    },
    {
      "epoch": 3.763518385003605,
      "grad_norm": 6.187328835949302e-05,
      "learning_rate": 1.236481614996395e-05,
      "loss": 0.0,
      "step": 5220
    },
    {
      "epoch": 3.785147801009373,
      "grad_norm": 0.0003874276007991284,
      "learning_rate": 1.2148521989906272e-05,
      "loss": 0.0002,
      "step": 5250
    },
    {
      "epoch": 3.8067772170151404,
      "grad_norm": 0.00010346725321142003,
      "learning_rate": 1.1932227829848595e-05,
      "loss": 0.0,
      "step": 5280
    },
    {
      "epoch": 3.8284066330209083,
      "grad_norm": 0.020370708778500557,
      "learning_rate": 1.1715933669790916e-05,
      "loss": 0.0,
      "step": 5310
    },
    {
      "epoch": 3.8500360490266763,
      "grad_norm": 0.000221878188312985,
      "learning_rate": 1.1499639509733238e-05,
      "loss": 0.0,
      "step": 5340
    },
    {
      "epoch": 3.8716654650324442,
      "grad_norm": 0.004732194356620312,
      "learning_rate": 1.1283345349675559e-05,
      "loss": 0.0,
      "step": 5370
    },
    {
      "epoch": 3.893294881038212,
      "grad_norm": 0.011568304151296616,
      "learning_rate": 1.106705118961788e-05,
      "loss": 0.0023,
      "step": 5400
    },
    {
      "epoch": 3.9149242970439797,
      "grad_norm": 0.0011147951008751988,
      "learning_rate": 1.0850757029560202e-05,
      "loss": 0.0,
      "step": 5430
    },
    {
      "epoch": 3.9365537130497477,
      "grad_norm": 0.00035450098221190274,
      "learning_rate": 1.0634462869502523e-05,
      "loss": 0.0,
      "step": 5460
    },
    {
      "epoch": 3.9581831290555156,
      "grad_norm": 0.0011028106091544032,
      "learning_rate": 1.0418168709444844e-05,
      "loss": 0.0012,
      "step": 5490
    },
    {
      "epoch": 3.979812545061283,
      "grad_norm": 0.00045547273475676775,
      "learning_rate": 1.0201874549387166e-05,
      "loss": 0.0,
      "step": 5520
    },
    {
      "epoch": 4.0014419610670515,
      "grad_norm": 7.938985800137743e-05,
      "learning_rate": 9.985580389329489e-06,
      "loss": 0.0,
      "step": 5550
    },
    {
      "epoch": 4.023071377072819,
      "grad_norm": 0.003888361854478717,
      "learning_rate": 9.76928622927181e-06,
      "loss": 0.0,
      "step": 5580
    },
    {
      "epoch": 4.0447007930785865,
      "grad_norm": 0.00029194954549893737,
      "learning_rate": 9.552992069214131e-06,
      "loss": 0.0,
      "step": 5610
    },
    {
      "epoch": 4.066330209084355,
      "grad_norm": 0.000853860517963767,
      "learning_rate": 9.336697909156453e-06,
      "loss": 0.0,
      "step": 5640
    },
    {
      "epoch": 4.0879596250901225,
      "grad_norm": 0.0010367140639573336,
      "learning_rate": 9.120403749098774e-06,
      "loss": 0.0,
      "step": 5670
    },
    {
      "epoch": 4.109589041095891,
      "grad_norm": 0.00010384136840002611,
      "learning_rate": 8.904109589041095e-06,
      "loss": 0.0,
      "step": 5700
    },
    {
      "epoch": 4.131218457101658,
      "grad_norm": 0.0006316221551969647,
      "learning_rate": 8.687815428983417e-06,
      "loss": 0.0,
      "step": 5730
    },
    {
      "epoch": 4.152847873107426,
      "grad_norm": 0.0005390524747781456,
      "learning_rate": 8.471521268925738e-06,
      "loss": 0.0,
      "step": 5760
    },
    {
      "epoch": 4.174477289113194,
      "grad_norm": 5.524172306060791,
      "learning_rate": 8.255227108868061e-06,
      "loss": 0.0022,
      "step": 5790
    },
    {
      "epoch": 4.196106705118962,
      "grad_norm": 0.0010216891532763839,
      "learning_rate": 8.038932948810383e-06,
      "loss": 0.0,
      "step": 5820
    },
    {
      "epoch": 4.217736121124729,
      "grad_norm": 0.0003730811004061252,
      "learning_rate": 7.822638788752704e-06,
      "loss": 0.0,
      "step": 5850
    },
    {
      "epoch": 4.239365537130498,
      "grad_norm": 0.001277286559343338,
      "learning_rate": 7.606344628695025e-06,
      "loss": 0.0,
      "step": 5880
    },
    {
      "epoch": 4.260994953136265,
      "grad_norm": 0.0025212133768945932,
      "learning_rate": 7.390050468637347e-06,
      "loss": 0.0,
      "step": 5910
    },
    {
      "epoch": 4.282624369142033,
      "grad_norm": 0.004302350804209709,
      "learning_rate": 7.173756308579668e-06,
      "loss": 0.0,
      "step": 5940
    },
    {
      "epoch": 4.304253785147801,
      "grad_norm": 0.0012293955078348517,
      "learning_rate": 6.95746214852199e-06,
      "loss": 0.0,
      "step": 5970
    },
    {
      "epoch": 4.325883201153569,
      "grad_norm": 0.0006065890775062144,
      "learning_rate": 6.7411679884643116e-06,
      "loss": 0.0,
      "step": 6000
    },
    {
      "epoch": 4.347512617159337,
      "grad_norm": 0.0015407137107104063,
      "learning_rate": 6.524873828406633e-06,
      "loss": 0.0,
      "step": 6030
    },
    {
      "epoch": 4.3691420331651045,
      "grad_norm": 0.0002221543254563585,
      "learning_rate": 6.308579668348954e-06,
      "loss": 0.0,
      "step": 6060
    },
    {
      "epoch": 4.390771449170872,
      "grad_norm": 0.007599177770316601,
      "learning_rate": 6.0922855082912765e-06,
      "loss": 0.0,
      "step": 6090
    },
    {
      "epoch": 4.41240086517664,
      "grad_norm": 0.003168993629515171,
      "learning_rate": 5.875991348233598e-06,
      "loss": 0.0,
      "step": 6120
    },
    {
      "epoch": 4.434030281182408,
      "grad_norm": 0.0012669307179749012,
      "learning_rate": 5.659697188175919e-06,
      "loss": 0.0,
      "step": 6150
    },
    {
      "epoch": 4.455659697188176,
      "grad_norm": 0.0024521639570593834,
      "learning_rate": 5.4434030281182405e-06,
      "loss": 0.0,
      "step": 6180
    },
    {
      "epoch": 4.477289113193944,
      "grad_norm": 0.03333938494324684,
      "learning_rate": 5.227108868060563e-06,
      "loss": 0.005,
      "step": 6210
    },
    {
      "epoch": 4.498918529199711,
      "grad_norm": 0.008298776112496853,
      "learning_rate": 5.010814708002884e-06,
      "loss": 0.0,
      "step": 6240
    },
    {
      "epoch": 4.52054794520548,
      "grad_norm": 0.0008691897382959723,
      "learning_rate": 4.7945205479452054e-06,
      "loss": 0.0,
      "step": 6270
    },
    {
      "epoch": 4.542177361211247,
      "grad_norm": 0.011668123304843903,
      "learning_rate": 4.578226387887527e-06,
      "loss": 0.0,
      "step": 6300
    },
    {
      "epoch": 4.563806777217015,
      "grad_norm": 0.0002202845789724961,
      "learning_rate": 4.361932227829848e-06,
      "loss": 0.0033,
      "step": 6330
    },
    {
      "epoch": 4.585436193222783,
      "grad_norm": 0.00031415498233400285,
      "learning_rate": 4.14563806777217e-06,
      "loss": 0.0,
      "step": 6360
    },
    {
      "epoch": 4.607065609228551,
      "grad_norm": 0.00010377545549999923,
      "learning_rate": 3.929343907714492e-06,
      "loss": 0.0,
      "step": 6390
    },
    {
      "epoch": 4.628695025234318,
      "grad_norm": 0.0029919766820967197,
      "learning_rate": 3.713049747656813e-06,
      "loss": 0.0,
      "step": 6420
    },
    {
      "epoch": 4.650324441240087,
      "grad_norm": 0.000775750435423106,
      "learning_rate": 3.496755587599135e-06,
      "loss": 0.0,
      "step": 6450
    },
    {
      "epoch": 4.671953857245854,
      "grad_norm": 0.0015133414417505264,
      "learning_rate": 3.280461427541456e-06,
      "loss": 0.0,
      "step": 6480
    },
    {
      "epoch": 4.6935832732516225,
      "grad_norm": 0.0023428713902831078,
      "learning_rate": 3.064167267483778e-06,
      "loss": 0.0,
      "step": 6510
    },
    {
      "epoch": 4.71521268925739,
      "grad_norm": 0.002895868383347988,
      "learning_rate": 2.8478731074260993e-06,
      "loss": 0.0,
      "step": 6540
    },
    {
      "epoch": 4.7368421052631575,
      "grad_norm": 0.00880496297031641,
      "learning_rate": 2.631578947368421e-06,
      "loss": 0.0,
      "step": 6570
    },
    {
      "epoch": 4.758471521268926,
      "grad_norm": 0.0009111640974879265,
      "learning_rate": 2.4152847873107425e-06,
      "loss": 0.0001,
      "step": 6600
    },
    {
      "epoch": 4.780100937274693,
      "grad_norm": 0.0007791388779878616,
      "learning_rate": 2.1989906272530642e-06,
      "loss": 0.0,
      "step": 6630
    },
    {
      "epoch": 4.801730353280462,
      "grad_norm": 0.00030644936487078667,
      "learning_rate": 1.9826964671953856e-06,
      "loss": 0.0001,
      "step": 6660
    },
    {
      "epoch": 4.823359769286229,
      "grad_norm": 0.00020950015459675342,
      "learning_rate": 1.7664023071377072e-06,
      "loss": 0.0004,
      "step": 6690
    },
    {
      "epoch": 4.844989185291997,
      "grad_norm": 0.0002440083189867437,
      "learning_rate": 1.5501081470800287e-06,
      "loss": 0.0,
      "step": 6720
    },
    {
      "epoch": 4.866618601297765,
      "grad_norm": 0.0005655589047819376,
      "learning_rate": 1.3338139870223503e-06,
      "loss": 0.0,
      "step": 6750
    },
    {
      "epoch": 4.888248017303533,
      "grad_norm": 0.0002683950588107109,
      "learning_rate": 1.1175198269646719e-06,
      "loss": 0.0,
      "step": 6780
    },
    {
      "epoch": 4.9098774333093,
      "grad_norm": 0.0026675432454794645,
      "learning_rate": 9.012256669069935e-07,
      "loss": 0.0,
      "step": 6810
    },
    {
      "epoch": 4.931506849315069,
      "grad_norm": 0.0039031701162457466,
      "learning_rate": 6.849315068493151e-07,
      "loss": 0.0,
      "step": 6840
    },
    {
      "epoch": 4.953136265320836,
      "grad_norm": 0.056871198117733,
      "learning_rate": 4.686373467916366e-07,
      "loss": 0.0,
      "step": 6870
    },
    {
      "epoch": 4.974765681326604,
      "grad_norm": 0.00018152125994674861,
      "learning_rate": 2.523431867339582e-07,
      "loss": 0.0008,
      "step": 6900
    },
    {
      "epoch": 4.996395097332372,
      "grad_norm": 0.0004676603712141514,
      "learning_rate": 3.604902667627974e-08,
      "loss": 0.0006,
      "step": 6930
    }
  ],
  "logging_steps": 30,
  "max_steps": 6935,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}