{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.992407324698526,
  "eval_steps": 500,
  "global_step": 837,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01786511835640911,
      "grad_norm": 0.5766569375991821,
      "learning_rate": 4.999559763441755e-05,
      "loss": 1.5453,
      "step": 5
    },
    {
      "epoch": 0.03573023671281822,
      "grad_norm": 0.4148772656917572,
      "learning_rate": 4.998239208813602e-05,
      "loss": 1.4724,
      "step": 10
    },
    {
      "epoch": 0.053595355069227336,
      "grad_norm": 0.3131537139415741,
      "learning_rate": 4.9960388012006784e-05,
      "loss": 1.2985,
      "step": 15
    },
    {
      "epoch": 0.07146047342563644,
      "grad_norm": 0.26158425211906433,
      "learning_rate": 4.992959315562887e-05,
      "loss": 1.3255,
      "step": 20
    },
    {
      "epoch": 0.08932559178204555,
      "grad_norm": 0.23404298722743988,
      "learning_rate": 4.9890018364619516e-05,
      "loss": 1.245,
      "step": 25
    },
    {
      "epoch": 0.10719071013845467,
      "grad_norm": 0.19169265031814575,
      "learning_rate": 4.984167757679458e-05,
      "loss": 1.2509,
      "step": 30
    },
    {
      "epoch": 0.12505582849486377,
      "grad_norm": 0.29275861382484436,
      "learning_rate": 4.9784587817259674e-05,
      "loss": 1.2937,
      "step": 35
    },
    {
      "epoch": 0.14292094685127288,
      "grad_norm": 0.1952601969242096,
      "learning_rate": 4.971876919241422e-05,
      "loss": 1.1818,
      "step": 40
    },
    {
      "epoch": 0.160786065207682,
      "grad_norm": 0.223506361246109,
      "learning_rate": 4.964424488287009e-05,
      "loss": 1.2293,
      "step": 45
    },
    {
      "epoch": 0.1786511835640911,
      "grad_norm": 0.23478031158447266,
      "learning_rate": 4.956104113528776e-05,
      "loss": 1.2531,
      "step": 50
    },
    {
      "epoch": 0.19651630192050024,
      "grad_norm": 0.26763561367988586,
      "learning_rate": 4.9469187253132356e-05,
      "loss": 1.2131,
      "step": 55
    },
    {
      "epoch": 0.21438142027690935,
      "grad_norm": 0.2691376507282257,
      "learning_rate": 4.936871558635346e-05,
      "loss": 1.2497,
      "step": 60
    },
    {
      "epoch": 0.23224653863331846,
      "grad_norm": 0.2545244097709656,
      "learning_rate": 4.925966151999168e-05,
      "loss": 1.1778,
      "step": 65
    },
    {
      "epoch": 0.25011165698972754,
      "grad_norm": 0.30710119009017944,
      "learning_rate": 4.914206346171651e-05,
      "loss": 1.2123,
      "step": 70
    },
    {
      "epoch": 0.2679767753461367,
      "grad_norm": 0.23636460304260254,
      "learning_rate": 4.901596282829948e-05,
      "loss": 1.2007,
      "step": 75
    },
    {
      "epoch": 0.28584189370254576,
      "grad_norm": 0.2662898600101471,
      "learning_rate": 4.888140403102768e-05,
      "loss": 1.1745,
      "step": 80
    },
    {
      "epoch": 0.3037070120589549,
      "grad_norm": 0.308378130197525,
      "learning_rate": 4.8738434460062524e-05,
      "loss": 1.1981,
      "step": 85
    },
    {
      "epoch": 0.321572130415364,
      "grad_norm": 0.23143641650676727,
      "learning_rate": 4.858710446774951e-05,
      "loss": 1.1297,
      "step": 90
    },
    {
      "epoch": 0.3394372487717731,
      "grad_norm": 0.2781310975551605,
      "learning_rate": 4.842746735088461e-05,
      "loss": 1.2359,
      "step": 95
    },
    {
      "epoch": 0.3573023671281822,
      "grad_norm": 0.29989567399024963,
      "learning_rate": 4.825957933194376e-05,
      "loss": 1.1798,
      "step": 100
    },
    {
      "epoch": 0.37516748548459133,
      "grad_norm": 0.3551138937473297,
      "learning_rate": 4.808349953928184e-05,
      "loss": 1.2055,
      "step": 105
    },
    {
      "epoch": 0.39303260384100047,
      "grad_norm": 0.3111821413040161,
      "learning_rate": 4.789928998630838e-05,
      "loss": 1.2145,
      "step": 110
    },
    {
      "epoch": 0.41089772219740955,
      "grad_norm": 0.27980926632881165,
      "learning_rate": 4.770701554964706e-05,
      "loss": 1.1673,
      "step": 115
    },
    {
      "epoch": 0.4287628405538187,
      "grad_norm": 0.2745010256767273,
      "learning_rate": 4.750674394628687e-05,
      "loss": 1.1494,
      "step": 120
    },
    {
      "epoch": 0.4466279589102278,
      "grad_norm": 0.2852526307106018,
      "learning_rate": 4.729854570973289e-05,
      "loss": 1.1219,
      "step": 125
    },
    {
      "epoch": 0.4644930772666369,
      "grad_norm": 0.35505807399749756,
      "learning_rate": 4.7082494165165216e-05,
      "loss": 1.1953,
      "step": 130
    },
    {
      "epoch": 0.482358195623046,
      "grad_norm": 0.32421571016311646,
      "learning_rate": 4.685866540361456e-05,
      "loss": 1.0871,
      "step": 135
    },
    {
      "epoch": 0.5002233139794551,
      "grad_norm": 0.27323105931282043,
      "learning_rate": 4.662713825516379e-05,
      "loss": 1.1604,
      "step": 140
    },
    {
      "epoch": 0.5180884323358642,
      "grad_norm": 0.3120843470096588,
      "learning_rate": 4.638799426118492e-05,
      "loss": 1.1202,
      "step": 145
    },
    {
      "epoch": 0.5359535506922734,
      "grad_norm": 0.26333290338516235,
      "learning_rate": 4.6141317645621e-05,
      "loss": 1.1539,
      "step": 150
    },
    {
      "epoch": 0.5538186690486825,
      "grad_norm": 0.270474374294281,
      "learning_rate": 4.588719528532342e-05,
      "loss": 1.1343,
      "step": 155
    },
    {
      "epoch": 0.5716837874050915,
      "grad_norm": 0.2665104866027832,
      "learning_rate": 4.5625716679454787e-05,
      "loss": 1.2117,
      "step": 160
    },
    {
      "epoch": 0.5895489057615007,
      "grad_norm": 0.29763564467430115,
      "learning_rate": 4.535697391796832e-05,
      "loss": 1.1323,
      "step": 165
    },
    {
      "epoch": 0.6074140241179098,
      "grad_norm": 0.36548730731010437,
      "learning_rate": 4.50810616491747e-05,
      "loss": 1.1848,
      "step": 170
    },
    {
      "epoch": 0.6252791424743189,
      "grad_norm": 0.32450002431869507,
      "learning_rate": 4.479807704640802e-05,
      "loss": 1.1829,
      "step": 175
    },
    {
      "epoch": 0.643144260830728,
      "grad_norm": 0.4330536425113678,
      "learning_rate": 4.45081197738023e-05,
      "loss": 1.1881,
      "step": 180
    },
    {
      "epoch": 0.6610093791871371,
      "grad_norm": 0.3612164258956909,
      "learning_rate": 4.421129195119094e-05,
      "loss": 1.23,
      "step": 185
    },
    {
      "epoch": 0.6788744975435462,
      "grad_norm": 0.3616299033164978,
      "learning_rate": 4.390769811814116e-05,
      "loss": 1.1593,
      "step": 190
    },
    {
      "epoch": 0.6967396158999554,
      "grad_norm": 0.3271953761577606,
      "learning_rate": 4.359744519713628e-05,
      "loss": 1.1454,
      "step": 195
    },
    {
      "epoch": 0.7146047342563644,
      "grad_norm": 0.3824130594730377,
      "learning_rate": 4.3280642455918806e-05,
      "loss": 1.1633,
      "step": 200
    },
    {
      "epoch": 0.7324698526127735,
      "grad_norm": 0.36359068751335144,
      "learning_rate": 4.2957401469007495e-05,
      "loss": 1.1841,
      "step": 205
    },
    {
      "epoch": 0.7503349709691827,
      "grad_norm": 0.41856512427330017,
      "learning_rate": 4.262783607840199e-05,
      "loss": 1.1567,
      "step": 210
    },
    {
      "epoch": 0.7682000893255918,
      "grad_norm": 0.2762792408466339,
      "learning_rate": 4.229206235348891e-05,
      "loss": 1.1147,
      "step": 215
    },
    {
      "epoch": 0.7860652076820009,
      "grad_norm": 0.37006062269210815,
      "learning_rate": 4.195019855016346e-05,
      "loss": 1.1194,
      "step": 220
    },
    {
      "epoch": 0.80393032603841,
      "grad_norm": 0.33580338954925537,
      "learning_rate": 4.160236506918098e-05,
      "loss": 1.1884,
      "step": 225
    },
    {
      "epoch": 0.8217954443948191,
      "grad_norm": 0.33786651492118835,
      "learning_rate": 4.124868441375307e-05,
      "loss": 1.1687,
      "step": 230
    },
    {
      "epoch": 0.8396605627512282,
      "grad_norm": 0.3432258665561676,
      "learning_rate": 4.08892811464033e-05,
      "loss": 1.1512,
      "step": 235
    },
    {
      "epoch": 0.8575256811076374,
      "grad_norm": 0.3321971595287323,
      "learning_rate": 4.052428184509762e-05,
      "loss": 1.1417,
      "step": 240
    },
    {
      "epoch": 0.8753907994640464,
      "grad_norm": 0.371055543422699,
      "learning_rate": 4.0153815058664976e-05,
      "loss": 1.1595,
      "step": 245
    },
    {
      "epoch": 0.8932559178204555,
      "grad_norm": 0.3127325773239136,
      "learning_rate": 3.977801126152376e-05,
      "loss": 1.1661,
      "step": 250
    },
    {
      "epoch": 0.9111210361768647,
      "grad_norm": 0.28069669008255005,
      "learning_rate": 3.9397002807730166e-05,
      "loss": 1.1271,
      "step": 255
    },
    {
      "epoch": 0.9289861545332738,
      "grad_norm": 0.39312201738357544,
      "learning_rate": 3.9010923884364467e-05,
      "loss": 1.0852,
      "step": 260
    },
    {
      "epoch": 0.9468512728896828,
      "grad_norm": 0.44475504755973816,
      "learning_rate": 3.861991046427182e-05,
      "loss": 1.0884,
      "step": 265
    },
    {
      "epoch": 0.964716391246092,
      "grad_norm": 0.3146402835845947,
      "learning_rate": 3.822410025817406e-05,
      "loss": 1.1159,
      "step": 270
    },
    {
      "epoch": 0.9825815096025011,
      "grad_norm": 0.4495191276073456,
      "learning_rate": 3.782363266616946e-05,
      "loss": 1.1642,
      "step": 275
    },
    {
      "epoch": 1.0013398838767307,
      "grad_norm": 0.30039986968040466,
      "learning_rate": 3.741864872863754e-05,
      "loss": 1.1599,
      "step": 280
    },
    {
      "epoch": 1.0192050022331398,
      "grad_norm": 0.3826463222503662,
      "learning_rate": 3.700929107656614e-05,
      "loss": 1.1167,
      "step": 285
    },
    {
      "epoch": 1.037070120589549,
      "grad_norm": 0.38993003964424133,
      "learning_rate": 3.659570388131832e-05,
      "loss": 1.087,
      "step": 290
    },
    {
      "epoch": 1.054935238945958,
      "grad_norm": 0.4054009020328522,
      "learning_rate": 3.61780328038568e-05,
      "loss": 1.0912,
      "step": 295
    },
    {
      "epoch": 1.0728003573023672,
      "grad_norm": 0.47545984387397766,
      "learning_rate": 3.575642494344365e-05,
      "loss": 1.0991,
      "step": 300
    },
    {
      "epoch": 1.0906654756587761,
      "grad_norm": 0.4191313683986664,
      "learning_rate": 3.533102878583361e-05,
      "loss": 1.1219,
      "step": 305
    },
    {
      "epoch": 1.1085305940151853,
      "grad_norm": 0.3124043643474579,
      "learning_rate": 3.490199415097892e-05,
      "loss": 1.1198,
      "step": 310
    },
    {
      "epoch": 1.1263957123715944,
      "grad_norm": 0.36894044280052185,
      "learning_rate": 3.44694721402644e-05,
      "loss": 1.1193,
      "step": 315
    },
    {
      "epoch": 1.1442608307280036,
      "grad_norm": 0.5683015584945679,
      "learning_rate": 3.4033615083291135e-05,
      "loss": 1.142,
      "step": 320
    },
    {
      "epoch": 1.1621259490844127,
      "grad_norm": 0.371493399143219,
      "learning_rate": 3.3594576484227655e-05,
      "loss": 1.128,
      "step": 325
    },
    {
      "epoch": 1.1799910674408218,
      "grad_norm": 0.38947540521621704,
      "learning_rate": 3.315251096774737e-05,
      "loss": 1.1046,
      "step": 330
    },
    {
      "epoch": 1.197856185797231,
      "grad_norm": 0.3288847804069519,
      "learning_rate": 3.2707574224571495e-05,
      "loss": 1.1209,
      "step": 335
    },
    {
      "epoch": 1.21572130415364,
      "grad_norm": 0.36680731177330017,
      "learning_rate": 3.225992295663639e-05,
      "loss": 1.0932,
      "step": 340
    },
    {
      "epoch": 1.233586422510049,
      "grad_norm": 0.4268665909767151,
      "learning_rate": 3.1809714821904834e-05,
      "loss": 1.1161,
      "step": 345
    },
    {
      "epoch": 1.2514515408664582,
      "grad_norm": 0.37560388445854187,
      "learning_rate": 3.1357108378840616e-05,
      "loss": 1.1264,
      "step": 350
    },
    {
      "epoch": 1.2693166592228673,
      "grad_norm": 0.35783258080482483,
      "learning_rate": 3.0902263030565925e-05,
      "loss": 1.1226,
      "step": 355
    },
    {
      "epoch": 1.2871817775792764,
      "grad_norm": 0.3974683880805969,
      "learning_rate": 3.0445338968721287e-05,
      "loss": 1.0651,
      "step": 360
    },
    {
      "epoch": 1.3050468959356856,
      "grad_norm": 0.36973434686660767,
      "learning_rate": 2.9986497117047797e-05,
      "loss": 1.1052,
      "step": 365
    },
    {
      "epoch": 1.3229120142920947,
      "grad_norm": 0.40590983629226685,
      "learning_rate": 2.9525899074711506e-05,
      "loss": 1.0979,
      "step": 370
    },
    {
      "epoch": 1.3407771326485038,
      "grad_norm": 0.3565100431442261,
      "learning_rate": 2.906370705938991e-05,
      "loss": 1.1254,
      "step": 375
    },
    {
      "epoch": 1.358642251004913,
      "grad_norm": 0.4463523328304291,
      "learning_rate": 2.8600083850140692e-05,
      "loss": 1.1275,
      "step": 380
    },
    {
      "epoch": 1.3765073693613221,
      "grad_norm": 0.4577709436416626,
      "learning_rate": 2.8135192730072598e-05,
      "loss": 1.0984,
      "step": 385
    },
    {
      "epoch": 1.3943724877177313,
      "grad_norm": 0.3673183023929596,
      "learning_rate": 2.7669197428838972e-05,
      "loss": 1.1126,
      "step": 390
    },
    {
      "epoch": 1.4122376060741402,
      "grad_norm": 0.3968072235584259,
      "learning_rate": 2.7202262064973873e-05,
      "loss": 1.1271,
      "step": 395
    },
    {
      "epoch": 1.4301027244305493,
      "grad_norm": 0.5759745240211487,
      "learning_rate": 2.6734551088091293e-05,
      "loss": 1.1135,
      "step": 400
    },
    {
      "epoch": 1.4479678427869584,
      "grad_norm": 0.4903964698314667,
      "learning_rate": 2.6266229220967818e-05,
      "loss": 1.1163,
      "step": 405
    },
    {
      "epoch": 1.4658329611433676,
      "grad_norm": 0.4729979932308197,
      "learning_rate": 2.5797461401529e-05,
      "loss": 1.1011,
      "step": 410
    },
    {
      "epoch": 1.4836980794997767,
      "grad_norm": 0.45003455877304077,
      "learning_rate": 2.5328412724759976e-05,
      "loss": 1.1506,
      "step": 415
    },
    {
      "epoch": 1.5015631978561856,
      "grad_norm": 0.4313727021217346,
      "learning_rate": 2.485924838456086e-05,
      "loss": 1.1257,
      "step": 420
    },
    {
      "epoch": 1.5194283162125948,
      "grad_norm": 0.39038538932800293,
      "learning_rate": 2.4390133615567136e-05,
      "loss": 1.0858,
      "step": 425
    },
    {
      "epoch": 1.537293434569004,
      "grad_norm": 0.49707311391830444,
      "learning_rate": 2.3921233634955868e-05,
      "loss": 1.1602,
      "step": 430
    },
    {
      "epoch": 1.555158552925413,
      "grad_norm": 0.46587345004081726,
      "learning_rate": 2.3452713584257955e-05,
      "loss": 1.1144,
      "step": 435
    },
    {
      "epoch": 1.5730236712818222,
      "grad_norm": 0.37671202421188354,
      "learning_rate": 2.2984738471197077e-05,
      "loss": 1.1793,
      "step": 440
    },
    {
      "epoch": 1.5908887896382313,
      "grad_norm": 0.4020828306674957,
      "learning_rate": 2.2517473111575725e-05,
      "loss": 1.1206,
      "step": 445
    },
    {
      "epoch": 1.6087539079946405,
      "grad_norm": 0.5164027214050293,
      "learning_rate": 2.2051082071228854e-05,
      "loss": 1.1177,
      "step": 450
    },
    {
      "epoch": 1.6266190263510496,
      "grad_norm": 0.3768448829650879,
      "learning_rate": 2.1585729608065595e-05,
      "loss": 1.0802,
      "step": 455
    },
    {
      "epoch": 1.6444841447074587,
      "grad_norm": 0.3956853747367859,
      "learning_rate": 2.1121579614219345e-05,
      "loss": 1.0913,
      "step": 460
    },
    {
      "epoch": 1.6623492630638679,
      "grad_norm": 0.36919939517974854,
      "learning_rate": 2.0658795558326743e-05,
      "loss": 1.0896,
      "step": 465
    },
    {
      "epoch": 1.680214381420277,
      "grad_norm": 0.45556601881980896,
      "learning_rate": 2.0197540427955777e-05,
      "loss": 1.1183,
      "step": 470
    },
    {
      "epoch": 1.6980794997766862,
      "grad_norm": 0.47595128417015076,
      "learning_rate": 1.97379766722033e-05,
      "loss": 1.2037,
      "step": 475
    },
    {
      "epoch": 1.7159446181330953,
      "grad_norm": 0.3943460285663605,
      "learning_rate": 1.928026614448221e-05,
      "loss": 1.0682,
      "step": 480
    },
    {
      "epoch": 1.7338097364895042,
      "grad_norm": 0.40548044443130493,
      "learning_rate": 1.8824570045518423e-05,
      "loss": 1.0664,
      "step": 485
    },
    {
      "epoch": 1.7516748548459133,
      "grad_norm": 0.4159083962440491,
      "learning_rate": 1.8371048866577713e-05,
      "loss": 1.1475,
      "step": 490
    },
    {
      "epoch": 1.7695399732023225,
      "grad_norm": 0.4973352551460266,
      "learning_rate": 1.79198623329424e-05,
      "loss": 1.0675,
      "step": 495
    },
    {
      "epoch": 1.7874050915587316,
      "grad_norm": 0.4413686692714691,
      "learning_rate": 1.747116934765785e-05,
      "loss": 1.0956,
      "step": 500
    },
    {
      "epoch": 1.8052702099151405,
      "grad_norm": 0.4255582094192505,
      "learning_rate": 1.70251279355685e-05,
      "loss": 1.071,
      "step": 505
    },
    {
      "epoch": 1.8231353282715497,
      "grad_norm": 0.7180905938148499,
      "learning_rate": 1.658189518766322e-05,
      "loss": 1.1428,
      "step": 510
    },
    {
      "epoch": 1.8410004466279588,
      "grad_norm": 0.6441544890403748,
      "learning_rate": 1.6141627205749563e-05,
      "loss": 1.1014,
      "step": 515
    },
    {
      "epoch": 1.858865564984368,
      "grad_norm": 0.415781170129776,
      "learning_rate": 1.5704479047476377e-05,
      "loss": 1.1245,
      "step": 520
    },
    {
      "epoch": 1.876730683340777,
      "grad_norm": 0.40880969166755676,
      "learning_rate": 1.5270604671724188e-05,
      "loss": 1.0354,
      "step": 525
    },
    {
      "epoch": 1.8945958016971862,
      "grad_norm": 0.5105963349342346,
      "learning_rate": 1.4840156884382491e-05,
      "loss": 1.0717,
      "step": 530
    },
    {
      "epoch": 1.9124609200535954,
      "grad_norm": 0.6085048317909241,
      "learning_rate": 1.4413287284533228e-05,
      "loss": 1.1494,
      "step": 535
    },
    {
      "epoch": 1.9303260384100045,
      "grad_norm": 0.6034563183784485,
      "learning_rate": 1.399014621105914e-05,
      "loss": 1.1347,
      "step": 540
    },
    {
      "epoch": 1.9481911567664136,
      "grad_norm": 0.5377976298332214,
      "learning_rate": 1.3570882689696127e-05,
      "loss": 1.1074,
      "step": 545
    },
    {
      "epoch": 1.9660562751228228,
      "grad_norm": 0.6293488144874573,
      "learning_rate": 1.3155644380547877e-05,
      "loss": 1.1264,
      "step": 550
    },
    {
      "epoch": 1.983921393479232,
      "grad_norm": 0.45377108454704285,
      "learning_rate": 1.2744577526081666e-05,
      "loss": 1.0915,
      "step": 555
    },
    {
      "epoch": 2.0026797677534613,
      "grad_norm": 0.38156697154045105,
      "learning_rate": 1.2337826899623265e-05,
      "loss": 1.1034,
      "step": 560
    },
    {
      "epoch": 2.0205448861098705,
      "grad_norm": 0.5213829278945923,
      "learning_rate": 1.193553575436935e-05,
      "loss": 1.0915,
      "step": 565
    },
    {
      "epoch": 2.0384100044662796,
      "grad_norm": 0.46200066804885864,
      "learning_rate": 1.1537845772935279e-05,
      "loss": 1.0797,
      "step": 570
    },
    {
      "epoch": 2.0562751228226888,
      "grad_norm": 0.6238696575164795,
      "learning_rate": 1.1144897017456e-05,
      "loss": 1.086,
      "step": 575
    },
    {
      "epoch": 2.074140241179098,
      "grad_norm": 0.45493337512016296,
      "learning_rate": 1.0756827880257661e-05,
      "loss": 1.1068,
      "step": 580
    },
    {
      "epoch": 2.092005359535507,
      "grad_norm": 0.535967230796814,
      "learning_rate": 1.0373775035117305e-05,
      "loss": 1.0574,
      "step": 585
    },
    {
      "epoch": 2.109870477891916,
      "grad_norm": 0.42889833450317383,
      "learning_rate": 9.995873389127895e-06,
      "loss": 1.1409,
      "step": 590
    },
    {
      "epoch": 2.1277355962483253,
      "grad_norm": 0.3753083646297455,
      "learning_rate": 9.623256035185382e-06,
      "loss": 1.1461,
      "step": 595
    },
    {
      "epoch": 2.1456007146047344,
      "grad_norm": 0.508271336555481,
      "learning_rate": 9.256054205114939e-06,
      "loss": 1.1039,
      "step": 600
    },
    {
      "epoch": 2.163465832961143,
      "grad_norm": 0.45634639263153076,
      "learning_rate": 8.894397223452453e-06,
      "loss": 1.0832,
      "step": 605
    },
    {
      "epoch": 2.1813309513175523,
      "grad_norm": 0.557949423789978,
      "learning_rate": 8.538412461897807e-06,
      "loss": 1.0621,
      "step": 610
    },
    {
      "epoch": 2.1991960696739614,
      "grad_norm": 0.5253410935401917,
      "learning_rate": 8.188225294455992e-06,
      "loss": 1.1078,
      "step": 615
    },
    {
      "epoch": 2.2170611880303706,
      "grad_norm": 0.4432696998119354,
      "learning_rate": 7.843959053281663e-06,
      "loss": 1.114,
      "step": 620
    },
    {
      "epoch": 2.2349263063867797,
      "grad_norm": 0.4199691712856293,
      "learning_rate": 7.505734985242927e-06,
      "loss": 1.0363,
      "step": 625
    },
    {
      "epoch": 2.252791424743189,
      "grad_norm": 0.4136619567871094,
      "learning_rate": 7.173672209219495e-06,
      "loss": 1.0577,
      "step": 630
    },
    {
      "epoch": 2.270656543099598,
      "grad_norm": 0.589055061340332,
      "learning_rate": 6.847887674150266e-06,
      "loss": 1.1141,
      "step": 635
    },
    {
      "epoch": 2.288521661456007,
      "grad_norm": 0.42672625184059143,
      "learning_rate": 6.528496117845215e-06,
      "loss": 1.0775,
      "step": 640
    },
    {
      "epoch": 2.3063867798124162,
      "grad_norm": 0.6587265133857727,
      "learning_rate": 6.215610026575916e-06,
      "loss": 1.1016,
      "step": 645
    },
    {
      "epoch": 2.3242518981688254,
      "grad_norm": 0.4289712607860565,
      "learning_rate": 5.909339595459123e-06,
      "loss": 1.0545,
      "step": 650
    },
    {
      "epoch": 2.3421170165252345,
      "grad_norm": 0.4478652775287628,
      "learning_rate": 5.609792689647222e-06,
      "loss": 1.1381,
      "step": 655
    },
    {
      "epoch": 2.3599821348816437,
      "grad_norm": 0.529575765132904,
      "learning_rate": 5.317074806339295e-06,
      "loss": 1.0688,
      "step": 660
    },
    {
      "epoch": 2.377847253238053,
      "grad_norm": 0.5171645879745483,
      "learning_rate": 5.0312890376261415e-06,
      "loss": 1.0419,
      "step": 665
    },
    {
      "epoch": 2.395712371594462,
      "grad_norm": 0.5134549140930176,
      "learning_rate": 4.752536034182312e-06,
      "loss": 1.0742,
      "step": 670
    },
    {
      "epoch": 2.413577489950871,
      "grad_norm": 0.36803004145622253,
      "learning_rate": 4.480913969818098e-06,
      "loss": 1.0799,
      "step": 675
    },
    {
      "epoch": 2.43144260830728,
      "grad_norm": 0.5440959334373474,
      "learning_rate": 4.216518506903683e-06,
      "loss": 1.0485,
      "step": 680
    },
    {
      "epoch": 2.4493077266636893,
      "grad_norm": 0.5819982290267944,
      "learning_rate": 3.959442762677967e-06,
      "loss": 1.0833,
      "step": 685
    },
    {
      "epoch": 2.467172845020098,
      "grad_norm": 0.4522368609905243,
      "learning_rate": 3.70977727645363e-06,
      "loss": 1.0997,
      "step": 690
    },
    {
      "epoch": 2.485037963376507,
      "grad_norm": 0.4565070569515228,
      "learning_rate": 3.467609977730155e-06,
      "loss": 1.0652,
      "step": 695
    },
    {
      "epoch": 2.5029030817329163,
      "grad_norm": 0.46055707335472107,
      "learning_rate": 3.233026155226046e-06,
      "loss": 1.0679,
      "step": 700
    },
    {
      "epoch": 2.5207682000893255,
      "grad_norm": 0.811050534248352,
      "learning_rate": 3.0061084268410006e-06,
      "loss": 1.1097,
      "step": 705
    },
    {
      "epoch": 2.5386333184457346,
      "grad_norm": 0.4883212447166443,
      "learning_rate": 2.786936710558821e-06,
      "loss": 1.1183,
      "step": 710
    },
    {
      "epoch": 2.5564984368021437,
      "grad_norm": 0.6575244069099426,
      "learning_rate": 2.575588196301146e-06,
      "loss": 1.0509,
      "step": 715
    },
    {
      "epoch": 2.574363555158553,
      "grad_norm": 0.4288465082645416,
      "learning_rate": 2.372137318741968e-06,
      "loss": 1.0662,
      "step": 720
    },
    {
      "epoch": 2.592228673514962,
      "grad_norm": 0.628170907497406,
      "learning_rate": 2.1766557310925916e-06,
      "loss": 1.0352,
      "step": 725
    },
    {
      "epoch": 2.610093791871371,
      "grad_norm": 0.4450254440307617,
      "learning_rate": 1.989212279866079e-06,
      "loss": 1.0813,
      "step": 730
    },
    {
      "epoch": 2.6279589102277803,
      "grad_norm": 0.5844019651412964,
      "learning_rate": 1.8098729806303116e-06,
      "loss": 1.1297,
      "step": 735
    },
    {
      "epoch": 2.6458240285841894,
      "grad_norm": 0.6221369504928589,
      "learning_rate": 1.638700994757955e-06,
      "loss": 1.0704,
      "step": 740
    },
    {
      "epoch": 2.6636891469405986,
      "grad_norm": 0.49307432770729065,
      "learning_rate": 1.4757566071817607e-06,
      "loss": 1.1033,
      "step": 745
    },
    {
      "epoch": 2.6815542652970077,
      "grad_norm": 0.6794072389602661,
      "learning_rate": 1.3210972051628328e-06,
      "loss": 1.0211,
      "step": 750
    },
    {
      "epoch": 2.699419383653417,
      "grad_norm": 0.4681510329246521,
      "learning_rate": 1.1747772580794348e-06,
      "loss": 1.0676,
      "step": 755
    },
    {
      "epoch": 2.717284502009826,
      "grad_norm": 0.4018392264842987,
      "learning_rate": 1.0368482982435063e-06,
      "loss": 1.0683,
      "step": 760
    },
    {
      "epoch": 2.7351496203662347,
      "grad_norm": 0.6274233460426331,
      "learning_rate": 9.073589027514789e-07,
      "loss": 1.0712,
      "step": 765
    },
    {
      "epoch": 2.7530147387226442,
      "grad_norm": 0.4873206615447998,
      "learning_rate": 7.863546763760055e-07,
      "loss": 1.0447,
      "step": 770
    },
    {
      "epoch": 2.770879857079053,
      "grad_norm": 0.6785792708396912,
      "learning_rate": 6.738782355044049e-07,
      "loss": 1.0881,
      "step": 775
    },
    {
      "epoch": 2.7887449754354625,
      "grad_norm": 0.6314323544502258,
      "learning_rate": 5.699691931296463e-07,
      "loss": 1.0517,
      "step": 780
    },
    {
      "epoch": 2.806610093791871,
      "grad_norm": 0.5061508417129517,
      "learning_rate": 4.746641448990785e-07,
      "loss": 1.0891,
      "step": 785
    },
    {
      "epoch": 2.8244752121482803,
      "grad_norm": 0.38295778632164,
      "learning_rate": 3.879966562258364e-07,
      "loss": 1.1203,
      "step": 790
    },
    {
      "epoch": 2.8423403305046895,
      "grad_norm": 0.7023996710777283,
      "learning_rate": 3.0999725046745866e-07,
      "loss": 1.0624,
      "step": 795
    },
    {
      "epoch": 2.8602054488610986,
      "grad_norm": 0.41944465041160583,
      "learning_rate": 2.406933981758952e-07,
      "loss": 1.0512,
      "step": 800
    },
    {
      "epoch": 2.8780705672175078,
      "grad_norm": 0.6278116703033447,
      "learning_rate": 1.801095074226683e-07,
      "loss": 1.0506,
      "step": 805
    },
    {
      "epoch": 2.895935685573917,
      "grad_norm": 0.523618757724762,
      "learning_rate": 1.2826691520262114e-07,
      "loss": 1.084,
      "step": 810
    },
    {
      "epoch": 2.913800803930326,
      "grad_norm": 0.552522599697113,
      "learning_rate": 8.518387991924837e-08,
      "loss": 1.1039,
      "step": 815
    },
    {
      "epoch": 2.931665922286735,
      "grad_norm": 0.5657618045806885,
      "learning_rate": 5.0875574954287186e-08,
      "loss": 1.0787,
      "step": 820
    },
    {
      "epoch": 2.9495310406431443,
      "grad_norm": 0.4286334216594696,
      "learning_rate": 2.535408332381417e-08,
      "loss": 1.0904,
      "step": 825
    },
    {
      "epoch": 2.9673961589995534,
      "grad_norm": 0.5537897348403931,
      "learning_rate": 8.62839342274102e-09,
      "loss": 1.0673,
      "step": 830
    },
    {
      "epoch": 2.9852612773559626,
      "grad_norm": 0.4285455644130707,
      "learning_rate": 7.043958591912425e-10,
      "loss": 1.0993,
      "step": 835
    }
  ],
  "logging_steps": 5,
  "max_steps": 837,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.422989809751163e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}