{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.591129819959406,
  "eval_steps": 500,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0036945613747462877,
      "grad_norm": 2.2292165756225586,
      "learning_rate": 5.911330049261084e-06,
      "loss": 0.9395,
      "step": 25
    },
    {
      "epoch": 0.007389122749492575,
      "grad_norm": 0.8381065130233765,
      "learning_rate": 1.206896551724138e-05,
      "loss": 0.89,
      "step": 50
    },
    {
      "epoch": 0.011083684124238862,
      "grad_norm": 0.9238471984863281,
      "learning_rate": 1.8226600985221676e-05,
      "loss": 0.8384,
      "step": 75
    },
    {
      "epoch": 0.01477824549898515,
      "grad_norm": 0.6837311387062073,
      "learning_rate": 2.438423645320197e-05,
      "loss": 0.784,
      "step": 100
    },
    {
      "epoch": 0.01847280687373144,
      "grad_norm": 0.7093706727027893,
      "learning_rate": 3.0541871921182266e-05,
      "loss": 0.7675,
      "step": 125
    },
    {
      "epoch": 0.022167368248477724,
      "grad_norm": 0.6208077073097229,
      "learning_rate": 3.669950738916256e-05,
      "loss": 0.7466,
      "step": 150
    },
    {
      "epoch": 0.025861929623224013,
      "grad_norm": 0.8929094076156616,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.7386,
      "step": 175
    },
    {
      "epoch": 0.0295564909979703,
      "grad_norm": 0.7828120589256287,
      "learning_rate": 4.901477832512316e-05,
      "loss": 0.7173,
      "step": 200
    },
    {
      "epoch": 0.03325105237271659,
      "grad_norm": 0.7134449481964111,
      "learning_rate": 5.517241379310345e-05,
      "loss": 0.7108,
      "step": 225
    },
    {
      "epoch": 0.03694561374746288,
      "grad_norm": 0.7464851140975952,
      "learning_rate": 6.133004926108375e-05,
      "loss": 0.7332,
      "step": 250
    },
    {
      "epoch": 0.04064017512220916,
      "grad_norm": 0.677793025970459,
      "learning_rate": 6.748768472906404e-05,
      "loss": 0.7194,
      "step": 275
    },
    {
      "epoch": 0.04433473649695545,
      "grad_norm": 0.7936354875564575,
      "learning_rate": 7.364532019704434e-05,
      "loss": 0.7253,
      "step": 300
    },
    {
      "epoch": 0.04802929787170174,
      "grad_norm": 0.6711100935935974,
      "learning_rate": 7.980295566502463e-05,
      "loss": 0.7018,
      "step": 325
    },
    {
      "epoch": 0.051723859246448026,
      "grad_norm": 0.5816489458084106,
      "learning_rate": 8.596059113300493e-05,
      "loss": 0.7298,
      "step": 350
    },
    {
      "epoch": 0.055418420621194314,
      "grad_norm": 0.6680681705474854,
      "learning_rate": 9.211822660098522e-05,
      "loss": 0.7149,
      "step": 375
    },
    {
      "epoch": 0.0591129819959406,
      "grad_norm": 0.5643934607505798,
      "learning_rate": 9.827586206896552e-05,
      "loss": 0.763,
      "step": 400
    },
    {
      "epoch": 0.06280754337068689,
      "grad_norm": 0.5739309191703796,
      "learning_rate": 9.986286759104069e-05,
      "loss": 0.7345,
      "step": 425
    },
    {
      "epoch": 0.06650210474543318,
      "grad_norm": 0.5929909944534302,
      "learning_rate": 9.967240591193052e-05,
      "loss": 0.7364,
      "step": 450
    },
    {
      "epoch": 0.07019666612017947,
      "grad_norm": 0.609235405921936,
      "learning_rate": 9.948194423282036e-05,
      "loss": 0.7294,
      "step": 475
    },
    {
      "epoch": 0.07389122749492576,
      "grad_norm": 0.4643324613571167,
      "learning_rate": 9.92914825537102e-05,
      "loss": 0.7344,
      "step": 500
    },
    {
      "epoch": 0.07758578886967203,
      "grad_norm": 0.5267598032951355,
      "learning_rate": 9.910102087460003e-05,
      "loss": 0.7249,
      "step": 525
    },
    {
      "epoch": 0.08128035024441832,
      "grad_norm": 0.47951069474220276,
      "learning_rate": 9.891055919548987e-05,
      "loss": 0.7256,
      "step": 550
    },
    {
      "epoch": 0.08497491161916461,
      "grad_norm": 0.4505012333393097,
      "learning_rate": 9.87200975163797e-05,
      "loss": 0.7359,
      "step": 575
    },
    {
      "epoch": 0.0886694729939109,
      "grad_norm": 0.5320091247558594,
      "learning_rate": 9.852963583726955e-05,
      "loss": 0.6856,
      "step": 600
    },
    {
      "epoch": 0.09236403436865719,
      "grad_norm": 0.5583036541938782,
      "learning_rate": 9.833917415815939e-05,
      "loss": 0.7235,
      "step": 625
    },
    {
      "epoch": 0.09605859574340347,
      "grad_norm": 0.5139252543449402,
      "learning_rate": 9.814871247904922e-05,
      "loss": 0.7272,
      "step": 650
    },
    {
      "epoch": 0.09975315711814976,
      "grad_norm": 0.4989326000213623,
      "learning_rate": 9.795825079993906e-05,
      "loss": 0.6952,
      "step": 675
    },
    {
      "epoch": 0.10344771849289605,
      "grad_norm": 0.47355732321739197,
      "learning_rate": 9.776778912082889e-05,
      "loss": 0.7266,
      "step": 700
    },
    {
      "epoch": 0.10714227986764234,
      "grad_norm": 0.3588508367538452,
      "learning_rate": 9.757732744171874e-05,
      "loss": 0.7406,
      "step": 725
    },
    {
      "epoch": 0.11083684124238863,
      "grad_norm": 0.4120556712150574,
      "learning_rate": 9.738686576260857e-05,
      "loss": 0.7443,
      "step": 750
    },
    {
      "epoch": 0.11453140261713492,
      "grad_norm": 0.5160555839538574,
      "learning_rate": 9.71964040834984e-05,
      "loss": 0.7134,
      "step": 775
    },
    {
      "epoch": 0.1182259639918812,
      "grad_norm": 0.5423145890235901,
      "learning_rate": 9.700594240438823e-05,
      "loss": 0.7289,
      "step": 800
    },
    {
      "epoch": 0.1219205253666275,
      "grad_norm": 0.5352346301078796,
      "learning_rate": 9.681548072527808e-05,
      "loss": 0.7144,
      "step": 825
    },
    {
      "epoch": 0.12561508674137378,
      "grad_norm": 0.47908860445022583,
      "learning_rate": 9.662501904616791e-05,
      "loss": 0.7175,
      "step": 850
    },
    {
      "epoch": 0.12930964811612006,
      "grad_norm": 0.47986069321632385,
      "learning_rate": 9.643455736705776e-05,
      "loss": 0.6983,
      "step": 875
    },
    {
      "epoch": 0.13300420949086636,
      "grad_norm": 0.6903620958328247,
      "learning_rate": 9.624409568794759e-05,
      "loss": 0.7086,
      "step": 900
    },
    {
      "epoch": 0.13669877086561263,
      "grad_norm": 0.44413208961486816,
      "learning_rate": 9.605363400883742e-05,
      "loss": 0.7306,
      "step": 925
    },
    {
      "epoch": 0.14039333224035894,
      "grad_norm": 0.4634678065776825,
      "learning_rate": 9.586317232972727e-05,
      "loss": 0.7061,
      "step": 950
    },
    {
      "epoch": 0.1440878936151052,
      "grad_norm": 0.5110129714012146,
      "learning_rate": 9.56727106506171e-05,
      "loss": 0.7406,
      "step": 975
    },
    {
      "epoch": 0.1477824549898515,
      "grad_norm": 0.5460866093635559,
      "learning_rate": 9.548224897150694e-05,
      "loss": 0.721,
      "step": 1000
    },
    {
      "epoch": 0.1514770163645978,
      "grad_norm": 0.5179885029792786,
      "learning_rate": 9.529178729239677e-05,
      "loss": 0.7068,
      "step": 1025
    },
    {
      "epoch": 0.15517157773934406,
      "grad_norm": 0.40280836820602417,
      "learning_rate": 9.51013256132866e-05,
      "loss": 0.7009,
      "step": 1050
    },
    {
      "epoch": 0.15886613911409037,
      "grad_norm": 1.2706756591796875,
      "learning_rate": 9.491086393417645e-05,
      "loss": 0.7125,
      "step": 1075
    },
    {
      "epoch": 0.16256070048883664,
      "grad_norm": 0.4963163435459137,
      "learning_rate": 9.47204022550663e-05,
      "loss": 0.7149,
      "step": 1100
    },
    {
      "epoch": 0.16625526186358294,
      "grad_norm": 0.5147728323936462,
      "learning_rate": 9.452994057595613e-05,
      "loss": 0.7245,
      "step": 1125
    },
    {
      "epoch": 0.16994982323832922,
      "grad_norm": 0.5933899879455566,
      "learning_rate": 9.433947889684596e-05,
      "loss": 0.7261,
      "step": 1150
    },
    {
      "epoch": 0.17364438461307552,
      "grad_norm": 0.4750466048717499,
      "learning_rate": 9.414901721773579e-05,
      "loss": 0.7206,
      "step": 1175
    },
    {
      "epoch": 0.1773389459878218,
      "grad_norm": 0.46546968817710876,
      "learning_rate": 9.395855553862564e-05,
      "loss": 0.7121,
      "step": 1200
    },
    {
      "epoch": 0.1810335073625681,
      "grad_norm": 0.6512172818183899,
      "learning_rate": 9.376809385951547e-05,
      "loss": 0.7212,
      "step": 1225
    },
    {
      "epoch": 0.18472806873731437,
      "grad_norm": 0.34932607412338257,
      "learning_rate": 9.35776321804053e-05,
      "loss": 0.704,
      "step": 1250
    },
    {
      "epoch": 0.18842263011206067,
      "grad_norm": 0.4648846983909607,
      "learning_rate": 9.338717050129514e-05,
      "loss": 0.7419,
      "step": 1275
    },
    {
      "epoch": 0.19211719148680695,
      "grad_norm": 0.4566064774990082,
      "learning_rate": 9.319670882218498e-05,
      "loss": 0.7318,
      "step": 1300
    },
    {
      "epoch": 0.19581175286155325,
      "grad_norm": 0.5357668399810791,
      "learning_rate": 9.300624714307481e-05,
      "loss": 0.6973,
      "step": 1325
    },
    {
      "epoch": 0.19950631423629953,
      "grad_norm": 0.4423241913318634,
      "learning_rate": 9.281578546396466e-05,
      "loss": 0.7298,
      "step": 1350
    },
    {
      "epoch": 0.20320087561104583,
      "grad_norm": 0.4530033767223358,
      "learning_rate": 9.26253237848545e-05,
      "loss": 0.7161,
      "step": 1375
    },
    {
      "epoch": 0.2068954369857921,
      "grad_norm": 0.4678841233253479,
      "learning_rate": 9.243486210574433e-05,
      "loss": 0.6972,
      "step": 1400
    },
    {
      "epoch": 0.21058999836053838,
      "grad_norm": 0.6039907336235046,
      "learning_rate": 9.224440042663417e-05,
      "loss": 0.7165,
      "step": 1425
    },
    {
      "epoch": 0.21428455973528468,
      "grad_norm": 0.4463271498680115,
      "learning_rate": 9.2053938747524e-05,
      "loss": 0.6863,
      "step": 1450
    },
    {
      "epoch": 0.21797912111003095,
      "grad_norm": 0.5739301443099976,
      "learning_rate": 9.186347706841384e-05,
      "loss": 0.6907,
      "step": 1475
    },
    {
      "epoch": 0.22167368248477726,
      "grad_norm": 0.4577805697917938,
      "learning_rate": 9.167301538930367e-05,
      "loss": 0.7114,
      "step": 1500
    },
    {
      "epoch": 0.22536824385952353,
      "grad_norm": 0.4522150158882141,
      "learning_rate": 9.14825537101935e-05,
      "loss": 0.6877,
      "step": 1525
    },
    {
      "epoch": 0.22906280523426983,
      "grad_norm": 0.49612903594970703,
      "learning_rate": 9.129209203108335e-05,
      "loss": 0.7112,
      "step": 1550
    },
    {
      "epoch": 0.2327573666090161,
      "grad_norm": 0.4710284471511841,
      "learning_rate": 9.11016303519732e-05,
      "loss": 0.7062,
      "step": 1575
    },
    {
      "epoch": 0.2364519279837624,
      "grad_norm": 0.5009223818778992,
      "learning_rate": 9.091116867286303e-05,
      "loss": 0.7275,
      "step": 1600
    },
    {
      "epoch": 0.24014648935850869,
      "grad_norm": 0.5547946691513062,
      "learning_rate": 9.072070699375286e-05,
      "loss": 0.6993,
      "step": 1625
    },
    {
      "epoch": 0.243841050733255,
      "grad_norm": 0.4580361843109131,
      "learning_rate": 9.05302453146427e-05,
      "loss": 0.7106,
      "step": 1650
    },
    {
      "epoch": 0.24753561210800126,
      "grad_norm": 0.4767173230648041,
      "learning_rate": 9.033978363553254e-05,
      "loss": 0.7103,
      "step": 1675
    },
    {
      "epoch": 0.25123017348274757,
      "grad_norm": 0.502202570438385,
      "learning_rate": 9.014932195642237e-05,
      "loss": 0.6921,
      "step": 1700
    },
    {
      "epoch": 0.25492473485749384,
      "grad_norm": 0.5283953547477722,
      "learning_rate": 8.99588602773122e-05,
      "loss": 0.7077,
      "step": 1725
    },
    {
      "epoch": 0.2586192962322401,
      "grad_norm": 0.4994209408760071,
      "learning_rate": 8.976839859820204e-05,
      "loss": 0.7,
      "step": 1750
    },
    {
      "epoch": 0.2623138576069864,
      "grad_norm": 0.48279210925102234,
      "learning_rate": 8.957793691909188e-05,
      "loss": 0.7114,
      "step": 1775
    },
    {
      "epoch": 0.2660084189817327,
      "grad_norm": 0.6055914759635925,
      "learning_rate": 8.938747523998172e-05,
      "loss": 0.7081,
      "step": 1800
    },
    {
      "epoch": 0.269702980356479,
      "grad_norm": 0.489519327878952,
      "learning_rate": 8.919701356087156e-05,
      "loss": 0.6946,
      "step": 1825
    },
    {
      "epoch": 0.27339754173122527,
      "grad_norm": 0.5379961133003235,
      "learning_rate": 8.90065518817614e-05,
      "loss": 0.6996,
      "step": 1850
    },
    {
      "epoch": 0.27709210310597154,
      "grad_norm": 0.47824332118034363,
      "learning_rate": 8.881609020265123e-05,
      "loss": 0.6881,
      "step": 1875
    },
    {
      "epoch": 0.2807866644807179,
      "grad_norm": 1551.0594482421875,
      "learning_rate": 8.862562852354107e-05,
      "loss": 0.6998,
      "step": 1900
    },
    {
      "epoch": 0.28448122585546415,
      "grad_norm": 0.4107681214809418,
      "learning_rate": 8.84351668444309e-05,
      "loss": 0.7088,
      "step": 1925
    },
    {
      "epoch": 0.2881757872302104,
      "grad_norm": 0.4558309316635132,
      "learning_rate": 8.824470516532074e-05,
      "loss": 0.712,
      "step": 1950
    },
    {
      "epoch": 0.2918703486049567,
      "grad_norm": 0.539107620716095,
      "learning_rate": 8.805424348621057e-05,
      "loss": 0.7159,
      "step": 1975
    },
    {
      "epoch": 0.295564909979703,
      "grad_norm": 0.5768142938613892,
      "learning_rate": 8.786378180710042e-05,
      "loss": 0.7072,
      "step": 2000
    },
    {
      "epoch": 0.2992594713544493,
      "grad_norm": 0.5575465559959412,
      "learning_rate": 8.767332012799025e-05,
      "loss": 0.7118,
      "step": 2025
    },
    {
      "epoch": 0.3029540327291956,
      "grad_norm": 0.5190144181251526,
      "learning_rate": 8.748285844888008e-05,
      "loss": 0.7099,
      "step": 2050
    },
    {
      "epoch": 0.30664859410394185,
      "grad_norm": 0.4934520125389099,
      "learning_rate": 8.729239676976993e-05,
      "loss": 0.692,
      "step": 2075
    },
    {
      "epoch": 0.3103431554786881,
      "grad_norm": 0.42613571882247925,
      "learning_rate": 8.710193509065976e-05,
      "loss": 0.7277,
      "step": 2100
    },
    {
      "epoch": 0.31403771685343446,
      "grad_norm": 0.5124602317810059,
      "learning_rate": 8.691147341154961e-05,
      "loss": 0.6801,
      "step": 2125
    },
    {
      "epoch": 0.31773227822818073,
      "grad_norm": 0.5284898281097412,
      "learning_rate": 8.672101173243944e-05,
      "loss": 0.7103,
      "step": 2150
    },
    {
      "epoch": 0.321426839602927,
      "grad_norm": 0.43099457025527954,
      "learning_rate": 8.653055005332927e-05,
      "loss": 0.7023,
      "step": 2175
    },
    {
      "epoch": 0.3251214009776733,
      "grad_norm": 0.5190865993499756,
      "learning_rate": 8.63400883742191e-05,
      "loss": 0.7144,
      "step": 2200
    },
    {
      "epoch": 0.3288159623524196,
      "grad_norm": 0.4722968637943268,
      "learning_rate": 8.614962669510895e-05,
      "loss": 0.7351,
      "step": 2225
    },
    {
      "epoch": 0.3325105237271659,
      "grad_norm": 0.6091466546058655,
      "learning_rate": 8.595916501599878e-05,
      "loss": 0.7062,
      "step": 2250
    },
    {
      "epoch": 0.33620508510191216,
      "grad_norm": 0.6135897040367126,
      "learning_rate": 8.576870333688862e-05,
      "loss": 0.7117,
      "step": 2275
    },
    {
      "epoch": 0.33989964647665843,
      "grad_norm": 0.5224157571792603,
      "learning_rate": 8.557824165777846e-05,
      "loss": 0.7068,
      "step": 2300
    },
    {
      "epoch": 0.34359420785140476,
      "grad_norm": 0.4863536059856415,
      "learning_rate": 8.53877799786683e-05,
      "loss": 0.6952,
      "step": 2325
    },
    {
      "epoch": 0.34728876922615104,
      "grad_norm": 0.4728885889053345,
      "learning_rate": 8.519731829955814e-05,
      "loss": 0.7289,
      "step": 2350
    },
    {
      "epoch": 0.3509833306008973,
      "grad_norm": 0.5152695775032043,
      "learning_rate": 8.500685662044798e-05,
      "loss": 0.6986,
      "step": 2375
    },
    {
      "epoch": 0.3546778919756436,
      "grad_norm": 0.4407690465450287,
      "learning_rate": 8.481639494133781e-05,
      "loss": 0.6983,
      "step": 2400
    },
    {
      "epoch": 0.3583724533503899,
      "grad_norm": 0.4756406545639038,
      "learning_rate": 8.462593326222764e-05,
      "loss": 0.6626,
      "step": 2425
    },
    {
      "epoch": 0.3620670147251362,
      "grad_norm": 0.48595255613327026,
      "learning_rate": 8.443547158311747e-05,
      "loss": 0.6892,
      "step": 2450
    },
    {
      "epoch": 0.36576157609988247,
      "grad_norm": 0.47830772399902344,
      "learning_rate": 8.424500990400732e-05,
      "loss": 0.7162,
      "step": 2475
    },
    {
      "epoch": 0.36945613747462874,
      "grad_norm": 0.43090149760246277,
      "learning_rate": 8.405454822489715e-05,
      "loss": 0.6913,
      "step": 2500
    },
    {
      "epoch": 0.373150698849375,
      "grad_norm": 0.5338718295097351,
      "learning_rate": 8.386408654578698e-05,
      "loss": 0.7159,
      "step": 2525
    },
    {
      "epoch": 0.37684526022412135,
      "grad_norm": 0.4907350242137909,
      "learning_rate": 8.367362486667683e-05,
      "loss": 0.7022,
      "step": 2550
    },
    {
      "epoch": 0.3805398215988676,
      "grad_norm": 0.44093430042266846,
      "learning_rate": 8.348316318756668e-05,
      "loss": 0.7124,
      "step": 2575
    },
    {
      "epoch": 0.3842343829736139,
      "grad_norm": 0.5388796925544739,
      "learning_rate": 8.329270150845651e-05,
      "loss": 0.7304,
      "step": 2600
    },
    {
      "epoch": 0.38792894434836017,
      "grad_norm": 0.4456349313259125,
      "learning_rate": 8.310223982934634e-05,
      "loss": 0.6964,
      "step": 2625
    },
    {
      "epoch": 0.3916235057231065,
      "grad_norm": 0.4602237343788147,
      "learning_rate": 8.291177815023617e-05,
      "loss": 0.6707,
      "step": 2650
    },
    {
      "epoch": 0.3953180670978528,
      "grad_norm": 0.4726378917694092,
      "learning_rate": 8.272131647112601e-05,
      "loss": 0.694,
      "step": 2675
    },
    {
      "epoch": 0.39901262847259905,
      "grad_norm": 0.500451922416687,
      "learning_rate": 8.253085479201585e-05,
      "loss": 0.6812,
      "step": 2700
    },
    {
      "epoch": 0.4027071898473453,
      "grad_norm": 0.4073813259601593,
      "learning_rate": 8.234039311290569e-05,
      "loss": 0.7028,
      "step": 2725
    },
    {
      "epoch": 0.40640175122209166,
      "grad_norm": 0.43644702434539795,
      "learning_rate": 8.214993143379552e-05,
      "loss": 0.6936,
      "step": 2750
    },
    {
      "epoch": 0.41009631259683793,
      "grad_norm": 0.5256789922714233,
      "learning_rate": 8.195946975468535e-05,
      "loss": 0.7035,
      "step": 2775
    },
    {
      "epoch": 0.4137908739715842,
      "grad_norm": 0.48385316133499146,
      "learning_rate": 8.17690080755752e-05,
      "loss": 0.6736,
      "step": 2800
    },
    {
      "epoch": 0.4174854353463305,
      "grad_norm": 0.49825233221054077,
      "learning_rate": 8.157854639646504e-05,
      "loss": 0.7019,
      "step": 2825
    },
    {
      "epoch": 0.42117999672107675,
      "grad_norm": 0.5086994171142578,
      "learning_rate": 8.138808471735488e-05,
      "loss": 0.6918,
      "step": 2850
    },
    {
      "epoch": 0.4248745580958231,
      "grad_norm": 0.5430259108543396,
      "learning_rate": 8.119762303824471e-05,
      "loss": 0.7175,
      "step": 2875
    },
    {
      "epoch": 0.42856911947056936,
      "grad_norm": 0.5889118313789368,
      "learning_rate": 8.100716135913454e-05,
      "loss": 0.6967,
      "step": 2900
    },
    {
      "epoch": 0.43226368084531563,
      "grad_norm": 0.5345672369003296,
      "learning_rate": 8.081669968002439e-05,
      "loss": 0.6926,
      "step": 2925
    },
    {
      "epoch": 0.4359582422200619,
      "grad_norm": 0.511101245880127,
      "learning_rate": 8.062623800091422e-05,
      "loss": 0.7248,
      "step": 2950
    },
    {
      "epoch": 0.43965280359480824,
      "grad_norm": 0.511239767074585,
      "learning_rate": 8.043577632180405e-05,
      "loss": 0.7319,
      "step": 2975
    },
    {
      "epoch": 0.4433473649695545,
      "grad_norm": 0.5121573805809021,
      "learning_rate": 8.024531464269389e-05,
      "loss": 0.7023,
      "step": 3000
    },
    {
      "epoch": 0.4470419263443008,
      "grad_norm": 0.5658753514289856,
      "learning_rate": 8.005485296358373e-05,
      "loss": 0.6934,
      "step": 3025
    },
    {
      "epoch": 0.45073648771904706,
      "grad_norm": 0.5475583672523499,
      "learning_rate": 7.986439128447358e-05,
      "loss": 0.6701,
      "step": 3050
    },
    {
      "epoch": 0.4544310490937934,
      "grad_norm": 0.6107661724090576,
      "learning_rate": 7.967392960536341e-05,
      "loss": 0.7056,
      "step": 3075
    },
    {
      "epoch": 0.45812561046853967,
      "grad_norm": 0.48424115777015686,
      "learning_rate": 7.948346792625324e-05,
      "loss": 0.7179,
      "step": 3100
    },
    {
      "epoch": 0.46182017184328594,
      "grad_norm": 0.6184881329536438,
      "learning_rate": 7.929300624714308e-05,
      "loss": 0.7141,
      "step": 3125
    },
    {
      "epoch": 0.4655147332180322,
      "grad_norm": 0.49919527769088745,
      "learning_rate": 7.910254456803291e-05,
      "loss": 0.7027,
      "step": 3150
    },
    {
      "epoch": 0.4692092945927785,
      "grad_norm": 0.5012905597686768,
      "learning_rate": 7.891208288892276e-05,
      "loss": 0.6852,
      "step": 3175
    },
    {
      "epoch": 0.4729038559675248,
      "grad_norm": 0.5033735036849976,
      "learning_rate": 7.872162120981259e-05,
      "loss": 0.7199,
      "step": 3200
    },
    {
      "epoch": 0.4765984173422711,
      "grad_norm": 0.5746079087257385,
      "learning_rate": 7.853115953070242e-05,
      "loss": 0.715,
      "step": 3225
    },
    {
      "epoch": 0.48029297871701737,
      "grad_norm": 0.5436145663261414,
      "learning_rate": 7.834069785159225e-05,
      "loss": 0.697,
      "step": 3250
    },
    {
      "epoch": 0.48398754009176365,
      "grad_norm": 0.5836604833602905,
      "learning_rate": 7.81502361724821e-05,
      "loss": 0.7129,
      "step": 3275
    },
    {
      "epoch": 0.48768210146651,
      "grad_norm": 0.5618935227394104,
      "learning_rate": 7.795977449337195e-05,
      "loss": 0.6993,
      "step": 3300
    },
    {
      "epoch": 0.49137666284125625,
      "grad_norm": 0.6251245737075806,
      "learning_rate": 7.776931281426178e-05,
      "loss": 0.7132,
      "step": 3325
    },
    {
      "epoch": 0.4950712242160025,
      "grad_norm": 0.5123202800750732,
      "learning_rate": 7.757885113515161e-05,
      "loss": 0.7218,
      "step": 3350
    },
    {
      "epoch": 0.4987657855907488,
      "grad_norm": 0.5818086862564087,
      "learning_rate": 7.738838945604144e-05,
      "loss": 0.7154,
      "step": 3375
    },
    {
      "epoch": 0.5024603469654951,
      "grad_norm": 0.5861947536468506,
      "learning_rate": 7.719792777693129e-05,
      "loss": 0.7013,
      "step": 3400
    },
    {
      "epoch": 0.5061549083402413,
      "grad_norm": 0.4849907457828522,
      "learning_rate": 7.700746609782112e-05,
      "loss": 0.6902,
      "step": 3425
    },
    {
      "epoch": 0.5098494697149877,
      "grad_norm": 0.5476916432380676,
      "learning_rate": 7.681700441871096e-05,
      "loss": 0.7147,
      "step": 3450
    },
    {
      "epoch": 0.513544031089734,
      "grad_norm": 0.5822548866271973,
      "learning_rate": 7.662654273960079e-05,
      "loss": 0.7431,
      "step": 3475
    },
    {
      "epoch": 0.5172385924644802,
      "grad_norm": 0.44818004965782166,
      "learning_rate": 7.643608106049063e-05,
      "loss": 0.718,
      "step": 3500
    },
    {
      "epoch": 0.5209331538392266,
      "grad_norm": 0.544068455696106,
      "learning_rate": 7.624561938138047e-05,
      "loss": 0.7201,
      "step": 3525
    },
    {
      "epoch": 0.5246277152139728,
      "grad_norm": 0.5535098910331726,
      "learning_rate": 7.605515770227031e-05,
      "loss": 0.6998,
      "step": 3550
    },
    {
      "epoch": 0.5283222765887191,
      "grad_norm": 0.5754445195198059,
      "learning_rate": 7.586469602316015e-05,
      "loss": 0.6918,
      "step": 3575
    },
    {
      "epoch": 0.5320168379634654,
      "grad_norm": 0.5976133942604065,
      "learning_rate": 7.567423434404998e-05,
      "loss": 0.693,
      "step": 3600
    },
    {
      "epoch": 0.5357113993382117,
      "grad_norm": 0.4844263195991516,
      "learning_rate": 7.548377266493982e-05,
      "loss": 0.68,
      "step": 3625
    },
    {
      "epoch": 0.539405960712958,
      "grad_norm": 0.5436462163925171,
      "learning_rate": 7.529331098582966e-05,
      "loss": 0.7075,
      "step": 3650
    },
    {
      "epoch": 0.5431005220877043,
      "grad_norm": 0.6490929126739502,
      "learning_rate": 7.510284930671949e-05,
      "loss": 0.6956,
      "step": 3675
    },
    {
      "epoch": 0.5467950834624505,
      "grad_norm": 0.47479814291000366,
      "learning_rate": 7.491238762760932e-05,
      "loss": 0.7026,
      "step": 3700
    },
    {
      "epoch": 0.5504896448371969,
      "grad_norm": 0.590874969959259,
      "learning_rate": 7.472192594849915e-05,
      "loss": 0.7117,
      "step": 3725
    },
    {
      "epoch": 0.5541842062119431,
      "grad_norm": 0.46487829089164734,
      "learning_rate": 7.4531464269389e-05,
      "loss": 0.6919,
      "step": 3750
    },
    {
      "epoch": 0.5578787675866894,
      "grad_norm": 0.6609780192375183,
      "learning_rate": 7.434100259027885e-05,
      "loss": 0.7089,
      "step": 3775
    },
    {
      "epoch": 0.5615733289614357,
      "grad_norm": 0.6165657639503479,
      "learning_rate": 7.415054091116868e-05,
      "loss": 0.7444,
      "step": 3800
    },
    {
      "epoch": 0.565267890336182,
      "grad_norm": 0.5194655656814575,
      "learning_rate": 7.396007923205851e-05,
      "loss": 0.7114,
      "step": 3825
    },
    {
      "epoch": 0.5689624517109283,
      "grad_norm": 0.4813441336154938,
      "learning_rate": 7.376961755294836e-05,
      "loss": 0.6953,
      "step": 3850
    },
    {
      "epoch": 0.5726570130856745,
      "grad_norm": 0.7607313990592957,
      "learning_rate": 7.357915587383819e-05,
      "loss": 0.7058,
      "step": 3875
    },
    {
      "epoch": 0.5763515744604208,
      "grad_norm": 0.48498719930648804,
      "learning_rate": 7.338869419472802e-05,
      "loss": 0.6866,
      "step": 3900
    },
    {
      "epoch": 0.5800461358351672,
      "grad_norm": 0.5969393253326416,
      "learning_rate": 7.319823251561786e-05,
      "loss": 0.6951,
      "step": 3925
    },
    {
      "epoch": 0.5837406972099134,
      "grad_norm": 0.6178887486457825,
      "learning_rate": 7.300777083650769e-05,
      "loss": 0.7036,
      "step": 3950
    },
    {
      "epoch": 0.5874352585846597,
      "grad_norm": 0.5318612456321716,
      "learning_rate": 7.281730915739754e-05,
      "loss": 0.6856,
      "step": 3975
    },
    {
      "epoch": 0.591129819959406,
      "grad_norm": 0.6101936101913452,
      "learning_rate": 7.262684747828737e-05,
      "loss": 0.7103,
      "step": 4000
    }
  ],
  "logging_steps": 25,
  "max_steps": 13532,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.6005238537023461e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}