{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500.0,
  "global_step": 939,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010659560293137908,
      "grad_norm": 65.16799013627562,
      "learning_rate": 1.7241379310344828e-07,
      "loss": 1.7501227855682373,
      "step": 1,
      "token_acc": 0.5974304068522484
    },
    {
      "epoch": 0.005329780146568954,
      "grad_norm": 49.60785562728193,
      "learning_rate": 8.620689655172415e-07,
      "loss": 1.9290522336959839,
      "step": 5,
      "token_acc": 0.6
    },
    {
      "epoch": 0.010659560293137908,
      "grad_norm": 19.615574554346505,
      "learning_rate": 1.724137931034483e-06,
      "loss": 1.787750244140625,
      "step": 10,
      "token_acc": 0.6390234811777861
    },
    {
      "epoch": 0.015989340439706862,
      "grad_norm": 11.065195164896805,
      "learning_rate": 2.5862068965517246e-06,
      "loss": 1.4807558059692383,
      "step": 15,
      "token_acc": 0.6601166054847765
    },
    {
      "epoch": 0.021319120586275817,
      "grad_norm": 6.817900936636832,
      "learning_rate": 3.448275862068966e-06,
      "loss": 1.398179054260254,
      "step": 20,
      "token_acc": 0.6557256131292643
    },
    {
      "epoch": 0.02664890073284477,
      "grad_norm": 6.32783528737025,
      "learning_rate": 4.310344827586207e-06,
      "loss": 1.3023101806640625,
      "step": 25,
      "token_acc": 0.6763397175469142
    },
    {
      "epoch": 0.031978680879413725,
      "grad_norm": 6.367128597932351,
      "learning_rate": 4.9999851020489695e-06,
      "loss": 1.2105424880981446,
      "step": 30,
      "token_acc": 0.6806692125841062
    },
    {
      "epoch": 0.037308461025982675,
      "grad_norm": 5.745679317649865,
      "learning_rate": 4.99946369240638e-06,
      "loss": 1.0424203872680664,
      "step": 35,
      "token_acc": 0.720532319391635
    },
    {
      "epoch": 0.04263824117255163,
      "grad_norm": 5.84995920075286,
      "learning_rate": 4.998197562761914e-06,
      "loss": 1.0592132568359376,
      "step": 40,
      "token_acc": 0.7028319235012872
    },
    {
      "epoch": 0.047968021319120584,
      "grad_norm": 6.027890463744797,
      "learning_rate": 4.9961870903613266e-06,
      "loss": 0.9895943641662598,
      "step": 45,
      "token_acc": 0.7256235827664399
    },
    {
      "epoch": 0.05329780146568954,
      "grad_norm": 5.138993659750872,
      "learning_rate": 4.9934328742287285e-06,
      "loss": 1.0118780136108398,
      "step": 50,
      "token_acc": 0.7169505271695052
    },
    {
      "epoch": 0.05862758161225849,
      "grad_norm": 5.517701738242315,
      "learning_rate": 4.989935734988098e-06,
      "loss": 1.1720707893371582,
      "step": 55,
      "token_acc": 0.6896987191051777
    },
    {
      "epoch": 0.06395736175882745,
      "grad_norm": 6.246846876317293,
      "learning_rate": 4.985696714618781e-06,
      "loss": 1.0461352348327637,
      "step": 60,
      "token_acc": 0.7217303822937626
    },
    {
      "epoch": 0.06928714190539641,
      "grad_norm": 5.1838327118545795,
      "learning_rate": 4.98071707614503e-06,
      "loss": 1.0784454345703125,
      "step": 65,
      "token_acc": 0.7065761258041459
    },
    {
      "epoch": 0.07461692205196535,
      "grad_norm": 5.207457333274206,
      "learning_rate": 4.9749983032596775e-06,
      "loss": 1.0711910247802734,
      "step": 70,
      "token_acc": 0.7140947288006112
    },
    {
      "epoch": 0.07994670219853431,
      "grad_norm": 5.076819409823014,
      "learning_rate": 4.968542099882077e-06,
      "loss": 0.9525423049926758,
      "step": 75,
      "token_acc": 0.7458682120626744
    },
    {
      "epoch": 0.08527648234510327,
      "grad_norm": 5.145333113069056,
      "learning_rate": 4.961350389650409e-06,
      "loss": 1.0395957946777343,
      "step": 80,
      "token_acc": 0.7181391712275215
    },
    {
      "epoch": 0.09060626249167222,
      "grad_norm": 5.708204831944594,
      "learning_rate": 4.953425315348534e-06,
      "loss": 1.015221405029297,
      "step": 85,
      "token_acc": 0.7219031993437244
    },
    {
      "epoch": 0.09593604263824117,
      "grad_norm": 5.356494906382105,
      "learning_rate": 4.944769238267542e-06,
      "loss": 1.065409278869629,
      "step": 90,
      "token_acc": 0.7133068878470971
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 5.561086978061129,
      "learning_rate": 4.935384737502204e-06,
      "loss": 0.8904417037963868,
      "step": 95,
      "token_acc": 0.7561363636363636
    },
    {
      "epoch": 0.10659560293137908,
      "grad_norm": 5.007929386757231,
      "learning_rate": 4.925274609182527e-06,
      "loss": 0.9963092803955078,
      "step": 100,
      "token_acc": 0.7157873859579532
    },
    {
      "epoch": 0.11192538307794804,
      "grad_norm": 4.729475065378666,
      "learning_rate": 4.914441865640636e-06,
      "loss": 1.0125768661499024,
      "step": 105,
      "token_acc": 0.7199248120300752
    },
    {
      "epoch": 0.11725516322451698,
      "grad_norm": 5.352763340267538,
      "learning_rate": 4.902889734513251e-06,
      "loss": 0.977642822265625,
      "step": 110,
      "token_acc": 0.7219261890665087
    },
    {
      "epoch": 0.12258494337108594,
      "grad_norm": 5.01951337624599,
      "learning_rate": 4.890621657780002e-06,
      "loss": 0.9350076675415039,
      "step": 115,
      "token_acc": 0.7369072164948454
    },
    {
      "epoch": 0.1279147235176549,
      "grad_norm": 5.682732846054211,
      "learning_rate": 4.8776412907378845e-06,
      "loss": 0.9863363265991211,
      "step": 120,
      "token_acc": 0.7295660948536832
    },
    {
      "epoch": 0.13324450366422386,
      "grad_norm": 5.009947806421316,
      "learning_rate": 4.86395250091216e-06,
      "loss": 0.9168167114257812,
      "step": 125,
      "token_acc": 0.7459949409780776
    },
    {
      "epoch": 0.13857428381079281,
      "grad_norm": 5.16090001119418,
      "learning_rate": 4.849559366904016e-06,
      "loss": 0.9731523513793945,
      "step": 130,
      "token_acc": 0.7264408099688473
    },
    {
      "epoch": 0.14390406395736177,
      "grad_norm": 5.393302939264074,
      "learning_rate": 4.83446617717534e-06,
      "loss": 0.9265727996826172,
      "step": 135,
      "token_acc": 0.7382675211908207
    },
    {
      "epoch": 0.1492338441039307,
      "grad_norm": 5.1348529508995915,
      "learning_rate": 4.818677428770965e-06,
      "loss": 0.9645535469055175,
      "step": 140,
      "token_acc": 0.7365953109072375
    },
    {
      "epoch": 0.15456362425049966,
      "grad_norm": 5.274944058378017,
      "learning_rate": 4.802197825978766e-06,
      "loss": 0.9219735145568848,
      "step": 145,
      "token_acc": 0.7424147311152961
    },
    {
      "epoch": 0.15989340439706862,
      "grad_norm": 4.9265099733335225,
      "learning_rate": 4.785032278928009e-06,
      "loss": 0.9755061149597168,
      "step": 150,
      "token_acc": 0.7261881478584001
    },
    {
      "epoch": 0.16522318454363757,
      "grad_norm": 5.186158859969355,
      "learning_rate": 4.7671859021263635e-06,
      "loss": 0.8955172538757324,
      "step": 155,
      "token_acc": 0.7485970819304153
    },
    {
      "epoch": 0.17055296469020653,
      "grad_norm": 4.8553098725317,
      "learning_rate": 4.748664012936035e-06,
      "loss": 0.9992611885070801,
      "step": 160,
      "token_acc": 0.7210042432814711
    },
    {
      "epoch": 0.1758827448367755,
      "grad_norm": 5.607452928309844,
      "learning_rate": 4.729472129989439e-06,
      "loss": 0.8785256385803223,
      "step": 165,
      "token_acc": 0.7405417024935511
    },
    {
      "epoch": 0.18121252498334445,
      "grad_norm": 5.063309155216159,
      "learning_rate": 4.709615971544912e-06,
      "loss": 0.958376121520996,
      "step": 170,
      "token_acc": 0.7236842105263158
    },
    {
      "epoch": 0.18654230512991338,
      "grad_norm": 4.958090896706627,
      "learning_rate": 4.689101453782954e-06,
      "loss": 0.9511463165283203,
      "step": 175,
      "token_acc": 0.7302402347331745
    },
    {
      "epoch": 0.19187208527648233,
      "grad_norm": 5.554535034316316,
      "learning_rate": 4.6679346890434864e-06,
      "loss": 0.9487571716308594,
      "step": 180,
      "token_acc": 0.7357374546842206
    },
    {
      "epoch": 0.1972018654230513,
      "grad_norm": 5.0455487989596985,
      "learning_rate": 4.646121984004666e-06,
      "loss": 0.9357378005981445,
      "step": 185,
      "token_acc": 0.7466236645837533
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 5.102368295977931,
      "learning_rate": 4.623669837803803e-06,
      "loss": 0.9165685653686524,
      "step": 190,
      "token_acc": 0.7344439983942191
    },
    {
      "epoch": 0.2078614257161892,
      "grad_norm": 6.03242586164789,
      "learning_rate": 4.600584940100929e-06,
      "loss": 0.8400386810302735,
      "step": 195,
      "token_acc": 0.7588924387646432
    },
    {
      "epoch": 0.21319120586275817,
      "grad_norm": 6.046048176929514,
      "learning_rate": 4.576874169085596e-06,
      "loss": 1.0382144927978516,
      "step": 200,
      "token_acc": 0.7164823773324119
    },
    {
      "epoch": 0.21852098600932712,
      "grad_norm": 5.3128952327016306,
      "learning_rate": 4.552544589427514e-06,
      "loss": 0.9587564468383789,
      "step": 205,
      "token_acc": 0.7335751713018944
    },
    {
      "epoch": 0.22385076615589608,
      "grad_norm": 4.856245265355434,
      "learning_rate": 4.527603450171614e-06,
      "loss": 1.0106466293334961,
      "step": 210,
      "token_acc": 0.7285714285714285
    },
    {
      "epoch": 0.229180546302465,
      "grad_norm": 4.364039021197141,
      "learning_rate": 4.502058182578181e-06,
      "loss": 0.8945587158203125,
      "step": 215,
      "token_acc": 0.7414500683994528
    },
    {
      "epoch": 0.23451032644903397,
      "grad_norm": 5.794013669019053,
      "learning_rate": 4.4759163979086985e-06,
      "loss": 0.9250093460083008,
      "step": 220,
      "token_acc": 0.7438412408759124
    },
    {
      "epoch": 0.23984010659560293,
      "grad_norm": 4.9248548258612574,
      "learning_rate": 4.449185885158056e-06,
      "loss": 0.9222723007202148,
      "step": 225,
      "token_acc": 0.7404455344369507
    },
    {
      "epoch": 0.24516988674217188,
      "grad_norm": 5.088119921434173,
      "learning_rate": 4.421874608733806e-06,
      "loss": 0.927369213104248,
      "step": 230,
      "token_acc": 0.7411545623836127
    },
    {
      "epoch": 0.2504996668887408,
      "grad_norm": 5.261396967134003,
      "learning_rate": 4.393990706083149e-06,
      "loss": 0.9975498199462891,
      "step": 235,
      "token_acc": 0.7302452316076294
    },
    {
      "epoch": 0.2558294470353098,
      "grad_norm": 4.270065742658553,
      "learning_rate": 4.3655424852683695e-06,
      "loss": 0.9474044799804687,
      "step": 240,
      "token_acc": 0.7366733466933868
    },
    {
      "epoch": 0.26115922718187873,
      "grad_norm": 4.0368542213252745,
      "learning_rate": 4.336538422491432e-06,
      "loss": 0.9614317893981934,
      "step": 245,
      "token_acc": 0.7291335803776249
    },
    {
      "epoch": 0.2664890073284477,
      "grad_norm": 5.940590055777755,
      "learning_rate": 4.3069871595684795e-06,
      "loss": 0.9075183868408203,
      "step": 250,
      "token_acc": 0.7441300421432872
    },
    {
      "epoch": 0.27181878747501664,
      "grad_norm": 6.020932164981831,
      "learning_rate": 4.276897501354987e-06,
      "loss": 0.9187180519104003,
      "step": 255,
      "token_acc": 0.7450124688279302
    },
    {
      "epoch": 0.27714856762158563,
      "grad_norm": 5.007173976721824,
      "learning_rate": 4.246278413122344e-06,
      "loss": 0.9306771278381347,
      "step": 260,
      "token_acc": 0.7377112881290072
    },
    {
      "epoch": 0.28247834776815456,
      "grad_norm": 5.037382296205618,
      "learning_rate": 4.21513901788662e-06,
      "loss": 0.9223703384399414,
      "step": 265,
      "token_acc": 0.7390941348361512
    },
    {
      "epoch": 0.28780812791472354,
      "grad_norm": 4.579320525829714,
      "learning_rate": 4.183488593690365e-06,
      "loss": 0.8956701278686523,
      "step": 270,
      "token_acc": 0.7352545629202689
    },
    {
      "epoch": 0.2931379080612925,
      "grad_norm": 4.422357549054009,
      "learning_rate": 4.151336570838182e-06,
      "loss": 0.9767134666442872,
      "step": 275,
      "token_acc": 0.729864575908767
    },
    {
      "epoch": 0.2984676882078614,
      "grad_norm": 5.387070334048114,
      "learning_rate": 4.118692529086956e-06,
      "loss": 0.8453518867492675,
      "step": 280,
      "token_acc": 0.7639063153492263
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 5.218152944895714,
      "learning_rate": 4.08556619479155e-06,
      "loss": 0.9285115242004395,
      "step": 285,
      "token_acc": 0.7405669997960432
    },
    {
      "epoch": 0.3091272485009993,
      "grad_norm": 4.473635157614191,
      "learning_rate": 4.051967438006812e-06,
      "loss": 0.9505875587463379,
      "step": 290,
      "token_acc": 0.7344601962922573
    },
    {
      "epoch": 0.3144570286475683,
      "grad_norm": 4.630841152213125,
      "learning_rate": 4.017906269546778e-06,
      "loss": 0.9147586822509766,
      "step": 295,
      "token_acc": 0.7461659457333858
    },
    {
      "epoch": 0.31978680879413723,
      "grad_norm": 4.4381854725202805,
      "learning_rate": 3.983392838001932e-06,
      "loss": 0.8980813980102539,
      "step": 300,
      "token_acc": 0.7419038272816487
    },
    {
      "epoch": 0.3251165889407062,
      "grad_norm": 4.3925972764182735,
      "learning_rate": 3.948437426715409e-06,
      "loss": 0.9489370346069336,
      "step": 305,
      "token_acc": 0.7271723122238586
    },
    {
      "epoch": 0.33044636908727515,
      "grad_norm": 4.777819792705776,
      "learning_rate": 3.913050450719057e-06,
      "loss": 0.9575244903564453,
      "step": 310,
      "token_acc": 0.7353364519611371
    },
    {
      "epoch": 0.3357761492338441,
      "grad_norm": 4.690307974879278,
      "learning_rate": 3.8772424536302565e-06,
      "loss": 0.8590595245361328,
      "step": 315,
      "token_acc": 0.7531313131313131
    },
    {
      "epoch": 0.34110592938041306,
      "grad_norm": 5.548729734061873,
      "learning_rate": 3.841024104510431e-06,
      "loss": 0.78990797996521,
      "step": 320,
      "token_acc": 0.7674789341835573
    },
    {
      "epoch": 0.346435709526982,
      "grad_norm": 4.989871534387572,
      "learning_rate": 3.8044061946861795e-06,
      "loss": 0.9506110191345215,
      "step": 325,
      "token_acc": 0.7373412059402398
    },
    {
      "epoch": 0.351765489673551,
      "grad_norm": 4.676259945487118,
      "learning_rate": 3.767399634533976e-06,
      "loss": 0.9495532035827636,
      "step": 330,
      "token_acc": 0.7358214416392242
    },
    {
      "epoch": 0.3570952698201199,
      "grad_norm": 4.980936255741834,
      "learning_rate": 3.7300154502294093e-06,
      "loss": 0.9292638778686524,
      "step": 335,
      "token_acc": 0.7407978415879746
    },
    {
      "epoch": 0.3624250499666889,
      "grad_norm": 4.965729215770431,
      "learning_rate": 3.6922647804619073e-06,
      "loss": 0.9855228424072265,
      "step": 340,
      "token_acc": 0.7225995961079493
    },
    {
      "epoch": 0.3677548301132578,
      "grad_norm": 4.936075061330924,
      "learning_rate": 3.654158873115944e-06,
      "loss": 0.9076862335205078,
      "step": 345,
      "token_acc": 0.7436141822340832
    },
    {
      "epoch": 0.37308461025982675,
      "grad_norm": 5.231662158108306,
      "learning_rate": 3.61570908191971e-06,
      "loss": 0.8928030014038086,
      "step": 350,
      "token_acc": 0.7541051756391602
    },
    {
      "epoch": 0.37841439040639574,
      "grad_norm": 4.522556964501456,
      "learning_rate": 3.5769268630622533e-06,
      "loss": 0.9126054763793945,
      "step": 355,
      "token_acc": 0.7474881516587678
    },
    {
      "epoch": 0.38374417055296467,
      "grad_norm": 5.1995597965754685,
      "learning_rate": 3.537823771780077e-06,
      "loss": 0.8765750885009765,
      "step": 360,
      "token_acc": 0.7457895779671092
    },
    {
      "epoch": 0.38907395069953365,
      "grad_norm": 5.281242833316695,
      "learning_rate": 3.4984114589142388e-06,
      "loss": 0.9568130493164062,
      "step": 365,
      "token_acc": 0.737781383086595
    },
    {
      "epoch": 0.3944037308461026,
      "grad_norm": 4.5694389026717435,
      "learning_rate": 3.458701667438965e-06,
      "loss": 0.8665159225463868,
      "step": 370,
      "token_acc": 0.7548999797938978
    },
    {
      "epoch": 0.39973351099267157,
      "grad_norm": 4.332384453495391,
      "learning_rate": 3.4187062289627983e-06,
      "loss": 0.882652759552002,
      "step": 375,
      "token_acc": 0.7516012810248198
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 5.463888132443461,
      "learning_rate": 3.3784370602033572e-06,
      "loss": 0.7812636852264404,
      "step": 380,
      "token_acc": 0.769391062746006
    },
    {
      "epoch": 0.4103930712858095,
      "grad_norm": 4.6396298492906505,
      "learning_rate": 3.337906159436721e-06,
      "loss": 0.891010856628418,
      "step": 385,
      "token_acc": 0.752018908804412
    },
    {
      "epoch": 0.4157228514323784,
      "grad_norm": 5.11189175613832,
      "learning_rate": 3.2971256029225253e-06,
      "loss": 0.8738397598266602,
      "step": 390,
      "token_acc": 0.7532467532467533
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 4.9436874254737075,
      "learning_rate": 3.2561075413058113e-06,
      "loss": 0.8521583557128907,
      "step": 395,
      "token_acc": 0.7503179313268334
    },
    {
      "epoch": 0.42638241172551633,
      "grad_norm": 4.666360407842893,
      "learning_rate": 3.214864195996723e-06,
      "loss": 0.9074520111083985,
      "step": 400,
      "token_acc": 0.7434575733544806
    },
    {
      "epoch": 0.43171219187208526,
      "grad_norm": 4.750112035885002,
      "learning_rate": 3.1734078555291137e-06,
      "loss": 0.9123194694519043,
      "step": 405,
      "token_acc": 0.7403057419835943
    },
    {
      "epoch": 0.43704197201865425,
      "grad_norm": 4.771140489181497,
      "learning_rate": 3.1317508718991496e-06,
      "loss": 0.8552797317504883,
      "step": 410,
      "token_acc": 0.7574031890660592
    },
    {
      "epoch": 0.4423717521652232,
      "grad_norm": 4.692545746773149,
      "learning_rate": 3.0899056568850148e-06,
      "loss": 0.9358634948730469,
      "step": 415,
      "token_acc": 0.7455585052072697
    },
    {
      "epoch": 0.44770153231179216,
      "grad_norm": 5.0036012753115235,
      "learning_rate": 3.047884678348795e-06,
      "loss": 0.9225390434265137,
      "step": 420,
      "token_acc": 0.739247795573151
    },
    {
      "epoch": 0.4530313124583611,
      "grad_norm": 4.741320422434045,
      "learning_rate": 3.0057004565216523e-06,
      "loss": 0.8901847839355469,
      "step": 425,
      "token_acc": 0.7495299770210988
    },
    {
      "epoch": 0.45836109260493,
      "grad_norm": 4.703918003843587,
      "learning_rate": 2.9633655602734035e-06,
      "loss": 0.9226030349731446,
      "step": 430,
      "token_acc": 0.7433257055682685
    },
    {
      "epoch": 0.463690872751499,
      "grad_norm": 5.477224800396244,
      "learning_rate": 2.920892603367596e-06,
      "loss": 0.8314413070678711,
      "step": 435,
      "token_acc": 0.7559628975265018
    },
    {
      "epoch": 0.46902065289806794,
      "grad_norm": 4.8329386010913264,
      "learning_rate": 2.8782942407032177e-06,
      "loss": 0.8740703582763671,
      "step": 440,
      "token_acc": 0.7501027538018906
    },
    {
      "epoch": 0.4743504330446369,
      "grad_norm": 5.087948549401884,
      "learning_rate": 2.835583164544139e-06,
      "loss": 0.8511223793029785,
      "step": 445,
      "token_acc": 0.7644836272040302
    },
    {
      "epoch": 0.47968021319120585,
      "grad_norm": 4.99727487177206,
      "learning_rate": 2.7927721007374347e-06,
      "loss": 0.815243911743164,
      "step": 450,
      "token_acc": 0.766592647421663
    },
    {
      "epoch": 0.48500999333777484,
      "grad_norm": 5.576288110570715,
      "learning_rate": 2.7498738049216904e-06,
      "loss": 0.9731996536254883,
      "step": 455,
      "token_acc": 0.7280547752808989
    },
    {
      "epoch": 0.49033977348434377,
      "grad_norm": 4.652800298925416,
      "learning_rate": 2.706901058726437e-06,
      "loss": 0.8669706344604492,
      "step": 460,
      "token_acc": 0.7513238289205703
    },
    {
      "epoch": 0.49566955363091275,
      "grad_norm": 4.67527794119894,
      "learning_rate": 2.6638666659638372e-06,
      "loss": 0.8978883743286132,
      "step": 465,
      "token_acc": 0.7504095004095004
    },
    {
      "epoch": 0.5009993337774816,
      "grad_norm": 4.500058654431643,
      "learning_rate": 2.620783448813768e-06,
      "loss": 0.803309440612793,
      "step": 470,
      "token_acc": 0.7682926829268293
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 5.549718692193634,
      "learning_rate": 2.577664244003426e-06,
      "loss": 0.8737884521484375,
      "step": 475,
      "token_acc": 0.7512740101920815
    },
    {
      "epoch": 0.5116588940706196,
      "grad_norm": 4.633775513369992,
      "learning_rate": 2.5345218989825985e-06,
      "loss": 0.8945508003234863,
      "step": 480,
      "token_acc": 0.7538946953263657
    },
    {
      "epoch": 0.5169886742171885,
      "grad_norm": 4.457350960855039,
      "learning_rate": 2.4913692680957472e-06,
      "loss": 0.8398324012756347,
      "step": 485,
      "token_acc": 0.763268931823229
    },
    {
      "epoch": 0.5223184543637575,
      "grad_norm": 4.9918603168538365,
      "learning_rate": 2.4482192087520225e-06,
      "loss": 0.8864654541015625,
      "step": 490,
      "token_acc": 0.7518545994065282
    },
    {
      "epoch": 0.5276482345103265,
      "grad_norm": 4.279486984922963,
      "learning_rate": 2.4050845775943816e-06,
      "loss": 0.951543140411377,
      "step": 495,
      "token_acc": 0.7330591811192562
    },
    {
      "epoch": 0.5329780146568954,
      "grad_norm": 4.9236086215774115,
      "learning_rate": 2.361978226668922e-06,
      "loss": 0.9079221725463867,
      "step": 500,
      "token_acc": 0.7499020759890325
    },
    {
      "epoch": 0.5383077948034644,
      "grad_norm": 5.043401669503223,
      "learning_rate": 2.3e-06,
      "loss": 0.9313335418701172,
      "step": 505,
      "token_acc": 0.7324805339265851
    },
    {
      "epoch": 0.5436375749500333,
      "grad_norm": 4.437011955068131,
      "learning_rate": 2.3e-06,
      "loss": 0.8004268646240235,
      "step": 510,
      "token_acc": 0.769280205655527
    },
    {
      "epoch": 0.5489673550966022,
      "grad_norm": 4.9461921832529905,
      "learning_rate": 2.3e-06,
      "loss": 0.843082332611084,
      "step": 515,
      "token_acc": 0.7563573136274723
    },
    {
      "epoch": 0.5542971352431713,
      "grad_norm": 4.791111901160093,
      "learning_rate": 2.3e-06,
      "loss": 0.8652103424072266,
      "step": 520,
      "token_acc": 0.7534468775344688
    },
    {
      "epoch": 0.5596269153897402,
      "grad_norm": 4.694299929140503,
      "learning_rate": 2.3e-06,
      "loss": 0.9284069061279296,
      "step": 525,
      "token_acc": 0.7399468556244464
    },
    {
      "epoch": 0.5649566955363091,
      "grad_norm": 4.590405768531116,
      "learning_rate": 2.3e-06,
      "loss": 0.8198009490966797,
      "step": 530,
      "token_acc": 0.7667011375387798
    },
    {
      "epoch": 0.570286475682878,
      "grad_norm": 5.139293375181794,
      "learning_rate": 2.3e-06,
      "loss": 0.9579063415527344,
      "step": 535,
      "token_acc": 0.7357558652319633
    },
    {
      "epoch": 0.5756162558294471,
      "grad_norm": 4.634359261863655,
      "learning_rate": 2.3e-06,
      "loss": 0.8792430877685546,
      "step": 540,
      "token_acc": 0.7548282171173003
    },
    {
      "epoch": 0.580946035976016,
      "grad_norm": 4.447120583644974,
      "learning_rate": 2.3e-06,
      "loss": 0.8046426773071289,
      "step": 545,
      "token_acc": 0.7633396107972379
    },
    {
      "epoch": 0.586275816122585,
      "grad_norm": 4.8496735934946855,
      "learning_rate": 2.3e-06,
      "loss": 0.8646127700805664,
      "step": 550,
      "token_acc": 0.7555690562980963
    },
    {
      "epoch": 0.5916055962691539,
      "grad_norm": 5.354316180442001,
      "learning_rate": 2.3e-06,
      "loss": 0.7959212303161621,
      "step": 555,
      "token_acc": 0.774977497749775
    },
    {
      "epoch": 0.5969353764157228,
      "grad_norm": 4.713142485275984,
      "learning_rate": 2.3e-06,
      "loss": 0.8876336097717286,
      "step": 560,
      "token_acc": 0.7479958890030832
    },
    {
      "epoch": 0.6022651565622918,
      "grad_norm": 4.497831071746374,
      "learning_rate": 2.3e-06,
      "loss": 0.7969367980957032,
      "step": 565,
      "token_acc": 0.7759058389106854
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 5.074530731396222,
      "learning_rate": 2.3e-06,
      "loss": 0.8135330200195312,
      "step": 570,
      "token_acc": 0.7633975852573607
    },
    {
      "epoch": 0.6129247168554297,
      "grad_norm": 5.439750380693716,
      "learning_rate": 2.3e-06,
      "loss": 0.8008811950683594,
      "step": 575,
      "token_acc": 0.7757837837837838
    },
    {
      "epoch": 0.6182544970019986,
      "grad_norm": 4.828635110227898,
      "learning_rate": 2.3e-06,
      "loss": 0.8349386215209961,
      "step": 580,
      "token_acc": 0.7681311751103637
    },
    {
      "epoch": 0.6235842771485676,
      "grad_norm": 4.665942163731842,
      "learning_rate": 2.3e-06,
      "loss": 0.8878559112548828,
      "step": 585,
      "token_acc": 0.7523170972194834
    },
    {
      "epoch": 0.6289140572951366,
      "grad_norm": 4.972690493936675,
      "learning_rate": 2.3e-06,
      "loss": 0.8758079528808593,
      "step": 590,
      "token_acc": 0.7523239000206569
    },
    {
      "epoch": 0.6342438374417055,
      "grad_norm": 5.350060088758041,
      "learning_rate": 2.3e-06,
      "loss": 0.8282616615295411,
      "step": 595,
      "token_acc": 0.7597730138713745
    },
    {
      "epoch": 0.6395736175882745,
      "grad_norm": 4.675665480589392,
      "learning_rate": 2.3e-06,
      "loss": 0.818998146057129,
      "step": 600,
      "token_acc": 0.7656360632523012
    },
    {
      "epoch": 0.6449033977348434,
      "grad_norm": 4.447533140940443,
      "learning_rate": 2.3e-06,
      "loss": 0.8747160911560059,
      "step": 605,
      "token_acc": 0.7465790914066777
    },
    {
      "epoch": 0.6502331778814124,
      "grad_norm": 5.130789428783895,
      "learning_rate": 2.3e-06,
      "loss": 0.8581184387207031,
      "step": 610,
      "token_acc": 0.7551923076923077
    },
    {
      "epoch": 0.6555629580279814,
      "grad_norm": 5.216571173791328,
      "learning_rate": 2.3e-06,
      "loss": 0.8216348648071289,
      "step": 615,
      "token_acc": 0.7595803896381931
    },
    {
      "epoch": 0.6608927381745503,
      "grad_norm": 4.541713854990473,
      "learning_rate": 2.3e-06,
      "loss": 0.8405349731445313,
      "step": 620,
      "token_acc": 0.7559499798305769
    },
    {
      "epoch": 0.6662225183211192,
      "grad_norm": 5.23891265693261,
      "learning_rate": 2.3e-06,
      "loss": 0.9148238182067872,
      "step": 625,
      "token_acc": 0.7366863905325444
    },
    {
      "epoch": 0.6715522984676882,
      "grad_norm": 5.162967835973132,
      "learning_rate": 2.3e-06,
      "loss": 0.9344933509826661,
      "step": 630,
      "token_acc": 0.7420375865479722
    },
    {
      "epoch": 0.6768820786142572,
      "grad_norm": 4.883300468904484,
      "learning_rate": 2.3e-06,
      "loss": 0.8580543518066406,
      "step": 635,
      "token_acc": 0.7618649383232281
    },
    {
      "epoch": 0.6822118587608261,
      "grad_norm": 5.135683049076609,
      "learning_rate": 2.3e-06,
      "loss": 0.7567629814147949,
      "step": 640,
      "token_acc": 0.7874684415882488
    },
    {
      "epoch": 0.6875416389073951,
      "grad_norm": 4.69251131886331,
      "learning_rate": 2.3e-06,
      "loss": 0.8729320526123047,
      "step": 645,
      "token_acc": 0.7532595811932042
    },
    {
      "epoch": 0.692871419053964,
      "grad_norm": 4.537917652494429,
      "learning_rate": 2.3e-06,
      "loss": 0.7952337265014648,
      "step": 650,
      "token_acc": 0.7685081637217624
    },
    {
      "epoch": 0.698201199200533,
      "grad_norm": 4.589779554748045,
      "learning_rate": 2.3e-06,
      "loss": 0.8533817291259765,
      "step": 655,
      "token_acc": 0.7492307692307693
    },
    {
      "epoch": 0.703530979347102,
      "grad_norm": 5.241045198399743,
      "learning_rate": 2.3e-06,
      "loss": 0.8691547393798829,
      "step": 660,
      "token_acc": 0.7476708074534162
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 4.4294360348908475,
      "learning_rate": 2.3e-06,
      "loss": 0.8178191184997559,
      "step": 665,
      "token_acc": 0.7612659819744289
    },
    {
      "epoch": 0.7141905396402398,
      "grad_norm": 5.0125004541858695,
      "learning_rate": 2.3e-06,
      "loss": 0.8978594779968262,
      "step": 670,
      "token_acc": 0.7502776749352091
    },
    {
      "epoch": 0.7195203197868087,
      "grad_norm": 4.89456621393514,
      "learning_rate": 2.3e-06,
      "loss": 0.8966169357299805,
      "step": 675,
      "token_acc": 0.7424985291233576
    },
    {
      "epoch": 0.7248500999333778,
      "grad_norm": 4.723699518324189,
      "learning_rate": 2.3e-06,
      "loss": 0.8785926818847656,
      "step": 680,
      "token_acc": 0.7525147086733726
    },
    {
      "epoch": 0.7301798800799467,
      "grad_norm": 4.749141256391105,
      "learning_rate": 2.3e-06,
      "loss": 0.8269707679748535,
      "step": 685,
      "token_acc": 0.7698035160289556
    },
    {
      "epoch": 0.7355096602265156,
      "grad_norm": 4.357558582129155,
      "learning_rate": 2.3e-06,
      "loss": 0.7781161308288574,
      "step": 690,
      "token_acc": 0.7590879759087976
    },
    {
      "epoch": 0.7408394403730846,
      "grad_norm": 4.715021183358473,
      "learning_rate": 2.3e-06,
      "loss": 0.8154525756835938,
      "step": 695,
      "token_acc": 0.7670259114881908
    },
    {
      "epoch": 0.7461692205196535,
      "grad_norm": 4.845394913724092,
      "learning_rate": 2.3e-06,
      "loss": 0.9302475929260254,
      "step": 700,
      "token_acc": 0.7333573876961934
    },
    {
      "epoch": 0.7514990006662225,
      "grad_norm": 4.458085929920139,
      "learning_rate": 2.3e-06,
      "loss": 0.8405391693115234,
      "step": 705,
      "token_acc": 0.7585416666666667
    },
    {
      "epoch": 0.7568287808127915,
      "grad_norm": 5.104919361304955,
      "learning_rate": 2.3e-06,
      "loss": 0.8558987617492676,
      "step": 710,
      "token_acc": 0.7531428571428571
    },
    {
      "epoch": 0.7621585609593604,
      "grad_norm": 4.50702112491344,
      "learning_rate": 2.3e-06,
      "loss": 0.8234317779541016,
      "step": 715,
      "token_acc": 0.7599072105161415
    },
    {
      "epoch": 0.7674883411059293,
      "grad_norm": 4.613405541301498,
      "learning_rate": 2.3e-06,
      "loss": 0.9495395660400391,
      "step": 720,
      "token_acc": 0.7383750638732755
    },
    {
      "epoch": 0.7728181212524984,
      "grad_norm": 5.0274346973418345,
      "learning_rate": 2.3e-06,
      "loss": 0.8480627059936523,
      "step": 725,
      "token_acc": 0.7663193706145014
    },
    {
      "epoch": 0.7781479013990673,
      "grad_norm": 4.868595479556196,
      "learning_rate": 2.3e-06,
      "loss": 0.8753684997558594,
      "step": 730,
      "token_acc": 0.74125
    },
    {
      "epoch": 0.7834776815456362,
      "grad_norm": 4.564850676362406,
      "learning_rate": 2.3e-06,
      "loss": 0.843782615661621,
      "step": 735,
      "token_acc": 0.765643924101736
    },
    {
      "epoch": 0.7888074616922052,
      "grad_norm": 5.05324907708148,
      "learning_rate": 2.3e-06,
      "loss": 0.8115653991699219,
      "step": 740,
      "token_acc": 0.7754152823920266
    },
    {
      "epoch": 0.7941372418387741,
      "grad_norm": 5.480865823369537,
      "learning_rate": 2.3e-06,
      "loss": 0.8137409210205078,
      "step": 745,
      "token_acc": 0.7597292724196277
    },
    {
      "epoch": 0.7994670219853431,
      "grad_norm": 5.369323266878849,
      "learning_rate": 2.3e-06,
      "loss": 0.8155117034912109,
      "step": 750,
      "token_acc": 0.7665171288743883
    },
    {
      "epoch": 0.8047968021319121,
      "grad_norm": 4.853665943115109,
      "learning_rate": 2.3e-06,
      "loss": 0.8802165985107422,
      "step": 755,
      "token_acc": 0.7480860371855632
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 4.808224943607436,
      "learning_rate": 2.3e-06,
      "loss": 0.82237548828125,
      "step": 760,
      "token_acc": 0.764473957252542
    },
    {
      "epoch": 0.8154563624250499,
      "grad_norm": 4.622557273028215,
      "learning_rate": 2.3e-06,
      "loss": 0.8893803596496582,
      "step": 765,
      "token_acc": 0.7448350851757883
    },
    {
      "epoch": 0.820786142571619,
      "grad_norm": 4.769533656387771,
      "learning_rate": 2.3e-06,
      "loss": 0.8182163238525391,
      "step": 770,
      "token_acc": 0.7690664387951293
    },
    {
      "epoch": 0.8261159227181879,
      "grad_norm": 4.781990293335522,
      "learning_rate": 2.3e-06,
      "loss": 0.8312844276428223,
      "step": 775,
      "token_acc": 0.761876127480457
    },
    {
      "epoch": 0.8314457028647568,
      "grad_norm": 4.988959771776016,
      "learning_rate": 2.3e-06,
      "loss": 0.814521312713623,
      "step": 780,
      "token_acc": 0.7755408010280574
    },
    {
      "epoch": 0.8367754830113258,
      "grad_norm": 4.087547272283349,
      "learning_rate": 2.3e-06,
      "loss": 0.8847762107849121,
      "step": 785,
      "token_acc": 0.7430968192939532
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 4.599551127627647,
      "learning_rate": 2.3e-06,
      "loss": 0.6787426471710205,
      "step": 790,
      "token_acc": 0.7971876690102758
    },
    {
      "epoch": 0.8474350433044637,
      "grad_norm": 4.776471715784441,
      "learning_rate": 2.3e-06,
      "loss": 0.8404502868652344,
      "step": 795,
      "token_acc": 0.7615946319321097
    },
    {
      "epoch": 0.8527648234510327,
      "grad_norm": 4.77324194182284,
      "learning_rate": 2.3e-06,
      "loss": 0.8979682922363281,
      "step": 800,
      "token_acc": 0.7428789906327662
    },
    {
      "epoch": 0.8580946035976016,
      "grad_norm": 4.186841899763308,
      "learning_rate": 2.3e-06,
      "loss": 0.8490333557128906,
      "step": 805,
      "token_acc": 0.7513409961685824
    },
    {
      "epoch": 0.8634243837441705,
      "grad_norm": 5.00996163860295,
      "learning_rate": 2.3e-06,
      "loss": 0.8072927474975586,
      "step": 810,
      "token_acc": 0.7662141779788839
    },
    {
      "epoch": 0.8687541638907396,
      "grad_norm": 4.818424223620723,
      "learning_rate": 2.3e-06,
      "loss": 0.7880621910095215,
      "step": 815,
      "token_acc": 0.7660944206008584
    },
    {
      "epoch": 0.8740839440373085,
      "grad_norm": 4.6356713282207975,
      "learning_rate": 2.3e-06,
      "loss": 0.8073809623718262,
      "step": 820,
      "token_acc": 0.7698314837648993
    },
    {
      "epoch": 0.8794137241838774,
      "grad_norm": 5.03662108299368,
      "learning_rate": 2.3e-06,
      "loss": 0.7883618354797364,
      "step": 825,
      "token_acc": 0.7801530153015301
    },
    {
      "epoch": 0.8847435043304464,
      "grad_norm": 4.494987433874659,
      "learning_rate": 2.3e-06,
      "loss": 0.8354556083679199,
      "step": 830,
      "token_acc": 0.7542094058447842
    },
    {
      "epoch": 0.8900732844770153,
      "grad_norm": 4.571643128763884,
      "learning_rate": 2.3e-06,
      "loss": 0.7755649089813232,
      "step": 835,
      "token_acc": 0.7809482943574341
    },
    {
      "epoch": 0.8954030646235843,
      "grad_norm": 4.515732991917952,
      "learning_rate": 2.3e-06,
      "loss": 0.8827999114990235,
      "step": 840,
      "token_acc": 0.7483820356932732
    },
    {
      "epoch": 0.9007328447701533,
      "grad_norm": 4.965366676519157,
      "learning_rate": 2.3e-06,
      "loss": 0.8320845603942871,
      "step": 845,
      "token_acc": 0.7568418980114195
    },
    {
      "epoch": 0.9060626249167222,
      "grad_norm": 5.153213542890202,
      "learning_rate": 2.3e-06,
      "loss": 0.8549650192260743,
      "step": 850,
      "token_acc": 0.7493156042236997
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 4.685348919973664,
      "learning_rate": 2.3e-06,
      "loss": 0.8040557861328125,
      "step": 855,
      "token_acc": 0.7747043918918919
    },
    {
      "epoch": 0.91672218520986,
      "grad_norm": 5.631387602760864,
      "learning_rate": 2.3e-06,
      "loss": 0.7609379768371582,
      "step": 860,
      "token_acc": 0.7761161681837885
    },
    {
      "epoch": 0.9220519653564291,
      "grad_norm": 4.29732725735809,
      "learning_rate": 2.3e-06,
      "loss": 0.8553963661193847,
      "step": 865,
      "token_acc": 0.7378821774794929
    },
    {
      "epoch": 0.927381745502998,
      "grad_norm": 4.4689622483360365,
      "learning_rate": 2.3e-06,
      "loss": 0.8411100387573243,
      "step": 870,
      "token_acc": 0.758266584514274
    },
    {
      "epoch": 0.9327115256495669,
      "grad_norm": 4.271516958949675,
      "learning_rate": 2.3e-06,
      "loss": 0.8458847045898438,
      "step": 875,
      "token_acc": 0.7636141138078727
    },
    {
      "epoch": 0.9380413057961359,
      "grad_norm": 5.280065922587605,
      "learning_rate": 2.3e-06,
      "loss": 0.7837121963500977,
      "step": 880,
      "token_acc": 0.7722984562607204
    },
    {
      "epoch": 0.9433710859427049,
      "grad_norm": 4.490709801842713,
      "learning_rate": 2.3e-06,
      "loss": 0.9280232429504395,
      "step": 885,
      "token_acc": 0.7421804375338998
    },
    {
      "epoch": 0.9487008660892738,
      "grad_norm": 5.2839081660696054,
      "learning_rate": 2.3e-06,
      "loss": 0.9062253952026367,
      "step": 890,
      "token_acc": 0.7496486649267216
    },
    {
      "epoch": 0.9540306462358428,
      "grad_norm": 4.659524061077068,
      "learning_rate": 2.3e-06,
      "loss": 0.7646889686584473,
      "step": 895,
      "token_acc": 0.7716500553709856
    },
    {
      "epoch": 0.9593604263824117,
      "grad_norm": 4.658005621045716,
      "learning_rate": 2.3e-06,
      "loss": 0.7846517562866211,
      "step": 900,
      "token_acc": 0.7760330578512397
    },
    {
      "epoch": 0.9646902065289806,
      "grad_norm": 4.827339570974992,
      "learning_rate": 2.3e-06,
      "loss": 0.8194332122802734,
      "step": 905,
      "token_acc": 0.7632450331125827
    },
    {
      "epoch": 0.9700199866755497,
      "grad_norm": 4.616508722821306,
      "learning_rate": 2.3e-06,
      "loss": 0.8886819839477539,
      "step": 910,
      "token_acc": 0.7549358843883575
    },
    {
      "epoch": 0.9753497668221186,
      "grad_norm": 4.587147736616801,
      "learning_rate": 2.3e-06,
      "loss": 0.82857666015625,
      "step": 915,
      "token_acc": 0.7579642579642579
    },
    {
      "epoch": 0.9806795469686875,
      "grad_norm": 4.385315976477473,
      "learning_rate": 2.3e-06,
      "loss": 0.8632021903991699,
      "step": 920,
      "token_acc": 0.7554684275691291
    },
    {
      "epoch": 0.9860093271152565,
      "grad_norm": 5.3105288038094605,
      "learning_rate": 2.3e-06,
      "loss": 0.7477989196777344,
      "step": 925,
      "token_acc": 0.7832802157391517
    },
    {
      "epoch": 0.9913391072618255,
      "grad_norm": 4.244619714135519,
      "learning_rate": 2.3e-06,
      "loss": 0.8497745513916015,
      "step": 930,
      "token_acc": 0.7574374885928089
    },
    {
      "epoch": 0.9966688874083944,
      "grad_norm": 4.693628875486247,
      "learning_rate": 2.3e-06,
      "loss": 0.7646096229553223,
      "step": 935,
      "token_acc": 0.7749592739120317
    }
  ],
  "logging_steps": 5,
  "max_steps": 939,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 1231477221752832.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}