{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.077741172308577,
  "eval_steps": 500,
  "global_step": 34800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.017688548875671613,
      "grad_norm": 4.9453349113464355,
      "learning_rate": 1.768972227136034e-06,
      "loss": 6.7403,
      "step": 200
    },
    {
      "epoch": 0.03537709775134323,
      "grad_norm": 8.81293773651123,
      "learning_rate": 3.537944454272068e-06,
      "loss": 4.3531,
      "step": 400
    },
    {
      "epoch": 0.05306564662701484,
      "grad_norm": 5.486693382263184,
      "learning_rate": 5.306916681408102e-06,
      "loss": 2.2861,
      "step": 600
    },
    {
      "epoch": 0.07075419550268645,
      "grad_norm": 2.244082450866699,
      "learning_rate": 7.075888908544136e-06,
      "loss": 1.162,
      "step": 800
    },
    {
      "epoch": 0.08844274437835806,
      "grad_norm": 0.6459134221076965,
      "learning_rate": 8.84486113568017e-06,
      "loss": 0.3574,
      "step": 1000
    },
    {
      "epoch": 0.10613129325402967,
      "grad_norm": 0.42250216007232666,
      "learning_rate": 1.0613833362816204e-05,
      "loss": 0.22,
      "step": 1200
    },
    {
      "epoch": 0.12381984212970129,
      "grad_norm": 0.4047611951828003,
      "learning_rate": 1.2382805589952239e-05,
      "loss": 0.1967,
      "step": 1400
    },
    {
      "epoch": 0.1415083910053729,
      "grad_norm": 0.34184110164642334,
      "learning_rate": 1.4151777817088272e-05,
      "loss": 0.1853,
      "step": 1600
    },
    {
      "epoch": 0.1591969398810445,
      "grad_norm": 0.32166364789009094,
      "learning_rate": 1.5920750044224307e-05,
      "loss": 0.1795,
      "step": 1800
    },
    {
      "epoch": 0.1768854887567161,
      "grad_norm": 0.30309006571769714,
      "learning_rate": 1.768972227136034e-05,
      "loss": 0.1777,
      "step": 2000
    },
    {
      "epoch": 0.19457403763238773,
      "grad_norm": 0.28535833954811096,
      "learning_rate": 1.9458694498496373e-05,
      "loss": 0.1728,
      "step": 2200
    },
    {
      "epoch": 0.21226258650805935,
      "grad_norm": 0.2785949409008026,
      "learning_rate": 2.1227666725632408e-05,
      "loss": 0.1707,
      "step": 2400
    },
    {
      "epoch": 0.22995113538373096,
      "grad_norm": 0.28039732575416565,
      "learning_rate": 2.2996638952768443e-05,
      "loss": 0.1682,
      "step": 2600
    },
    {
      "epoch": 0.24763968425940258,
      "grad_norm": 0.2688646614551544,
      "learning_rate": 2.4765611179904478e-05,
      "loss": 0.1663,
      "step": 2800
    },
    {
      "epoch": 0.2653282331350742,
      "grad_norm": 0.2585384249687195,
      "learning_rate": 2.653458340704051e-05,
      "loss": 0.1672,
      "step": 3000
    },
    {
      "epoch": 0.2830167820107458,
      "grad_norm": 0.2583986222743988,
      "learning_rate": 2.8303555634176544e-05,
      "loss": 0.1646,
      "step": 3200
    },
    {
      "epoch": 0.30070533088641743,
      "grad_norm": 0.23997661471366882,
      "learning_rate": 3.007252786131258e-05,
      "loss": 0.1647,
      "step": 3400
    },
    {
      "epoch": 0.318393879762089,
      "grad_norm": 0.2527500092983246,
      "learning_rate": 3.1841500088448614e-05,
      "loss": 0.1629,
      "step": 3600
    },
    {
      "epoch": 0.3360824286377606,
      "grad_norm": 0.23295536637306213,
      "learning_rate": 3.3610472315584645e-05,
      "loss": 0.1615,
      "step": 3800
    },
    {
      "epoch": 0.3537709775134322,
      "grad_norm": 0.24165655672550201,
      "learning_rate": 3.537944454272068e-05,
      "loss": 0.1626,
      "step": 4000
    },
    {
      "epoch": 0.37145952638910384,
      "grad_norm": 0.22286227345466614,
      "learning_rate": 3.7148416769856715e-05,
      "loss": 0.1611,
      "step": 4200
    },
    {
      "epoch": 0.38914807526477546,
      "grad_norm": 0.23912444710731506,
      "learning_rate": 3.8917388996992746e-05,
      "loss": 0.1639,
      "step": 4400
    },
    {
      "epoch": 0.4068366241404471,
      "grad_norm": 0.22330383956432343,
      "learning_rate": 4.0686361224128784e-05,
      "loss": 0.1613,
      "step": 4600
    },
    {
      "epoch": 0.4245251730161187,
      "grad_norm": 0.2139132171869278,
      "learning_rate": 4.2455333451264816e-05,
      "loss": 0.1601,
      "step": 4800
    },
    {
      "epoch": 0.4422137218917903,
      "grad_norm": 0.19250735640525818,
      "learning_rate": 4.4224305678400854e-05,
      "loss": 0.1595,
      "step": 5000
    },
    {
      "epoch": 0.45990227076746193,
      "grad_norm": 0.19502046704292297,
      "learning_rate": 4.5993277905536885e-05,
      "loss": 0.159,
      "step": 5200
    },
    {
      "epoch": 0.47759081964313355,
      "grad_norm": 0.19040720164775848,
      "learning_rate": 4.776225013267292e-05,
      "loss": 0.1578,
      "step": 5400
    },
    {
      "epoch": 0.49527936851880516,
      "grad_norm": 0.20574906468391418,
      "learning_rate": 4.9531222359808955e-05,
      "loss": 0.1603,
      "step": 5600
    },
    {
      "epoch": 0.5129679173944768,
      "grad_norm": 0.19954811036586761,
      "learning_rate": 4.999976891136569e-05,
      "loss": 0.1582,
      "step": 5800
    },
    {
      "epoch": 0.5306564662701484,
      "grad_norm": 0.18808187544345856,
      "learning_rate": 4.999871234414489e-05,
      "loss": 0.1589,
      "step": 6000
    },
    {
      "epoch": 0.54834501514582,
      "grad_norm": 0.17781756818294525,
      "learning_rate": 4.999680029474971e-05,
      "loss": 0.158,
      "step": 6200
    },
    {
      "epoch": 0.5660335640214916,
      "grad_norm": 0.18824966251850128,
      "learning_rate": 4.999403282861248e-05,
      "loss": 0.1572,
      "step": 6400
    },
    {
      "epoch": 0.5837221128971632,
      "grad_norm": 0.20633359253406525,
      "learning_rate": 4.999041004043882e-05,
      "loss": 0.1572,
      "step": 6600
    },
    {
      "epoch": 0.6014106617728349,
      "grad_norm": 0.17536261677742004,
      "learning_rate": 4.998593205420432e-05,
      "loss": 0.1565,
      "step": 6800
    },
    {
      "epoch": 0.6190992106485064,
      "grad_norm": 0.1713794469833374,
      "learning_rate": 4.998059902315038e-05,
      "loss": 0.1554,
      "step": 7000
    },
    {
      "epoch": 0.636787759524178,
      "grad_norm": 0.16785737872123718,
      "learning_rate": 4.997441112977891e-05,
      "loss": 0.1526,
      "step": 7200
    },
    {
      "epoch": 0.6544763083998496,
      "grad_norm": 0.18514026701450348,
      "learning_rate": 4.996736858584613e-05,
      "loss": 0.153,
      "step": 7400
    },
    {
      "epoch": 0.6721648572755212,
      "grad_norm": 0.17102603614330292,
      "learning_rate": 4.995947163235527e-05,
      "loss": 0.1517,
      "step": 7600
    },
    {
      "epoch": 0.6898534061511928,
      "grad_norm": 0.16847601532936096,
      "learning_rate": 4.9950720539548384e-05,
      "loss": 0.1519,
      "step": 7800
    },
    {
      "epoch": 0.7075419550268645,
      "grad_norm": 0.16419926285743713,
      "learning_rate": 4.9941115606897036e-05,
      "loss": 0.1507,
      "step": 8000
    },
    {
      "epoch": 0.7252305039025361,
      "grad_norm": 0.17619839310646057,
      "learning_rate": 4.9930657163092123e-05,
      "loss": 0.15,
      "step": 8200
    },
    {
      "epoch": 0.7429190527782077,
      "grad_norm": 0.16663217544555664,
      "learning_rate": 4.991934556603254e-05,
      "loss": 0.1502,
      "step": 8400
    },
    {
      "epoch": 0.7606076016538793,
      "grad_norm": 0.18116818368434906,
      "learning_rate": 4.990718120281304e-05,
      "loss": 0.1511,
      "step": 8600
    },
    {
      "epoch": 0.7782961505295509,
      "grad_norm": 0.1595318764448166,
      "learning_rate": 4.989416448971088e-05,
      "loss": 0.1502,
      "step": 8800
    },
    {
      "epoch": 0.7959846994052225,
      "grad_norm": 0.15985798835754395,
      "learning_rate": 4.988029587217165e-05,
      "loss": 0.1513,
      "step": 9000
    },
    {
      "epoch": 0.8136732482808942,
      "grad_norm": 0.17328430712223053,
      "learning_rate": 4.9865575824794e-05,
      "loss": 0.1491,
      "step": 9200
    },
    {
      "epoch": 0.8313617971565658,
      "grad_norm": 0.1536557823419571,
      "learning_rate": 4.98500048513134e-05,
      "loss": 0.1474,
      "step": 9400
    },
    {
      "epoch": 0.8490503460322374,
      "grad_norm": 0.15362966060638428,
      "learning_rate": 4.983358348458491e-05,
      "loss": 0.1458,
      "step": 9600
    },
    {
      "epoch": 0.866738894907909,
      "grad_norm": 0.16235561668872833,
      "learning_rate": 4.9816312286564926e-05,
      "loss": 0.1454,
      "step": 9800
    },
    {
      "epoch": 0.8844274437835806,
      "grad_norm": 0.6078733801841736,
      "learning_rate": 4.979819184829197e-05,
      "loss": 0.148,
      "step": 10000
    },
    {
      "epoch": 0.9021159926592522,
      "grad_norm": 0.15071770548820496,
      "learning_rate": 4.9779222789866476e-05,
      "loss": 0.1461,
      "step": 10200
    },
    {
      "epoch": 0.9198045415349239,
      "grad_norm": 0.15512414276599884,
      "learning_rate": 4.9759405760429524e-05,
      "loss": 0.1455,
      "step": 10400
    },
    {
      "epoch": 0.9374930904105955,
      "grad_norm": 0.15411154925823212,
      "learning_rate": 4.9738741438140644e-05,
      "loss": 0.1437,
      "step": 10600
    },
    {
      "epoch": 0.9551816392862671,
      "grad_norm": 0.15488554537296295,
      "learning_rate": 4.9717230530154657e-05,
      "loss": 0.1445,
      "step": 10800
    },
    {
      "epoch": 0.9728701881619387,
      "grad_norm": 0.14975008368492126,
      "learning_rate": 4.9694873772597396e-05,
      "loss": 0.1437,
      "step": 11000
    },
    {
      "epoch": 0.9905587370376103,
      "grad_norm": 0.15676584839820862,
      "learning_rate": 4.967167193054058e-05,
      "loss": 0.1432,
      "step": 11200
    },
    {
      "epoch": 1.0082251752271874,
      "grad_norm": 0.15392336249351501,
      "learning_rate": 4.964762579797558e-05,
      "loss": 0.1355,
      "step": 11400
    },
    {
      "epoch": 1.0259137241028589,
      "grad_norm": 0.1635395586490631,
      "learning_rate": 4.962273619778632e-05,
      "loss": 0.1268,
      "step": 11600
    },
    {
      "epoch": 1.0436022729785306,
      "grad_norm": 0.15909574925899506,
      "learning_rate": 4.959700398172101e-05,
      "loss": 0.1263,
      "step": 11800
    },
    {
      "epoch": 1.061290821854202,
      "grad_norm": 0.16021914780139923,
      "learning_rate": 4.957043003036311e-05,
      "loss": 0.1264,
      "step": 12000
    },
    {
      "epoch": 1.0789793707298738,
      "grad_norm": 0.163675919175148,
      "learning_rate": 4.954301525310113e-05,
      "loss": 0.1262,
      "step": 12200
    },
    {
      "epoch": 1.0966679196055453,
      "grad_norm": 0.15520550310611725,
      "learning_rate": 4.951476058809751e-05,
      "loss": 0.1275,
      "step": 12400
    },
    {
      "epoch": 1.114356468481217,
      "grad_norm": 0.1602706015110016,
      "learning_rate": 4.948566700225654e-05,
      "loss": 0.1268,
      "step": 12600
    },
    {
      "epoch": 1.1320450173568886,
      "grad_norm": 0.15892288088798523,
      "learning_rate": 4.945573549119128e-05,
      "loss": 0.1269,
      "step": 12800
    },
    {
      "epoch": 1.1497335662325603,
      "grad_norm": 0.16024959087371826,
      "learning_rate": 4.9424967079189434e-05,
      "loss": 0.1265,
      "step": 13000
    },
    {
      "epoch": 1.1674221151082318,
      "grad_norm": 0.15324904024600983,
      "learning_rate": 4.939336281917837e-05,
      "loss": 0.1265,
      "step": 13200
    },
    {
      "epoch": 1.1851106639839033,
      "grad_norm": 0.15968984365463257,
      "learning_rate": 4.936092379268902e-05,
      "loss": 0.1269,
      "step": 13400
    },
    {
      "epoch": 1.202799212859575,
      "grad_norm": 0.1491909772157669,
      "learning_rate": 4.932765110981894e-05,
      "loss": 0.1261,
      "step": 13600
    },
    {
      "epoch": 1.2204877617352468,
      "grad_norm": 0.15451796352863312,
      "learning_rate": 4.929354590919424e-05,
      "loss": 0.1273,
      "step": 13800
    },
    {
      "epoch": 1.2381763106109183,
      "grad_norm": 0.1513613909482956,
      "learning_rate": 4.9258609357930686e-05,
      "loss": 0.1264,
      "step": 14000
    },
    {
      "epoch": 1.2558648594865898,
      "grad_norm": 0.1397712677717209,
      "learning_rate": 4.9222842651593736e-05,
      "loss": 0.1268,
      "step": 14200
    },
    {
      "epoch": 1.2735534083622615,
      "grad_norm": 0.13348053395748138,
      "learning_rate": 4.918624701415763e-05,
      "loss": 0.1267,
      "step": 14400
    },
    {
      "epoch": 1.2912419572379332,
      "grad_norm": 0.15651994943618774,
      "learning_rate": 4.9148823697963465e-05,
      "loss": 0.1258,
      "step": 14600
    },
    {
      "epoch": 1.3089305061136047,
      "grad_norm": 0.16271665692329407,
      "learning_rate": 4.9110573983676414e-05,
      "loss": 0.1258,
      "step": 14800
    },
    {
      "epoch": 1.3266190549892762,
      "grad_norm": 0.15141454339027405,
      "learning_rate": 4.907149918024185e-05,
      "loss": 0.1252,
      "step": 15000
    },
    {
      "epoch": 1.344307603864948,
      "grad_norm": 0.14476826786994934,
      "learning_rate": 4.903160062484056e-05,
      "loss": 0.1263,
      "step": 15200
    },
    {
      "epoch": 1.3619961527406195,
      "grad_norm": 0.14604584872722626,
      "learning_rate": 4.8990879682842964e-05,
      "loss": 0.1267,
      "step": 15400
    },
    {
      "epoch": 1.3796847016162912,
      "grad_norm": 0.1551065295934677,
      "learning_rate": 4.8949337747762465e-05,
      "loss": 0.1268,
      "step": 15600
    },
    {
      "epoch": 1.3973732504919627,
      "grad_norm": 0.15430127084255219,
      "learning_rate": 4.890697624120767e-05,
      "loss": 0.1258,
      "step": 15800
    },
    {
      "epoch": 1.4150617993676344,
      "grad_norm": 0.1534918248653412,
      "learning_rate": 4.886379661283379e-05,
      "loss": 0.1245,
      "step": 16000
    },
    {
      "epoch": 1.432750348243306,
      "grad_norm": 0.15945963561534882,
      "learning_rate": 4.881980034029303e-05,
      "loss": 0.1251,
      "step": 16200
    },
    {
      "epoch": 1.4504388971189777,
      "grad_norm": 0.15190242230892181,
      "learning_rate": 4.877498892918403e-05,
      "loss": 0.1246,
      "step": 16400
    },
    {
      "epoch": 1.4681274459946492,
      "grad_norm": 0.14410296082496643,
      "learning_rate": 4.872936391300029e-05,
      "loss": 0.1251,
      "step": 16600
    },
    {
      "epoch": 1.485815994870321,
      "grad_norm": 0.13909801840782166,
      "learning_rate": 4.868292685307776e-05,
      "loss": 0.1255,
      "step": 16800
    },
    {
      "epoch": 1.5035045437459924,
      "grad_norm": 0.14646555483341217,
      "learning_rate": 4.8635679338541364e-05,
      "loss": 0.1243,
      "step": 17000
    },
    {
      "epoch": 1.521193092621664,
      "grad_norm": 0.14244422316551208,
      "learning_rate": 4.858762298625065e-05,
      "loss": 0.1248,
      "step": 17200
    },
    {
      "epoch": 1.5388816414973356,
      "grad_norm": 0.15146219730377197,
      "learning_rate": 4.853875944074442e-05,
      "loss": 0.1235,
      "step": 17400
    },
    {
      "epoch": 1.5565701903730074,
      "grad_norm": 0.14190447330474854,
      "learning_rate": 4.848909037418449e-05,
      "loss": 0.1242,
      "step": 17600
    },
    {
      "epoch": 1.5742587392486789,
      "grad_norm": 0.13961580395698547,
      "learning_rate": 4.8438617486298455e-05,
      "loss": 0.1235,
      "step": 17800
    },
    {
      "epoch": 1.5919472881243504,
      "grad_norm": 0.13622143864631653,
      "learning_rate": 4.838734250432152e-05,
      "loss": 0.1231,
      "step": 18000
    },
    {
      "epoch": 1.609635837000022,
      "grad_norm": 0.140974760055542,
      "learning_rate": 4.833526718293736e-05,
      "loss": 0.1229,
      "step": 18200
    },
    {
      "epoch": 1.6273243858756938,
      "grad_norm": 0.1440068930387497,
      "learning_rate": 4.828239330421815e-05,
      "loss": 0.1233,
      "step": 18400
    },
    {
      "epoch": 1.6450129347513653,
      "grad_norm": 0.13277290761470795,
      "learning_rate": 4.822872267756351e-05,
      "loss": 0.122,
      "step": 18600
    },
    {
      "epoch": 1.6627014836270368,
      "grad_norm": 0.1427055150270462,
      "learning_rate": 4.817425713963861e-05,
      "loss": 0.123,
      "step": 18800
    },
    {
      "epoch": 1.6803900325027086,
      "grad_norm": 0.14716538786888123,
      "learning_rate": 4.8118998554311336e-05,
      "loss": 0.1235,
      "step": 19000
    },
    {
      "epoch": 1.6980785813783803,
      "grad_norm": 0.14058035612106323,
      "learning_rate": 4.806294881258846e-05,
      "loss": 0.1219,
      "step": 19200
    },
    {
      "epoch": 1.7157671302540518,
      "grad_norm": 0.1426689326763153,
      "learning_rate": 4.800610983255098e-05,
      "loss": 0.1223,
      "step": 19400
    },
    {
      "epoch": 1.7334556791297233,
      "grad_norm": 0.142822727560997,
      "learning_rate": 4.7948483559288445e-05,
      "loss": 0.1217,
      "step": 19600
    },
    {
      "epoch": 1.751144228005395,
      "grad_norm": 0.14739733934402466,
      "learning_rate": 4.7890071964832426e-05,
      "loss": 0.122,
      "step": 19800
    },
    {
      "epoch": 1.7688327768810668,
      "grad_norm": 0.1422308385372162,
      "learning_rate": 4.7830877048088974e-05,
      "loss": 0.1209,
      "step": 20000
    },
    {
      "epoch": 1.7865213257567383,
      "grad_norm": 0.13312865793704987,
      "learning_rate": 4.777090083477027e-05,
      "loss": 0.1209,
      "step": 20200
    },
    {
      "epoch": 1.8042098746324098,
      "grad_norm": 0.1410498321056366,
      "learning_rate": 4.771014537732529e-05,
      "loss": 0.1202,
      "step": 20400
    },
    {
      "epoch": 1.8218984235080815,
      "grad_norm": 0.14671219885349274,
      "learning_rate": 4.764861275486956e-05,
      "loss": 0.1195,
      "step": 20600
    },
    {
      "epoch": 1.839586972383753,
      "grad_norm": 0.13390739262104034,
      "learning_rate": 4.758630507311399e-05,
      "loss": 0.1204,
      "step": 20800
    },
    {
      "epoch": 1.8572755212594245,
      "grad_norm": 0.1377139389514923,
      "learning_rate": 4.7523224464292855e-05,
      "loss": 0.1194,
      "step": 21000
    },
    {
      "epoch": 1.8749640701350963,
      "grad_norm": 0.13953597843647003,
      "learning_rate": 4.745937308709079e-05,
      "loss": 0.1196,
      "step": 21200
    },
    {
      "epoch": 1.892652619010768,
      "grad_norm": 0.14722049236297607,
      "learning_rate": 4.739475312656895e-05,
      "loss": 0.1189,
      "step": 21400
    },
    {
      "epoch": 1.9103411678864395,
      "grad_norm": 0.129732146859169,
      "learning_rate": 4.7329366794090205e-05,
      "loss": 0.1195,
      "step": 21600
    },
    {
      "epoch": 1.928029716762111,
      "grad_norm": 0.13127955794334412,
      "learning_rate": 4.726321632724346e-05,
      "loss": 0.1188,
      "step": 21800
    },
    {
      "epoch": 1.9457182656377827,
      "grad_norm": 0.1353120505809784,
      "learning_rate": 4.719630398976714e-05,
      "loss": 0.1184,
      "step": 22000
    },
    {
      "epoch": 1.9634068145134544,
      "grad_norm": 0.14046898484230042,
      "learning_rate": 4.7128632071471667e-05,
      "loss": 0.1185,
      "step": 22200
    },
    {
      "epoch": 1.981095363389126,
      "grad_norm": 0.13207530975341797,
      "learning_rate": 4.7060202888161106e-05,
      "loss": 0.1174,
      "step": 22400
    },
    {
      "epoch": 1.9987839122647975,
      "grad_norm": 0.1329745352268219,
      "learning_rate": 4.6991018781553926e-05,
      "loss": 0.1181,
      "step": 22600
    },
    {
      "epoch": 2.0164503504543747,
      "grad_norm": 0.13534873723983765,
      "learning_rate": 4.692108211920287e-05,
      "loss": 0.097,
      "step": 22800
    },
    {
      "epoch": 2.0341388993300464,
      "grad_norm": 0.13999226689338684,
      "learning_rate": 4.685039529441393e-05,
      "loss": 0.096,
      "step": 23000
    },
    {
      "epoch": 2.0518274482057177,
      "grad_norm": 0.14331400394439697,
      "learning_rate": 4.677896072616444e-05,
      "loss": 0.0956,
      "step": 23200
    },
    {
      "epoch": 2.0695159970813894,
      "grad_norm": 0.1391351968050003,
      "learning_rate": 4.67067808590203e-05,
      "loss": 0.0955,
      "step": 23400
    },
    {
      "epoch": 2.087204545957061,
      "grad_norm": 0.15080305933952332,
      "learning_rate": 4.6633858163052324e-05,
      "loss": 0.0966,
      "step": 23600
    },
    {
      "epoch": 2.1048930948327325,
      "grad_norm": 0.14853306114673615,
      "learning_rate": 4.656019513375171e-05,
      "loss": 0.0955,
      "step": 23800
    },
    {
      "epoch": 2.122581643708404,
      "grad_norm": 0.14308440685272217,
      "learning_rate": 4.648579429194463e-05,
      "loss": 0.0959,
      "step": 24000
    },
    {
      "epoch": 2.140270192584076,
      "grad_norm": 0.14518137276172638,
      "learning_rate": 4.641065818370597e-05,
      "loss": 0.0964,
      "step": 24200
    },
    {
      "epoch": 2.1579587414597476,
      "grad_norm": 0.14385883510112762,
      "learning_rate": 4.6334789380272235e-05,
      "loss": 0.0966,
      "step": 24400
    },
    {
      "epoch": 2.175647290335419,
      "grad_norm": 0.14189130067825317,
      "learning_rate": 4.625819047795349e-05,
      "loss": 0.0969,
      "step": 24600
    },
    {
      "epoch": 2.1933358392110907,
      "grad_norm": 0.14711739122867584,
      "learning_rate": 4.6180864098044584e-05,
      "loss": 0.0967,
      "step": 24800
    },
    {
      "epoch": 2.2110243880867624,
      "grad_norm": 0.14622507989406586,
      "learning_rate": 4.610281288673539e-05,
      "loss": 0.0967,
      "step": 25000
    },
    {
      "epoch": 2.228712936962434,
      "grad_norm": 0.1425817757844925,
      "learning_rate": 4.6024039515020276e-05,
      "loss": 0.0981,
      "step": 25200
    },
    {
      "epoch": 2.2464014858381054,
      "grad_norm": 0.1472136527299881,
      "learning_rate": 4.5944546678606706e-05,
      "loss": 0.0993,
      "step": 25400
    },
    {
      "epoch": 2.264090034713777,
      "grad_norm": 0.14232853055000305,
      "learning_rate": 4.586433709782296e-05,
      "loss": 0.0985,
      "step": 25600
    },
    {
      "epoch": 2.281778583589449,
      "grad_norm": 0.1511968970298767,
      "learning_rate": 4.578341351752511e-05,
      "loss": 0.098,
      "step": 25800
    },
    {
      "epoch": 2.2994671324651206,
      "grad_norm": 0.14743036031723022,
      "learning_rate": 4.570177870700298e-05,
      "loss": 0.0974,
      "step": 26000
    },
    {
      "epoch": 2.317155681340792,
      "grad_norm": 0.1461828500032425,
      "learning_rate": 4.561943545988548e-05,
      "loss": 0.0969,
      "step": 26200
    },
    {
      "epoch": 2.3348442302164636,
      "grad_norm": 0.1614687293767929,
      "learning_rate": 4.5536386594044956e-05,
      "loss": 0.0988,
      "step": 26400
    },
    {
      "epoch": 2.3525327790921353,
      "grad_norm": 0.1400858461856842,
      "learning_rate": 4.5452634951500745e-05,
      "loss": 0.0969,
      "step": 26600
    },
    {
      "epoch": 2.3702213279678066,
      "grad_norm": 0.14338116347789764,
      "learning_rate": 4.536818339832197e-05,
      "loss": 0.0961,
      "step": 26800
    },
    {
      "epoch": 2.3879098768434783,
      "grad_norm": 0.1501627415418625,
      "learning_rate": 4.528303482452943e-05,
      "loss": 0.0969,
      "step": 27000
    },
    {
      "epoch": 2.40559842571915,
      "grad_norm": 0.14346472918987274,
      "learning_rate": 4.519719214399667e-05,
      "loss": 0.0958,
      "step": 27200
    },
    {
      "epoch": 2.423286974594822,
      "grad_norm": 0.1539745032787323,
      "learning_rate": 4.5110658294350326e-05,
      "loss": 0.0966,
      "step": 27400
    },
    {
      "epoch": 2.4409755234704935,
      "grad_norm": 0.1377008557319641,
      "learning_rate": 4.502343623686956e-05,
      "loss": 0.0971,
      "step": 27600
    },
    {
      "epoch": 2.458664072346165,
      "grad_norm": 0.1407323032617569,
      "learning_rate": 4.493552895638472e-05,
      "loss": 0.0974,
      "step": 27800
    },
    {
      "epoch": 2.4763526212218365,
      "grad_norm": 0.1503271609544754,
      "learning_rate": 4.48469394611752e-05,
      "loss": 0.0975,
      "step": 28000
    },
    {
      "epoch": 2.4940411700975083,
      "grad_norm": 0.14528031647205353,
      "learning_rate": 4.475767078286652e-05,
      "loss": 0.0974,
      "step": 28200
    },
    {
      "epoch": 2.5117297189731795,
      "grad_norm": 0.14073877036571503,
      "learning_rate": 4.466772597632654e-05,
      "loss": 0.0963,
      "step": 28400
    },
    {
      "epoch": 2.5294182678488513,
      "grad_norm": 0.14889408648014069,
      "learning_rate": 4.457710811956094e-05,
      "loss": 0.0963,
      "step": 28600
    },
    {
      "epoch": 2.547106816724523,
      "grad_norm": 0.14946310222148895,
      "learning_rate": 4.4485820313607906e-05,
      "loss": 0.0967,
      "step": 28800
    },
    {
      "epoch": 2.5647953656001947,
      "grad_norm": 0.1508057564496994,
      "learning_rate": 4.4393865682431955e-05,
      "loss": 0.0973,
      "step": 29000
    },
    {
      "epoch": 2.5824839144758664,
      "grad_norm": 0.1469687819480896,
      "learning_rate": 4.4301247372817077e-05,
      "loss": 0.0971,
      "step": 29200
    },
    {
      "epoch": 2.6001724633515377,
      "grad_norm": 0.14045564830303192,
      "learning_rate": 4.420796855425905e-05,
      "loss": 0.0959,
      "step": 29400
    },
    {
      "epoch": 2.6178610122272095,
      "grad_norm": 0.14008161425590515,
      "learning_rate": 4.411403241885693e-05,
      "loss": 0.0966,
      "step": 29600
    },
    {
      "epoch": 2.635549561102881,
      "grad_norm": 0.15664730966091156,
      "learning_rate": 4.4019442181203884e-05,
      "loss": 0.0969,
      "step": 29800
    },
    {
      "epoch": 2.6532381099785525,
      "grad_norm": 0.146159827709198,
      "learning_rate": 4.3924201078277105e-05,
      "loss": 0.0969,
      "step": 30000
    },
    {
      "epoch": 2.670926658854224,
      "grad_norm": 0.14852313697338104,
      "learning_rate": 4.382831236932711e-05,
      "loss": 0.0955,
      "step": 30200
    },
    {
      "epoch": 2.688615207729896,
      "grad_norm": 0.15001444518566132,
      "learning_rate": 4.3731779335766154e-05,
      "loss": 0.0959,
      "step": 30400
    },
    {
      "epoch": 2.7063037566055677,
      "grad_norm": 0.14691965281963348,
      "learning_rate": 4.363460528105597e-05,
      "loss": 0.0957,
      "step": 30600
    },
    {
      "epoch": 2.723992305481239,
      "grad_norm": 0.14318469166755676,
      "learning_rate": 4.35367935305947e-05,
      "loss": 0.0967,
      "step": 30800
    },
    {
      "epoch": 2.7416808543569107,
      "grad_norm": 0.14508825540542603,
      "learning_rate": 4.34383474316031e-05,
      "loss": 0.0951,
      "step": 31000
    },
    {
      "epoch": 2.7593694032325824,
      "grad_norm": 0.15151719748973846,
      "learning_rate": 4.333927035301001e-05,
      "loss": 0.0958,
      "step": 31200
    },
    {
      "epoch": 2.7770579521082537,
      "grad_norm": 0.1437998265028,
      "learning_rate": 4.3239565685337044e-05,
      "loss": 0.0955,
      "step": 31400
    },
    {
      "epoch": 2.7947465009839254,
      "grad_norm": 0.14122720062732697,
      "learning_rate": 4.3139236840582575e-05,
      "loss": 0.0951,
      "step": 31600
    },
    {
      "epoch": 2.812435049859597,
      "grad_norm": 0.1338931769132614,
      "learning_rate": 4.303828725210498e-05,
      "loss": 0.0959,
      "step": 31800
    },
    {
      "epoch": 2.830123598735269,
      "grad_norm": 0.144461989402771,
      "learning_rate": 4.293672037450512e-05,
      "loss": 0.0951,
      "step": 32000
    },
    {
      "epoch": 2.8478121476109406,
      "grad_norm": 0.13668565452098846,
      "learning_rate": 4.2834539683508166e-05,
      "loss": 0.0959,
      "step": 32200
    },
    {
      "epoch": 2.865500696486612,
      "grad_norm": 0.15980184078216553,
      "learning_rate": 4.27317486758446e-05,
      "loss": 0.0951,
      "step": 32400
    },
    {
      "epoch": 2.8831892453622836,
      "grad_norm": 0.14222297072410583,
      "learning_rate": 4.262835086913058e-05,
      "loss": 0.0948,
      "step": 32600
    },
    {
      "epoch": 2.9008777942379553,
      "grad_norm": 0.1509125828742981,
      "learning_rate": 4.25243498017476e-05,
      "loss": 0.0949,
      "step": 32800
    },
    {
      "epoch": 2.9185663431136266,
      "grad_norm": 0.14582620561122894,
      "learning_rate": 4.241974903272132e-05,
      "loss": 0.0943,
      "step": 33000
    },
    {
      "epoch": 2.9362548919892983,
      "grad_norm": 0.130904421210289,
      "learning_rate": 4.231455214159985e-05,
      "loss": 0.0945,
      "step": 33200
    },
    {
      "epoch": 2.95394344086497,
      "grad_norm": 0.1373138129711151,
      "learning_rate": 4.220876272833121e-05,
      "loss": 0.0936,
      "step": 33400
    },
    {
      "epoch": 2.971631989740642,
      "grad_norm": 0.14548689126968384,
      "learning_rate": 4.210238441314017e-05,
      "loss": 0.0934,
      "step": 33600
    },
    {
      "epoch": 2.9893205386163135,
      "grad_norm": 0.13954471051692963,
      "learning_rate": 4.199542083640432e-05,
      "loss": 0.0937,
      "step": 33800
    },
    {
      "epoch": 3.0069869768058903,
      "grad_norm": 0.14478568732738495,
      "learning_rate": 4.188787565852952e-05,
      "loss": 0.0847,
      "step": 34000
    },
    {
      "epoch": 3.024675525681562,
      "grad_norm": 0.15503637492656708,
      "learning_rate": 4.177975255982463e-05,
      "loss": 0.0718,
      "step": 34200
    },
    {
      "epoch": 3.0423640745572333,
      "grad_norm": 0.16748104989528656,
      "learning_rate": 4.1671055240375575e-05,
      "loss": 0.0714,
      "step": 34400
    },
    {
      "epoch": 3.060052623432905,
      "grad_norm": 0.15309254825115204,
      "learning_rate": 4.156178741991872e-05,
      "loss": 0.0724,
      "step": 34600
    },
    {
      "epoch": 3.077741172308577,
      "grad_norm": 0.15351246297359467,
      "learning_rate": 4.1451952837713556e-05,
      "loss": 0.0722,
      "step": 34800
    }
  ],
  "logging_steps": 200,
  "max_steps": 113060,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0180563807244386e+20,
  "train_batch_size": 9,
  "trial_name": null,
  "trial_params": null
}