{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.39222519170640535,
  "eval_steps": 1000000,
  "global_step": 44909,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0043668885046027,
      "grad_norm": 1.5595725774765015,
      "learning_rate": 9.99956331114954e-06,
      "loss": 3.1525,
      "step": 500
    },
    {
      "epoch": 0.0087337770092054,
      "grad_norm": 1.5938678979873657,
      "learning_rate": 9.99912662229908e-06,
      "loss": 3.1559,
      "step": 1000
    },
    {
      "epoch": 0.013100665513808101,
      "grad_norm": 1.6471061706542969,
      "learning_rate": 9.99868993344862e-06,
      "loss": 3.1529,
      "step": 1500
    },
    {
      "epoch": 0.0174675540184108,
      "grad_norm": 1.5785421133041382,
      "learning_rate": 9.99825324459816e-06,
      "loss": 3.1604,
      "step": 2000
    },
    {
      "epoch": 0.021834442523013503,
      "grad_norm": 1.598698377609253,
      "learning_rate": 9.9978165557477e-06,
      "loss": 3.1565,
      "step": 2500
    },
    {
      "epoch": 0.026201331027616202,
      "grad_norm": 1.5545878410339355,
      "learning_rate": 9.997379866897238e-06,
      "loss": 3.1531,
      "step": 3000
    },
    {
      "epoch": 0.030568219532218905,
      "grad_norm": 1.5569171905517578,
      "learning_rate": 9.996943178046778e-06,
      "loss": 3.1572,
      "step": 3500
    },
    {
      "epoch": 0.0349351080368216,
      "grad_norm": 1.569543719291687,
      "learning_rate": 9.996506489196318e-06,
      "loss": 3.1594,
      "step": 4000
    },
    {
      "epoch": 0.03930199654142431,
      "grad_norm": 1.6514673233032227,
      "learning_rate": 9.996069800345858e-06,
      "loss": 3.1553,
      "step": 4500
    },
    {
      "epoch": 0.043668885046027006,
      "grad_norm": 1.604501485824585,
      "learning_rate": 9.995633111495399e-06,
      "loss": 3.1547,
      "step": 5000
    },
    {
      "epoch": 0.048035773550629705,
      "grad_norm": 1.6219558715820312,
      "learning_rate": 9.995196422644939e-06,
      "loss": 3.1534,
      "step": 5500
    },
    {
      "epoch": 0.052402662055232405,
      "grad_norm": 1.606451392173767,
      "learning_rate": 9.994759733794479e-06,
      "loss": 3.1522,
      "step": 6000
    },
    {
      "epoch": 0.056769550559835104,
      "grad_norm": 1.6143338680267334,
      "learning_rate": 9.994323044944017e-06,
      "loss": 3.1445,
      "step": 6500
    },
    {
      "epoch": 0.06113643906443781,
      "grad_norm": 1.6684553623199463,
      "learning_rate": 9.993886356093557e-06,
      "loss": 3.1551,
      "step": 7000
    },
    {
      "epoch": 0.06550332756904051,
      "grad_norm": 1.6013703346252441,
      "learning_rate": 9.993449667243097e-06,
      "loss": 3.1498,
      "step": 7500
    },
    {
      "epoch": 0.0698702160736432,
      "grad_norm": 1.626556396484375,
      "learning_rate": 9.993012978392636e-06,
      "loss": 3.1509,
      "step": 8000
    },
    {
      "epoch": 0.07423710457824591,
      "grad_norm": 1.659162998199463,
      "learning_rate": 9.992576289542176e-06,
      "loss": 3.153,
      "step": 8500
    },
    {
      "epoch": 0.07860399308284861,
      "grad_norm": 1.6168526411056519,
      "learning_rate": 9.992139600691716e-06,
      "loss": 3.1429,
      "step": 9000
    },
    {
      "epoch": 0.0829708815874513,
      "grad_norm": 1.6810740232467651,
      "learning_rate": 9.991702911841256e-06,
      "loss": 3.1438,
      "step": 9500
    },
    {
      "epoch": 0.08733777009205401,
      "grad_norm": 1.7077081203460693,
      "learning_rate": 9.991266222990796e-06,
      "loss": 3.1474,
      "step": 10000
    },
    {
      "epoch": 0.0917046585966567,
      "grad_norm": 1.5795377492904663,
      "learning_rate": 9.990829534140335e-06,
      "loss": 3.1564,
      "step": 10500
    },
    {
      "epoch": 0.09607154710125941,
      "grad_norm": 1.6760386228561401,
      "learning_rate": 9.990392845289875e-06,
      "loss": 3.1534,
      "step": 11000
    },
    {
      "epoch": 0.10043843560586212,
      "grad_norm": 1.5983505249023438,
      "learning_rate": 9.989956156439415e-06,
      "loss": 3.1464,
      "step": 11500
    },
    {
      "epoch": 0.10480532411046481,
      "grad_norm": 1.6111019849777222,
      "learning_rate": 9.989519467588953e-06,
      "loss": 3.1333,
      "step": 12000
    },
    {
      "epoch": 0.10917221261506752,
      "grad_norm": 1.5885874032974243,
      "learning_rate": 9.989082778738493e-06,
      "loss": 3.1461,
      "step": 12500
    },
    {
      "epoch": 0.11353910111967021,
      "grad_norm": 1.6448547840118408,
      "learning_rate": 9.988646089888033e-06,
      "loss": 3.1459,
      "step": 13000
    },
    {
      "epoch": 0.11790598962427291,
      "grad_norm": 1.6023292541503906,
      "learning_rate": 9.988209401037573e-06,
      "loss": 3.1467,
      "step": 13500
    },
    {
      "epoch": 0.12227287812887562,
      "grad_norm": 1.5931400060653687,
      "learning_rate": 9.987772712187114e-06,
      "loss": 3.1388,
      "step": 14000
    },
    {
      "epoch": 0.1266397666334783,
      "grad_norm": 1.625355839729309,
      "learning_rate": 9.987336023336654e-06,
      "loss": 3.1473,
      "step": 14500
    },
    {
      "epoch": 0.13100665513808102,
      "grad_norm": 1.6307705640792847,
      "learning_rate": 9.986899334486194e-06,
      "loss": 3.1424,
      "step": 15000
    },
    {
      "epoch": 0.13537354364268372,
      "grad_norm": 1.6311821937561035,
      "learning_rate": 9.986462645635732e-06,
      "loss": 3.1477,
      "step": 15500
    },
    {
      "epoch": 0.1397404321472864,
      "grad_norm": 1.6085231304168701,
      "learning_rate": 9.986025956785272e-06,
      "loss": 3.1358,
      "step": 16000
    },
    {
      "epoch": 0.1441073206518891,
      "grad_norm": 1.6187548637390137,
      "learning_rate": 9.985589267934812e-06,
      "loss": 3.1459,
      "step": 16500
    },
    {
      "epoch": 0.14847420915649182,
      "grad_norm": 1.5847587585449219,
      "learning_rate": 9.985152579084352e-06,
      "loss": 3.1359,
      "step": 17000
    },
    {
      "epoch": 0.15284109766109452,
      "grad_norm": 1.6051414012908936,
      "learning_rate": 9.984715890233891e-06,
      "loss": 3.1368,
      "step": 17500
    },
    {
      "epoch": 0.15720798616569723,
      "grad_norm": 1.6769737005233765,
      "learning_rate": 9.984279201383431e-06,
      "loss": 3.1366,
      "step": 18000
    },
    {
      "epoch": 0.1615748746702999,
      "grad_norm": 1.586376428604126,
      "learning_rate": 9.983842512532971e-06,
      "loss": 3.1454,
      "step": 18500
    },
    {
      "epoch": 0.1659417631749026,
      "grad_norm": 1.6034488677978516,
      "learning_rate": 9.98340582368251e-06,
      "loss": 3.1395,
      "step": 19000
    },
    {
      "epoch": 0.17030865167950532,
      "grad_norm": 1.5914573669433594,
      "learning_rate": 9.98296913483205e-06,
      "loss": 3.1394,
      "step": 19500
    },
    {
      "epoch": 0.17467554018410802,
      "grad_norm": 1.6094170808792114,
      "learning_rate": 9.98253244598159e-06,
      "loss": 3.1452,
      "step": 20000
    },
    {
      "epoch": 0.17904242868871073,
      "grad_norm": 1.5861003398895264,
      "learning_rate": 9.982095757131129e-06,
      "loss": 3.1337,
      "step": 20500
    },
    {
      "epoch": 0.1834093171933134,
      "grad_norm": 1.5987565517425537,
      "learning_rate": 9.981659068280668e-06,
      "loss": 3.1437,
      "step": 21000
    },
    {
      "epoch": 0.18777620569791612,
      "grad_norm": 1.5903559923171997,
      "learning_rate": 9.981222379430208e-06,
      "loss": 3.1364,
      "step": 21500
    },
    {
      "epoch": 0.19214309420251882,
      "grad_norm": 1.5559889078140259,
      "learning_rate": 9.980785690579748e-06,
      "loss": 3.1412,
      "step": 22000
    },
    {
      "epoch": 0.19650998270712153,
      "grad_norm": 1.6113758087158203,
      "learning_rate": 9.980349001729288e-06,
      "loss": 3.1364,
      "step": 22500
    },
    {
      "epoch": 0.20087687121172423,
      "grad_norm": 1.6110438108444214,
      "learning_rate": 9.97991231287883e-06,
      "loss": 3.1404,
      "step": 23000
    },
    {
      "epoch": 0.2052437597163269,
      "grad_norm": 1.6391644477844238,
      "learning_rate": 9.97947562402837e-06,
      "loss": 3.1398,
      "step": 23500
    },
    {
      "epoch": 0.20961064822092962,
      "grad_norm": 1.5945662260055542,
      "learning_rate": 9.979038935177907e-06,
      "loss": 3.1351,
      "step": 24000
    },
    {
      "epoch": 0.21397753672553232,
      "grad_norm": 1.7338697910308838,
      "learning_rate": 9.978602246327447e-06,
      "loss": 3.1294,
      "step": 24500
    },
    {
      "epoch": 0.21834442523013503,
      "grad_norm": 1.574798345565796,
      "learning_rate": 9.978165557476987e-06,
      "loss": 3.1328,
      "step": 25000
    },
    {
      "epoch": 0.22271131373473774,
      "grad_norm": 1.5656156539916992,
      "learning_rate": 9.977728868626527e-06,
      "loss": 3.1307,
      "step": 25500
    },
    {
      "epoch": 0.22707820223934042,
      "grad_norm": 1.6264046430587769,
      "learning_rate": 9.977292179776067e-06,
      "loss": 3.1382,
      "step": 26000
    },
    {
      "epoch": 0.23144509074394312,
      "grad_norm": 1.591504454612732,
      "learning_rate": 9.976855490925607e-06,
      "loss": 3.1317,
      "step": 26500
    },
    {
      "epoch": 0.23581197924854583,
      "grad_norm": 1.7037614583969116,
      "learning_rate": 9.976418802075146e-06,
      "loss": 3.1284,
      "step": 27000
    },
    {
      "epoch": 0.24017886775314853,
      "grad_norm": 1.6053491830825806,
      "learning_rate": 9.975982113224686e-06,
      "loss": 3.1391,
      "step": 27500
    },
    {
      "epoch": 0.24454575625775124,
      "grad_norm": 1.574787974357605,
      "learning_rate": 9.975545424374226e-06,
      "loss": 3.133,
      "step": 28000
    },
    {
      "epoch": 0.24891264476235392,
      "grad_norm": 1.5841468572616577,
      "learning_rate": 9.975108735523766e-06,
      "loss": 3.1298,
      "step": 28500
    },
    {
      "epoch": 0.2532795332669566,
      "grad_norm": 1.5428208112716675,
      "learning_rate": 9.974672046673306e-06,
      "loss": 3.1313,
      "step": 29000
    },
    {
      "epoch": 0.25764642177155933,
      "grad_norm": 1.6238336563110352,
      "learning_rate": 9.974235357822844e-06,
      "loss": 3.1199,
      "step": 29500
    },
    {
      "epoch": 0.26201331027616204,
      "grad_norm": 1.6074714660644531,
      "learning_rate": 9.973798668972384e-06,
      "loss": 3.1322,
      "step": 30000
    },
    {
      "epoch": 0.26638019878076474,
      "grad_norm": 1.583072304725647,
      "learning_rate": 9.973361980121923e-06,
      "loss": 3.133,
      "step": 30500
    },
    {
      "epoch": 0.27074708728536745,
      "grad_norm": 1.6968551874160767,
      "learning_rate": 9.972925291271463e-06,
      "loss": 3.1281,
      "step": 31000
    },
    {
      "epoch": 0.27511397578997016,
      "grad_norm": 1.6063077449798584,
      "learning_rate": 9.972488602421003e-06,
      "loss": 3.1247,
      "step": 31500
    },
    {
      "epoch": 0.2794808642945728,
      "grad_norm": 1.5899564027786255,
      "learning_rate": 9.972051913570545e-06,
      "loss": 3.1215,
      "step": 32000
    },
    {
      "epoch": 0.2838477527991755,
      "grad_norm": 1.5838756561279297,
      "learning_rate": 9.971615224720084e-06,
      "loss": 3.1262,
      "step": 32500
    },
    {
      "epoch": 0.2882146413037782,
      "grad_norm": 1.5518513917922974,
      "learning_rate": 9.971178535869623e-06,
      "loss": 3.1181,
      "step": 33000
    },
    {
      "epoch": 0.2925815298083809,
      "grad_norm": 1.5394465923309326,
      "learning_rate": 9.970741847019162e-06,
      "loss": 3.1205,
      "step": 33500
    },
    {
      "epoch": 0.29694841831298363,
      "grad_norm": 1.6238704919815063,
      "learning_rate": 9.970305158168702e-06,
      "loss": 3.1205,
      "step": 34000
    },
    {
      "epoch": 0.30131530681758634,
      "grad_norm": 1.6042754650115967,
      "learning_rate": 9.969868469318242e-06,
      "loss": 3.1283,
      "step": 34500
    },
    {
      "epoch": 0.30568219532218904,
      "grad_norm": 1.6376352310180664,
      "learning_rate": 9.969431780467782e-06,
      "loss": 3.1242,
      "step": 35000
    },
    {
      "epoch": 0.31004908382679175,
      "grad_norm": 1.6078994274139404,
      "learning_rate": 9.968995091617322e-06,
      "loss": 3.1188,
      "step": 35500
    },
    {
      "epoch": 0.31441597233139446,
      "grad_norm": 1.6135457754135132,
      "learning_rate": 9.968558402766862e-06,
      "loss": 3.1187,
      "step": 36000
    },
    {
      "epoch": 0.31878286083599716,
      "grad_norm": 1.6336065530776978,
      "learning_rate": 9.968121713916401e-06,
      "loss": 3.1183,
      "step": 36500
    },
    {
      "epoch": 0.3231497493405998,
      "grad_norm": 1.6096062660217285,
      "learning_rate": 9.967685025065941e-06,
      "loss": 3.1183,
      "step": 37000
    },
    {
      "epoch": 0.3275166378452025,
      "grad_norm": 1.570556879043579,
      "learning_rate": 9.967248336215481e-06,
      "loss": 3.122,
      "step": 37500
    },
    {
      "epoch": 0.3318835263498052,
      "grad_norm": 1.5272282361984253,
      "learning_rate": 9.96681164736502e-06,
      "loss": 3.1193,
      "step": 38000
    },
    {
      "epoch": 0.33625041485440793,
      "grad_norm": 1.5485848188400269,
      "learning_rate": 9.966374958514559e-06,
      "loss": 3.1222,
      "step": 38500
    },
    {
      "epoch": 0.34061730335901064,
      "grad_norm": 1.5924766063690186,
      "learning_rate": 9.965938269664099e-06,
      "loss": 3.1183,
      "step": 39000
    },
    {
      "epoch": 0.34498419186361334,
      "grad_norm": 1.601727843284607,
      "learning_rate": 9.965501580813639e-06,
      "loss": 3.1194,
      "step": 39500
    },
    {
      "epoch": 0.34935108036821605,
      "grad_norm": 1.6034713983535767,
      "learning_rate": 9.965064891963178e-06,
      "loss": 3.1146,
      "step": 40000
    },
    {
      "epoch": 0.35371796887281876,
      "grad_norm": 1.6022893190383911,
      "learning_rate": 9.964628203112718e-06,
      "loss": 3.116,
      "step": 40500
    },
    {
      "epoch": 0.35808485737742146,
      "grad_norm": 1.5923199653625488,
      "learning_rate": 9.96419151426226e-06,
      "loss": 3.1191,
      "step": 41000
    },
    {
      "epoch": 0.36245174588202417,
      "grad_norm": 1.5369057655334473,
      "learning_rate": 9.9637548254118e-06,
      "loss": 3.1153,
      "step": 41500
    },
    {
      "epoch": 0.3668186343866268,
      "grad_norm": 1.574439525604248,
      "learning_rate": 9.963318136561338e-06,
      "loss": 3.1164,
      "step": 42000
    },
    {
      "epoch": 0.3711855228912295,
      "grad_norm": 1.5916707515716553,
      "learning_rate": 9.962881447710878e-06,
      "loss": 3.1115,
      "step": 42500
    },
    {
      "epoch": 0.37555241139583223,
      "grad_norm": 1.551416277885437,
      "learning_rate": 9.962444758860417e-06,
      "loss": 3.1154,
      "step": 43000
    },
    {
      "epoch": 0.37991929990043494,
      "grad_norm": 1.6488794088363647,
      "learning_rate": 9.962008070009957e-06,
      "loss": 3.1175,
      "step": 43500
    },
    {
      "epoch": 0.38428618840503764,
      "grad_norm": 1.5746173858642578,
      "learning_rate": 9.961571381159497e-06,
      "loss": 3.1112,
      "step": 44000
    },
    {
      "epoch": 0.38865307690964035,
      "grad_norm": 1.5648127794265747,
      "learning_rate": 9.961134692309037e-06,
      "loss": 3.1055,
      "step": 44500
    },
    {
      "epoch": 0.39222519170640535,
      "step": 44909,
      "total_flos": 1.1264989248576553e+18,
      "train_loss": 3.1353410507916006,
      "train_runtime": 46794.5499,
      "train_samples_per_second": 11744.671,
      "train_steps_per_second": 244.682
    }
  ],
  "logging_steps": 500,
  "max_steps": 11449800,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 1000000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1264989248576553e+18,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": null
}