{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4298328355080438,
  "eval_steps": 1000000,
  "global_step": 49215,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0043668885046027,
      "grad_norm": 1.499011516571045,
      "learning_rate": 9.99956331114954e-06,
      "loss": 3.0752,
      "step": 500
    },
    {
      "epoch": 0.0087337770092054,
      "grad_norm": 1.530051350593567,
      "learning_rate": 9.99912662229908e-06,
      "loss": 3.0792,
      "step": 1000
    },
    {
      "epoch": 0.013100665513808101,
      "grad_norm": 1.5715222358703613,
      "learning_rate": 9.99868993344862e-06,
      "loss": 3.0762,
      "step": 1500
    },
    {
      "epoch": 0.0174675540184108,
      "grad_norm": 1.519605040550232,
      "learning_rate": 9.99825324459816e-06,
      "loss": 3.0838,
      "step": 2000
    },
    {
      "epoch": 0.021834442523013503,
      "grad_norm": 1.5175423622131348,
      "learning_rate": 9.9978165557477e-06,
      "loss": 3.0795,
      "step": 2500
    },
    {
      "epoch": 0.026201331027616202,
      "grad_norm": 1.503777027130127,
      "learning_rate": 9.997379866897238e-06,
      "loss": 3.0768,
      "step": 3000
    },
    {
      "epoch": 0.030568219532218905,
      "grad_norm": 1.493525743484497,
      "learning_rate": 9.996943178046778e-06,
      "loss": 3.0818,
      "step": 3500
    },
    {
      "epoch": 0.0349351080368216,
      "grad_norm": 1.5032507181167603,
      "learning_rate": 9.996506489196318e-06,
      "loss": 3.0838,
      "step": 4000
    },
    {
      "epoch": 0.03930199654142431,
      "grad_norm": 1.5484732389450073,
      "learning_rate": 9.996069800345858e-06,
      "loss": 3.0798,
      "step": 4500
    },
    {
      "epoch": 0.043668885046027006,
      "grad_norm": 1.5356857776641846,
      "learning_rate": 9.995633111495399e-06,
      "loss": 3.08,
      "step": 5000
    },
    {
      "epoch": 0.048035773550629705,
      "grad_norm": 1.5570900440216064,
      "learning_rate": 9.995196422644939e-06,
      "loss": 3.0784,
      "step": 5500
    },
    {
      "epoch": 0.052402662055232405,
      "grad_norm": 1.5267707109451294,
      "learning_rate": 9.994759733794479e-06,
      "loss": 3.0775,
      "step": 6000
    },
    {
      "epoch": 0.056769550559835104,
      "grad_norm": 1.5429790019989014,
      "learning_rate": 9.994323044944017e-06,
      "loss": 3.0695,
      "step": 6500
    },
    {
      "epoch": 0.06113643906443781,
      "grad_norm": 1.8057048320770264,
      "learning_rate": 9.993886356093557e-06,
      "loss": 3.0802,
      "step": 7000
    },
    {
      "epoch": 0.06550332756904051,
      "grad_norm": 1.5101138353347778,
      "learning_rate": 9.993449667243097e-06,
      "loss": 3.0753,
      "step": 7500
    },
    {
      "epoch": 0.0698702160736432,
      "grad_norm": 1.5532760620117188,
      "learning_rate": 9.993012978392636e-06,
      "loss": 3.0767,
      "step": 8000
    },
    {
      "epoch": 0.07423710457824591,
      "grad_norm": 1.5693330764770508,
      "learning_rate": 9.992576289542176e-06,
      "loss": 3.0792,
      "step": 8500
    },
    {
      "epoch": 0.07860399308284861,
      "grad_norm": 1.5401012897491455,
      "learning_rate": 9.992139600691716e-06,
      "loss": 3.069,
      "step": 9000
    },
    {
      "epoch": 0.0829708815874513,
      "grad_norm": 1.5863893032073975,
      "learning_rate": 9.991702911841256e-06,
      "loss": 3.0701,
      "step": 9500
    },
    {
      "epoch": 0.08733777009205401,
      "grad_norm": 1.588409185409546,
      "learning_rate": 9.991266222990796e-06,
      "loss": 3.0735,
      "step": 10000
    },
    {
      "epoch": 0.0917046585966567,
      "grad_norm": 1.5264829397201538,
      "learning_rate": 9.990829534140335e-06,
      "loss": 3.0825,
      "step": 10500
    },
    {
      "epoch": 0.09607154710125941,
      "grad_norm": 1.5645649433135986,
      "learning_rate": 9.990392845289875e-06,
      "loss": 3.0802,
      "step": 11000
    },
    {
      "epoch": 0.10043843560586212,
      "grad_norm": 1.536049485206604,
      "learning_rate": 9.989956156439415e-06,
      "loss": 3.0739,
      "step": 11500
    },
    {
      "epoch": 0.10480532411046481,
      "grad_norm": 1.5308794975280762,
      "learning_rate": 9.989519467588953e-06,
      "loss": 3.0606,
      "step": 12000
    },
    {
      "epoch": 0.10917221261506752,
      "grad_norm": 1.5116829872131348,
      "learning_rate": 9.989082778738493e-06,
      "loss": 3.0734,
      "step": 12500
    },
    {
      "epoch": 0.11353910111967021,
      "grad_norm": 1.5693246126174927,
      "learning_rate": 9.988646089888033e-06,
      "loss": 3.0732,
      "step": 13000
    },
    {
      "epoch": 0.11790598962427291,
      "grad_norm": 1.524153709411621,
      "learning_rate": 9.988209401037573e-06,
      "loss": 3.0743,
      "step": 13500
    },
    {
      "epoch": 0.12227287812887562,
      "grad_norm": 1.5216927528381348,
      "learning_rate": 9.987772712187114e-06,
      "loss": 3.0668,
      "step": 14000
    },
    {
      "epoch": 0.1266397666334783,
      "grad_norm": 1.5442315340042114,
      "learning_rate": 9.987336023336654e-06,
      "loss": 3.0755,
      "step": 14500
    },
    {
      "epoch": 0.13100665513808102,
      "grad_norm": 1.555234670639038,
      "learning_rate": 9.986899334486194e-06,
      "loss": 3.0704,
      "step": 15000
    },
    {
      "epoch": 0.13537354364268372,
      "grad_norm": 1.5696550607681274,
      "learning_rate": 9.986462645635732e-06,
      "loss": 3.0757,
      "step": 15500
    },
    {
      "epoch": 0.1397404321472864,
      "grad_norm": 1.531101107597351,
      "learning_rate": 9.986025956785272e-06,
      "loss": 3.0655,
      "step": 16000
    },
    {
      "epoch": 0.1441073206518891,
      "grad_norm": 1.544026494026184,
      "learning_rate": 9.985589267934812e-06,
      "loss": 3.0747,
      "step": 16500
    },
    {
      "epoch": 0.14847420915649182,
      "grad_norm": 1.5147799253463745,
      "learning_rate": 9.985152579084352e-06,
      "loss": 3.0649,
      "step": 17000
    },
    {
      "epoch": 0.15284109766109452,
      "grad_norm": 1.5733822584152222,
      "learning_rate": 9.984715890233891e-06,
      "loss": 3.0663,
      "step": 17500
    },
    {
      "epoch": 0.15720798616569723,
      "grad_norm": 1.590242862701416,
      "learning_rate": 9.984279201383431e-06,
      "loss": 3.0663,
      "step": 18000
    },
    {
      "epoch": 0.1615748746702999,
      "grad_norm": 1.5203492641448975,
      "learning_rate": 9.983842512532971e-06,
      "loss": 3.0747,
      "step": 18500
    },
    {
      "epoch": 0.1659417631749026,
      "grad_norm": 1.5461294651031494,
      "learning_rate": 9.98340582368251e-06,
      "loss": 3.0689,
      "step": 19000
    },
    {
      "epoch": 0.17030865167950532,
      "grad_norm": 1.5351060628890991,
      "learning_rate": 9.98296913483205e-06,
      "loss": 3.0689,
      "step": 19500
    },
    {
      "epoch": 0.17467554018410802,
      "grad_norm": 1.5561175346374512,
      "learning_rate": 9.98253244598159e-06,
      "loss": 3.0756,
      "step": 20000
    },
    {
      "epoch": 0.17904242868871073,
      "grad_norm": 1.51571524143219,
      "learning_rate": 9.982095757131129e-06,
      "loss": 3.0644,
      "step": 20500
    },
    {
      "epoch": 0.1834093171933134,
      "grad_norm": 1.5357000827789307,
      "learning_rate": 9.981659068280668e-06,
      "loss": 3.0745,
      "step": 21000
    },
    {
      "epoch": 0.18777620569791612,
      "grad_norm": 1.5208464860916138,
      "learning_rate": 9.981222379430208e-06,
      "loss": 3.0661,
      "step": 21500
    },
    {
      "epoch": 0.19214309420251882,
      "grad_norm": 1.4807689189910889,
      "learning_rate": 9.980785690579748e-06,
      "loss": 3.0718,
      "step": 22000
    },
    {
      "epoch": 0.19650998270712153,
      "grad_norm": 1.5423370599746704,
      "learning_rate": 9.980349001729288e-06,
      "loss": 3.0672,
      "step": 22500
    },
    {
      "epoch": 0.20087687121172423,
      "grad_norm": 1.5475423336029053,
      "learning_rate": 9.97991231287883e-06,
      "loss": 3.0707,
      "step": 23000
    },
    {
      "epoch": 0.2052437597163269,
      "grad_norm": 1.5928481817245483,
      "learning_rate": 9.97947562402837e-06,
      "loss": 3.0704,
      "step": 23500
    },
    {
      "epoch": 0.20961064822092962,
      "grad_norm": 1.5184266567230225,
      "learning_rate": 9.979038935177907e-06,
      "loss": 3.0668,
      "step": 24000
    },
    {
      "epoch": 0.21397753672553232,
      "grad_norm": 1.6194623708724976,
      "learning_rate": 9.978602246327447e-06,
      "loss": 3.0604,
      "step": 24500
    },
    {
      "epoch": 0.21834442523013503,
      "grad_norm": 1.5114210844039917,
      "learning_rate": 9.978165557476987e-06,
      "loss": 3.0643,
      "step": 25000
    },
    {
      "epoch": 0.22271131373473774,
      "grad_norm": 1.4907346963882446,
      "learning_rate": 9.977728868626527e-06,
      "loss": 3.0627,
      "step": 25500
    },
    {
      "epoch": 0.22707820223934042,
      "grad_norm": 1.55216383934021,
      "learning_rate": 9.977292179776067e-06,
      "loss": 3.0705,
      "step": 26000
    },
    {
      "epoch": 0.23144509074394312,
      "grad_norm": 1.5280052423477173,
      "learning_rate": 9.976855490925607e-06,
      "loss": 3.0642,
      "step": 26500
    },
    {
      "epoch": 0.23581197924854583,
      "grad_norm": 1.6076959371566772,
      "learning_rate": 9.976418802075146e-06,
      "loss": 3.0607,
      "step": 27000
    },
    {
      "epoch": 0.24017886775314853,
      "grad_norm": 1.5258777141571045,
      "learning_rate": 9.975982113224686e-06,
      "loss": 3.0711,
      "step": 27500
    },
    {
      "epoch": 0.24454575625775124,
      "grad_norm": 1.5106350183486938,
      "learning_rate": 9.975545424374226e-06,
      "loss": 3.0659,
      "step": 28000
    },
    {
      "epoch": 0.24891264476235392,
      "grad_norm": 1.5026884078979492,
      "learning_rate": 9.975108735523766e-06,
      "loss": 3.0629,
      "step": 28500
    },
    {
      "epoch": 0.2532795332669566,
      "grad_norm": 1.4955607652664185,
      "learning_rate": 9.974672046673306e-06,
      "loss": 3.0642,
      "step": 29000
    },
    {
      "epoch": 0.25764642177155933,
      "grad_norm": 1.5555437803268433,
      "learning_rate": 9.974235357822844e-06,
      "loss": 3.0525,
      "step": 29500
    },
    {
      "epoch": 0.26201331027616204,
      "grad_norm": 1.5264055728912354,
      "learning_rate": 9.973798668972384e-06,
      "loss": 3.0656,
      "step": 30000
    },
    {
      "epoch": 0.26638019878076474,
      "grad_norm": 1.5161739587783813,
      "learning_rate": 9.973361980121923e-06,
      "loss": 3.067,
      "step": 30500
    },
    {
      "epoch": 0.27074708728536745,
      "grad_norm": 1.640349268913269,
      "learning_rate": 9.972925291271463e-06,
      "loss": 3.0617,
      "step": 31000
    },
    {
      "epoch": 0.27511397578997016,
      "grad_norm": 1.5388609170913696,
      "learning_rate": 9.972488602421003e-06,
      "loss": 3.0584,
      "step": 31500
    },
    {
      "epoch": 0.2794808642945728,
      "grad_norm": 1.5143171548843384,
      "learning_rate": 9.972051913570545e-06,
      "loss": 3.0563,
      "step": 32000
    },
    {
      "epoch": 0.2838477527991755,
      "grad_norm": 1.5235260725021362,
      "learning_rate": 9.971615224720084e-06,
      "loss": 3.0601,
      "step": 32500
    },
    {
      "epoch": 0.2882146413037782,
      "grad_norm": 1.4914684295654297,
      "learning_rate": 9.971178535869623e-06,
      "loss": 3.052,
      "step": 33000
    },
    {
      "epoch": 0.2925815298083809,
      "grad_norm": 1.4766864776611328,
      "learning_rate": 9.970741847019162e-06,
      "loss": 3.0549,
      "step": 33500
    },
    {
      "epoch": 0.29694841831298363,
      "grad_norm": 1.5566754341125488,
      "learning_rate": 9.970305158168702e-06,
      "loss": 3.055,
      "step": 34000
    },
    {
      "epoch": 0.30131530681758634,
      "grad_norm": 1.546927809715271,
      "learning_rate": 9.969868469318242e-06,
      "loss": 3.063,
      "step": 34500
    },
    {
      "epoch": 0.30568219532218904,
      "grad_norm": 1.5629053115844727,
      "learning_rate": 9.969431780467782e-06,
      "loss": 3.059,
      "step": 35000
    },
    {
      "epoch": 0.31004908382679175,
      "grad_norm": 1.538415789604187,
      "learning_rate": 9.968995091617322e-06,
      "loss": 3.054,
      "step": 35500
    },
    {
      "epoch": 0.31441597233139446,
      "grad_norm": 1.541513442993164,
      "learning_rate": 9.968558402766862e-06,
      "loss": 3.0538,
      "step": 36000
    },
    {
      "epoch": 0.31878286083599716,
      "grad_norm": 1.5637692213058472,
      "learning_rate": 9.968121713916401e-06,
      "loss": 3.0538,
      "step": 36500
    },
    {
      "epoch": 0.3231497493405998,
      "grad_norm": 1.5472180843353271,
      "learning_rate": 9.967685025065941e-06,
      "loss": 3.0539,
      "step": 37000
    },
    {
      "epoch": 0.3275166378452025,
      "grad_norm": 1.4945605993270874,
      "learning_rate": 9.967248336215481e-06,
      "loss": 3.0576,
      "step": 37500
    },
    {
      "epoch": 0.3318835263498052,
      "grad_norm": 1.4775537252426147,
      "learning_rate": 9.96681164736502e-06,
      "loss": 3.0553,
      "step": 38000
    },
    {
      "epoch": 0.33625041485440793,
      "grad_norm": 1.487927794456482,
      "learning_rate": 9.966374958514559e-06,
      "loss": 3.0577,
      "step": 38500
    },
    {
      "epoch": 0.34061730335901064,
      "grad_norm": 1.538345456123352,
      "learning_rate": 9.965938269664099e-06,
      "loss": 3.0541,
      "step": 39000
    },
    {
      "epoch": 0.34498419186361334,
      "grad_norm": 1.543130874633789,
      "learning_rate": 9.965501580813639e-06,
      "loss": 3.0559,
      "step": 39500
    },
    {
      "epoch": 0.34935108036821605,
      "grad_norm": 1.5487135648727417,
      "learning_rate": 9.965064891963178e-06,
      "loss": 3.0514,
      "step": 40000
    },
    {
      "epoch": 0.35371796887281876,
      "grad_norm": 1.5321663618087769,
      "learning_rate": 9.964628203112718e-06,
      "loss": 3.0524,
      "step": 40500
    },
    {
      "epoch": 0.35808485737742146,
      "grad_norm": 1.5257816314697266,
      "learning_rate": 9.96419151426226e-06,
      "loss": 3.056,
      "step": 41000
    },
    {
      "epoch": 0.36245174588202417,
      "grad_norm": 1.478402018547058,
      "learning_rate": 9.9637548254118e-06,
      "loss": 3.0518,
      "step": 41500
    },
    {
      "epoch": 0.3668186343866268,
      "grad_norm": 1.5002671480178833,
      "learning_rate": 9.963318136561338e-06,
      "loss": 3.0531,
      "step": 42000
    },
    {
      "epoch": 0.3711855228912295,
      "grad_norm": 1.5446417331695557,
      "learning_rate": 9.962881447710878e-06,
      "loss": 3.0483,
      "step": 42500
    },
    {
      "epoch": 0.37555241139583223,
      "grad_norm": 1.4993116855621338,
      "learning_rate": 9.962444758860417e-06,
      "loss": 3.0522,
      "step": 43000
    },
    {
      "epoch": 0.37991929990043494,
      "grad_norm": 1.576460838317871,
      "learning_rate": 9.962008070009957e-06,
      "loss": 3.0547,
      "step": 43500
    },
    {
      "epoch": 0.38428618840503764,
      "grad_norm": 1.5173957347869873,
      "learning_rate": 9.961571381159497e-06,
      "loss": 3.0484,
      "step": 44000
    },
    {
      "epoch": 0.38865307690964035,
      "grad_norm": 1.5036677122116089,
      "learning_rate": 9.961134692309037e-06,
      "loss": 3.0432,
      "step": 44500
    },
    {
      "epoch": 0.39301996541424306,
      "grad_norm": 1.4855741262435913,
      "learning_rate": 9.960698003458577e-06,
      "loss": 3.0552,
      "step": 45000
    },
    {
      "epoch": 0.39738685391884576,
      "grad_norm": 1.4877578020095825,
      "learning_rate": 9.960261314608117e-06,
      "loss": 3.077,
      "step": 45500
    },
    {
      "epoch": 0.40175374242344847,
      "grad_norm": 1.5206501483917236,
      "learning_rate": 9.959824625757656e-06,
      "loss": 3.0821,
      "step": 46000
    },
    {
      "epoch": 0.4061206309280511,
      "grad_norm": 1.4510574340820312,
      "learning_rate": 9.959387936907196e-06,
      "loss": 3.0795,
      "step": 46500
    },
    {
      "epoch": 0.4104875194326538,
      "grad_norm": 1.5214266777038574,
      "learning_rate": 9.958951248056736e-06,
      "loss": 3.076,
      "step": 47000
    },
    {
      "epoch": 0.41485440793725653,
      "grad_norm": 1.553749442100525,
      "learning_rate": 9.958514559206274e-06,
      "loss": 3.0868,
      "step": 47500
    },
    {
      "epoch": 0.41922129644185924,
      "grad_norm": 1.5479093790054321,
      "learning_rate": 9.958077870355814e-06,
      "loss": 3.0771,
      "step": 48000
    },
    {
      "epoch": 0.42358818494646194,
      "grad_norm": 1.4554359912872314,
      "learning_rate": 9.957641181505354e-06,
      "loss": 3.0783,
      "step": 48500
    },
    {
      "epoch": 0.42795507345106465,
      "grad_norm": 1.482832908630371,
      "learning_rate": 9.957204492654894e-06,
      "loss": 3.074,
      "step": 49000
    },
    {
      "epoch": 0.4298328355080438,
      "step": 49215,
      "total_flos": 1.2345107794414797e+18,
      "train_loss": 3.067183213959159,
      "train_runtime": 46797.6916,
      "train_samples_per_second": 11743.883,
      "train_steps_per_second": 244.666
    }
  ],
  "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": 1.2345107794414797e+18,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": null
}