{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 9851,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00010152026598309688,
      "grad_norm": 20.906557083129883,
      "learning_rate": 0.0,
      "loss": 13.5217,
      "step": 1
    },
    {
      "epoch": 0.010152026598309688,
      "grad_norm": 33.71811294555664,
      "learning_rate": 2.008113590263692e-05,
      "loss": 10.8681,
      "step": 100
    },
    {
      "epoch": 0.020304053196619376,
      "grad_norm": 27.14719581604004,
      "learning_rate": 4.036511156186613e-05,
      "loss": 7.3949,
      "step": 200
    },
    {
      "epoch": 0.030456079794929064,
      "grad_norm": 22.181039810180664,
      "learning_rate": 6.064908722109534e-05,
      "loss": 5.9034,
      "step": 300
    },
    {
      "epoch": 0.04060810639323875,
      "grad_norm": 14.685601234436035,
      "learning_rate": 8.093306288032456e-05,
      "loss": 5.0405,
      "step": 400
    },
    {
      "epoch": 0.05076013299154844,
      "grad_norm": 12.026344299316406,
      "learning_rate": 9.9935883735841e-05,
      "loss": 4.4838,
      "step": 500
    },
    {
      "epoch": 0.06091215958985813,
      "grad_norm": 14.332210540771484,
      "learning_rate": 9.886727933319086e-05,
      "loss": 4.0977,
      "step": 600
    },
    {
      "epoch": 0.07106418618816782,
      "grad_norm": 10.582870483398438,
      "learning_rate": 9.779867493054072e-05,
      "loss": 3.8085,
      "step": 700
    },
    {
      "epoch": 0.0812162127864775,
      "grad_norm": 8.66326904296875,
      "learning_rate": 9.673007052789058e-05,
      "loss": 3.6189,
      "step": 800
    },
    {
      "epoch": 0.09136823938478719,
      "grad_norm": 8.398235321044922,
      "learning_rate": 9.566146612524043e-05,
      "loss": 3.4115,
      "step": 900
    },
    {
      "epoch": 0.10152026598309688,
      "grad_norm": 8.620134353637695,
      "learning_rate": 9.45928617225903e-05,
      "loss": 3.3107,
      "step": 1000
    },
    {
      "epoch": 0.11167229258140657,
      "grad_norm": 8.263737678527832,
      "learning_rate": 9.352425731994017e-05,
      "loss": 3.2162,
      "step": 1100
    },
    {
      "epoch": 0.12182431917971626,
      "grad_norm": 7.5817646980285645,
      "learning_rate": 9.245565291729003e-05,
      "loss": 3.14,
      "step": 1200
    },
    {
      "epoch": 0.13197634577802594,
      "grad_norm": 8.10513973236084,
      "learning_rate": 9.138704851463988e-05,
      "loss": 3.0052,
      "step": 1300
    },
    {
      "epoch": 0.14212837237633563,
      "grad_norm": 9.334915161132812,
      "learning_rate": 9.031844411198974e-05,
      "loss": 2.9633,
      "step": 1400
    },
    {
      "epoch": 0.15228039897464532,
      "grad_norm": 6.445065975189209,
      "learning_rate": 8.924983970933962e-05,
      "loss": 2.9094,
      "step": 1500
    },
    {
      "epoch": 0.162432425572955,
      "grad_norm": 7.780778884887695,
      "learning_rate": 8.818123530668947e-05,
      "loss": 2.8677,
      "step": 1600
    },
    {
      "epoch": 0.1725844521712647,
      "grad_norm": 7.130796432495117,
      "learning_rate": 8.711263090403933e-05,
      "loss": 2.7924,
      "step": 1700
    },
    {
      "epoch": 0.18273647876957438,
      "grad_norm": 6.419895648956299,
      "learning_rate": 8.604402650138918e-05,
      "loss": 2.7478,
      "step": 1800
    },
    {
      "epoch": 0.19288850536788407,
      "grad_norm": 7.287320613861084,
      "learning_rate": 8.497542209873905e-05,
      "loss": 2.6924,
      "step": 1900
    },
    {
      "epoch": 0.20304053196619376,
      "grad_norm": 9.301968574523926,
      "learning_rate": 8.390681769608892e-05,
      "loss": 2.7025,
      "step": 2000
    },
    {
      "epoch": 0.21319255856450345,
      "grad_norm": 6.076135635375977,
      "learning_rate": 8.283821329343877e-05,
      "loss": 2.6538,
      "step": 2100
    },
    {
      "epoch": 0.22334458516281314,
      "grad_norm": 8.065519332885742,
      "learning_rate": 8.176960889078863e-05,
      "loss": 2.5985,
      "step": 2200
    },
    {
      "epoch": 0.23349661176112282,
      "grad_norm": 6.776027679443359,
      "learning_rate": 8.07010044881385e-05,
      "loss": 2.5731,
      "step": 2300
    },
    {
      "epoch": 0.2436486383594325,
      "grad_norm": 7.288259029388428,
      "learning_rate": 7.963240008548835e-05,
      "loss": 2.5714,
      "step": 2400
    },
    {
      "epoch": 0.2538006649577422,
      "grad_norm": 7.614097595214844,
      "learning_rate": 7.856379568283822e-05,
      "loss": 2.5819,
      "step": 2500
    },
    {
      "epoch": 0.2639526915560519,
      "grad_norm": 6.544124603271484,
      "learning_rate": 7.749519128018808e-05,
      "loss": 2.5085,
      "step": 2600
    },
    {
      "epoch": 0.27410471815436155,
      "grad_norm": 6.303622722625732,
      "learning_rate": 7.642658687753794e-05,
      "loss": 2.4782,
      "step": 2700
    },
    {
      "epoch": 0.28425674475267126,
      "grad_norm": 6.230086803436279,
      "learning_rate": 7.53579824748878e-05,
      "loss": 2.4933,
      "step": 2800
    },
    {
      "epoch": 0.2944087713509809,
      "grad_norm": 7.580053806304932,
      "learning_rate": 7.428937807223765e-05,
      "loss": 2.4438,
      "step": 2900
    },
    {
      "epoch": 0.30456079794929064,
      "grad_norm": 7.0785675048828125,
      "learning_rate": 7.322077366958752e-05,
      "loss": 2.4569,
      "step": 3000
    },
    {
      "epoch": 0.3147128245476003,
      "grad_norm": 7.264654159545898,
      "learning_rate": 7.215216926693739e-05,
      "loss": 2.348,
      "step": 3100
    },
    {
      "epoch": 0.32486485114591,
      "grad_norm": 8.996079444885254,
      "learning_rate": 7.108356486428725e-05,
      "loss": 2.4008,
      "step": 3200
    },
    {
      "epoch": 0.3350168777442197,
      "grad_norm": 7.104641914367676,
      "learning_rate": 7.00149604616371e-05,
      "loss": 2.3889,
      "step": 3300
    },
    {
      "epoch": 0.3451689043425294,
      "grad_norm": 6.1596221923828125,
      "learning_rate": 6.894635605898697e-05,
      "loss": 2.3832,
      "step": 3400
    },
    {
      "epoch": 0.35532093094083905,
      "grad_norm": 6.6536712646484375,
      "learning_rate": 6.787775165633682e-05,
      "loss": 2.3509,
      "step": 3500
    },
    {
      "epoch": 0.36547295753914877,
      "grad_norm": 8.533254623413086,
      "learning_rate": 6.680914725368669e-05,
      "loss": 2.3616,
      "step": 3600
    },
    {
      "epoch": 0.37562498413745843,
      "grad_norm": 6.878735542297363,
      "learning_rate": 6.574054285103655e-05,
      "loss": 2.2491,
      "step": 3700
    },
    {
      "epoch": 0.38577701073576814,
      "grad_norm": 6.529376029968262,
      "learning_rate": 6.467193844838642e-05,
      "loss": 2.239,
      "step": 3800
    },
    {
      "epoch": 0.3959290373340778,
      "grad_norm": 8.117202758789062,
      "learning_rate": 6.360333404573627e-05,
      "loss": 2.311,
      "step": 3900
    },
    {
      "epoch": 0.4060810639323875,
      "grad_norm": 7.16500186920166,
      "learning_rate": 6.253472964308614e-05,
      "loss": 2.2687,
      "step": 4000
    },
    {
      "epoch": 0.4162330905306972,
      "grad_norm": 7.756877899169922,
      "learning_rate": 6.1466125240436e-05,
      "loss": 2.2649,
      "step": 4100
    },
    {
      "epoch": 0.4263851171290069,
      "grad_norm": 6.9812469482421875,
      "learning_rate": 6.039752083778586e-05,
      "loss": 2.2575,
      "step": 4200
    },
    {
      "epoch": 0.43653714372731656,
      "grad_norm": 7.193202972412109,
      "learning_rate": 5.932891643513572e-05,
      "loss": 2.2156,
      "step": 4300
    },
    {
      "epoch": 0.4466891703256263,
      "grad_norm": 6.245234489440918,
      "learning_rate": 5.826031203248558e-05,
      "loss": 2.2088,
      "step": 4400
    },
    {
      "epoch": 0.45684119692393593,
      "grad_norm": 7.823598861694336,
      "learning_rate": 5.719170762983543e-05,
      "loss": 2.2095,
      "step": 4500
    },
    {
      "epoch": 0.46699322352224565,
      "grad_norm": 9.34554386138916,
      "learning_rate": 5.61231032271853e-05,
      "loss": 2.2101,
      "step": 4600
    },
    {
      "epoch": 0.4771452501205553,
      "grad_norm": 5.98472261428833,
      "learning_rate": 5.505449882453516e-05,
      "loss": 2.2042,
      "step": 4700
    },
    {
      "epoch": 0.487297276718865,
      "grad_norm": 6.595754146575928,
      "learning_rate": 5.398589442188502e-05,
      "loss": 2.1979,
      "step": 4800
    },
    {
      "epoch": 0.4974493033171747,
      "grad_norm": 6.549368858337402,
      "learning_rate": 5.291729001923488e-05,
      "loss": 2.1515,
      "step": 4900
    },
    {
      "epoch": 0.5076013299154843,
      "grad_norm": 6.540451526641846,
      "learning_rate": 5.184868561658475e-05,
      "loss": 2.1269,
      "step": 5000
    },
    {
      "epoch": 0.5177533565137941,
      "grad_norm": 5.891973972320557,
      "learning_rate": 5.07800812139346e-05,
      "loss": 2.145,
      "step": 5100
    },
    {
      "epoch": 0.5279053831121038,
      "grad_norm": 7.685993194580078,
      "learning_rate": 4.9711476811284465e-05,
      "loss": 2.141,
      "step": 5200
    },
    {
      "epoch": 0.5380574097104135,
      "grad_norm": 7.306082725524902,
      "learning_rate": 4.8642872408634326e-05,
      "loss": 2.1274,
      "step": 5300
    },
    {
      "epoch": 0.5482094363087231,
      "grad_norm": 5.640676498413086,
      "learning_rate": 4.757426800598419e-05,
      "loss": 2.129,
      "step": 5400
    },
    {
      "epoch": 0.5583614629070328,
      "grad_norm": 5.850844860076904,
      "learning_rate": 4.650566360333405e-05,
      "loss": 2.0921,
      "step": 5500
    },
    {
      "epoch": 0.5685134895053425,
      "grad_norm": 7.10348653793335,
      "learning_rate": 4.5437059200683905e-05,
      "loss": 2.135,
      "step": 5600
    },
    {
      "epoch": 0.5786655161036522,
      "grad_norm": 4.819798946380615,
      "learning_rate": 4.436845479803377e-05,
      "loss": 2.0731,
      "step": 5700
    },
    {
      "epoch": 0.5888175427019618,
      "grad_norm": 8.465038299560547,
      "learning_rate": 4.329985039538363e-05,
      "loss": 2.0882,
      "step": 5800
    },
    {
      "epoch": 0.5989695693002716,
      "grad_norm": 6.31635046005249,
      "learning_rate": 4.223124599273349e-05,
      "loss": 2.1084,
      "step": 5900
    },
    {
      "epoch": 0.6091215958985813,
      "grad_norm": 7.215450286865234,
      "learning_rate": 4.116264159008335e-05,
      "loss": 2.0788,
      "step": 6000
    },
    {
      "epoch": 0.619273622496891,
      "grad_norm": 6.831523895263672,
      "learning_rate": 4.009403718743321e-05,
      "loss": 2.0606,
      "step": 6100
    },
    {
      "epoch": 0.6294256490952006,
      "grad_norm": 6.920884132385254,
      "learning_rate": 3.9025432784783075e-05,
      "loss": 2.0504,
      "step": 6200
    },
    {
      "epoch": 0.6395776756935103,
      "grad_norm": 5.9327802658081055,
      "learning_rate": 3.7956828382132936e-05,
      "loss": 2.061,
      "step": 6300
    },
    {
      "epoch": 0.64972970229182,
      "grad_norm": 7.0381550788879395,
      "learning_rate": 3.68882239794828e-05,
      "loss": 2.021,
      "step": 6400
    },
    {
      "epoch": 0.6598817288901296,
      "grad_norm": 5.229250431060791,
      "learning_rate": 3.581961957683266e-05,
      "loss": 2.0095,
      "step": 6500
    },
    {
      "epoch": 0.6700337554884394,
      "grad_norm": 5.998141288757324,
      "learning_rate": 3.475101517418252e-05,
      "loss": 2.0756,
      "step": 6600
    },
    {
      "epoch": 0.6801857820867491,
      "grad_norm": 7.094350814819336,
      "learning_rate": 3.368241077153238e-05,
      "loss": 2.019,
      "step": 6700
    },
    {
      "epoch": 0.6903378086850588,
      "grad_norm": 6.863796710968018,
      "learning_rate": 3.2613806368882245e-05,
      "loss": 2.0057,
      "step": 6800
    },
    {
      "epoch": 0.7004898352833684,
      "grad_norm": 7.739531993865967,
      "learning_rate": 3.15452019662321e-05,
      "loss": 2.0452,
      "step": 6900
    },
    {
      "epoch": 0.7106418618816781,
      "grad_norm": 6.940340995788574,
      "learning_rate": 3.0476597563581964e-05,
      "loss": 2.0123,
      "step": 7000
    },
    {
      "epoch": 0.7207938884799878,
      "grad_norm": 6.965139865875244,
      "learning_rate": 2.9407993160931823e-05,
      "loss": 2.0076,
      "step": 7100
    },
    {
      "epoch": 0.7309459150782975,
      "grad_norm": 6.2129411697387695,
      "learning_rate": 2.8339388758281688e-05,
      "loss": 2.0221,
      "step": 7200
    },
    {
      "epoch": 0.7410979416766071,
      "grad_norm": 6.450573921203613,
      "learning_rate": 2.7270784355631546e-05,
      "loss": 1.9899,
      "step": 7300
    },
    {
      "epoch": 0.7512499682749169,
      "grad_norm": 6.273707866668701,
      "learning_rate": 2.6202179952981408e-05,
      "loss": 1.9801,
      "step": 7400
    },
    {
      "epoch": 0.7614019948732266,
      "grad_norm": 6.587959289550781,
      "learning_rate": 2.5133575550331266e-05,
      "loss": 1.973,
      "step": 7500
    },
    {
      "epoch": 0.7715540214715363,
      "grad_norm": 7.1678948402404785,
      "learning_rate": 2.4064971147681128e-05,
      "loss": 2.0192,
      "step": 7600
    },
    {
      "epoch": 0.7817060480698459,
      "grad_norm": 6.3077874183654785,
      "learning_rate": 2.299636674503099e-05,
      "loss": 1.9815,
      "step": 7700
    },
    {
      "epoch": 0.7918580746681556,
      "grad_norm": 7.233158111572266,
      "learning_rate": 2.192776234238085e-05,
      "loss": 2.0093,
      "step": 7800
    },
    {
      "epoch": 0.8020101012664653,
      "grad_norm": 6.688365459442139,
      "learning_rate": 2.0859157939730713e-05,
      "loss": 1.9382,
      "step": 7900
    },
    {
      "epoch": 0.812162127864775,
      "grad_norm": 6.085184574127197,
      "learning_rate": 1.9790553537080574e-05,
      "loss": 1.967,
      "step": 8000
    },
    {
      "epoch": 0.8223141544630846,
      "grad_norm": 6.654317378997803,
      "learning_rate": 1.8721949134430436e-05,
      "loss": 1.9463,
      "step": 8100
    },
    {
      "epoch": 0.8324661810613944,
      "grad_norm": 6.44837760925293,
      "learning_rate": 1.7653344731780298e-05,
      "loss": 1.9443,
      "step": 8200
    },
    {
      "epoch": 0.8426182076597041,
      "grad_norm": 6.596092224121094,
      "learning_rate": 1.658474032913016e-05,
      "loss": 1.9514,
      "step": 8300
    },
    {
      "epoch": 0.8527702342580138,
      "grad_norm": 6.20723295211792,
      "learning_rate": 1.5516135926480017e-05,
      "loss": 1.9623,
      "step": 8400
    },
    {
      "epoch": 0.8629222608563234,
      "grad_norm": 6.767848491668701,
      "learning_rate": 1.4447531523829879e-05,
      "loss": 1.959,
      "step": 8500
    },
    {
      "epoch": 0.8730742874546331,
      "grad_norm": 7.135315418243408,
      "learning_rate": 1.337892712117974e-05,
      "loss": 1.9661,
      "step": 8600
    },
    {
      "epoch": 0.8832263140529428,
      "grad_norm": 6.851596832275391,
      "learning_rate": 1.2310322718529602e-05,
      "loss": 1.8859,
      "step": 8700
    },
    {
      "epoch": 0.8933783406512525,
      "grad_norm": 5.696228504180908,
      "learning_rate": 1.1241718315879462e-05,
      "loss": 1.9249,
      "step": 8800
    },
    {
      "epoch": 0.9035303672495621,
      "grad_norm": 7.561517238616943,
      "learning_rate": 1.0173113913229324e-05,
      "loss": 1.8969,
      "step": 8900
    },
    {
      "epoch": 0.9136823938478719,
      "grad_norm": 5.516517639160156,
      "learning_rate": 9.104509510579184e-06,
      "loss": 1.952,
      "step": 9000
    },
    {
      "epoch": 0.9238344204461816,
      "grad_norm": 6.028636932373047,
      "learning_rate": 8.035905107929046e-06,
      "loss": 1.957,
      "step": 9100
    },
    {
      "epoch": 0.9339864470444913,
      "grad_norm": 5.458266735076904,
      "learning_rate": 6.9673007052789065e-06,
      "loss": 1.9445,
      "step": 9200
    },
    {
      "epoch": 0.9441384736428009,
      "grad_norm": 7.229398727416992,
      "learning_rate": 5.898696302628767e-06,
      "loss": 1.8976,
      "step": 9300
    },
    {
      "epoch": 0.9542905002411106,
      "grad_norm": 5.801294803619385,
      "learning_rate": 4.830091899978628e-06,
      "loss": 1.9468,
      "step": 9400
    },
    {
      "epoch": 0.9644425268394203,
      "grad_norm": 4.766051769256592,
      "learning_rate": 3.7614874973284893e-06,
      "loss": 1.8822,
      "step": 9500
    },
    {
      "epoch": 0.97459455343773,
      "grad_norm": 5.5660200119018555,
      "learning_rate": 2.69288309467835e-06,
      "loss": 1.8705,
      "step": 9600
    },
    {
      "epoch": 0.9847465800360397,
      "grad_norm": 7.402261734008789,
      "learning_rate": 1.6242786920282113e-06,
      "loss": 1.9297,
      "step": 9700
    },
    {
      "epoch": 0.9948986066343494,
      "grad_norm": 8.760640144348145,
      "learning_rate": 5.556742893780722e-07,
      "loss": 1.8924,
      "step": 9800
    }
  ],
  "logging_steps": 100,
  "max_steps": 9851,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5355046335700656.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}