{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9997818023128955,
  "eval_steps": 500,
  "global_step": 36660,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.027274710888064588,
      "grad_norm": 8.132363319396973,
      "learning_rate": 1.976e-05,
      "loss": 3.5991,
      "step": 500
    },
    {
      "epoch": 0.054549421776129176,
      "grad_norm": 7.099025249481201,
      "learning_rate": 1.9818702290076336e-05,
      "loss": 2.8754,
      "step": 1000
    },
    {
      "epoch": 0.08182413266419376,
      "grad_norm": 7.820577621459961,
      "learning_rate": 1.963520258367587e-05,
      "loss": 2.6895,
      "step": 1500
    },
    {
      "epoch": 0.10909884355225835,
      "grad_norm": 5.448477268218994,
      "learning_rate": 1.94517028772754e-05,
      "loss": 2.6494,
      "step": 2000
    },
    {
      "epoch": 0.13637355444032292,
      "grad_norm": 6.391147613525391,
      "learning_rate": 1.926820317087493e-05,
      "loss": 2.5544,
      "step": 2500
    },
    {
      "epoch": 0.16364826532838753,
      "grad_norm": 5.117228031158447,
      "learning_rate": 1.908470346447446e-05,
      "loss": 2.5431,
      "step": 3000
    },
    {
      "epoch": 0.1909229762164521,
      "grad_norm": 4.08992338180542,
      "learning_rate": 1.890120375807399e-05,
      "loss": 2.5359,
      "step": 3500
    },
    {
      "epoch": 0.2181976871045167,
      "grad_norm": 3.984097957611084,
      "learning_rate": 1.871770405167352e-05,
      "loss": 2.4834,
      "step": 4000
    },
    {
      "epoch": 0.24547239799258128,
      "grad_norm": 3.8590118885040283,
      "learning_rate": 1.853420434527305e-05,
      "loss": 2.4899,
      "step": 4500
    },
    {
      "epoch": 0.27274710888064585,
      "grad_norm": 4.541988372802734,
      "learning_rate": 1.835070463887258e-05,
      "loss": 2.4608,
      "step": 5000
    },
    {
      "epoch": 0.30002181976871045,
      "grad_norm": 4.910805702209473,
      "learning_rate": 1.816720493247211e-05,
      "loss": 2.402,
      "step": 5500
    },
    {
      "epoch": 0.32729653065677505,
      "grad_norm": 4.220065116882324,
      "learning_rate": 1.798370522607164e-05,
      "loss": 2.4483,
      "step": 6000
    },
    {
      "epoch": 0.3545712415448396,
      "grad_norm": 3.9413557052612305,
      "learning_rate": 1.780020551967117e-05,
      "loss": 2.4271,
      "step": 6500
    },
    {
      "epoch": 0.3818459524329042,
      "grad_norm": 2.890425443649292,
      "learning_rate": 1.76167058132707e-05,
      "loss": 2.3884,
      "step": 7000
    },
    {
      "epoch": 0.4091206633209688,
      "grad_norm": 3.186286211013794,
      "learning_rate": 1.743320610687023e-05,
      "loss": 2.364,
      "step": 7500
    },
    {
      "epoch": 0.4363953742090334,
      "grad_norm": 3.4042603969573975,
      "learning_rate": 1.7249706400469762e-05,
      "loss": 2.4374,
      "step": 8000
    },
    {
      "epoch": 0.46367008509709795,
      "grad_norm": 2.714256763458252,
      "learning_rate": 1.7066206694069292e-05,
      "loss": 2.3535,
      "step": 8500
    },
    {
      "epoch": 0.49094479598516255,
      "grad_norm": 3.9176571369171143,
      "learning_rate": 1.6882706987668822e-05,
      "loss": 2.379,
      "step": 9000
    },
    {
      "epoch": 0.5182195068732272,
      "grad_norm": 3.3017940521240234,
      "learning_rate": 1.669920728126835e-05,
      "loss": 2.3246,
      "step": 9500
    },
    {
      "epoch": 0.5454942177612917,
      "grad_norm": 3.9203879833221436,
      "learning_rate": 1.651570757486788e-05,
      "loss": 2.3452,
      "step": 10000
    },
    {
      "epoch": 0.5727689286493564,
      "grad_norm": 3.474233627319336,
      "learning_rate": 1.6332207868467413e-05,
      "loss": 2.3192,
      "step": 10500
    },
    {
      "epoch": 0.6000436395374209,
      "grad_norm": 3.47390079498291,
      "learning_rate": 1.6148708162066943e-05,
      "loss": 2.3775,
      "step": 11000
    },
    {
      "epoch": 0.6273183504254854,
      "grad_norm": 3.5392141342163086,
      "learning_rate": 1.5965208455666473e-05,
      "loss": 2.3436,
      "step": 11500
    },
    {
      "epoch": 0.6545930613135501,
      "grad_norm": 3.213120698928833,
      "learning_rate": 1.5782075748678803e-05,
      "loss": 2.2958,
      "step": 12000
    },
    {
      "epoch": 0.6818677722016147,
      "grad_norm": 3.356034278869629,
      "learning_rate": 1.5598576042278333e-05,
      "loss": 2.3013,
      "step": 12500
    },
    {
      "epoch": 0.7091424830896792,
      "grad_norm": 2.832211494445801,
      "learning_rate": 1.5415076335877863e-05,
      "loss": 2.3138,
      "step": 13000
    },
    {
      "epoch": 0.7364171939777439,
      "grad_norm": 3.0613503456115723,
      "learning_rate": 1.5231576629477394e-05,
      "loss": 2.302,
      "step": 13500
    },
    {
      "epoch": 0.7636919048658084,
      "grad_norm": 3.324085235595703,
      "learning_rate": 1.5048076923076924e-05,
      "loss": 2.3072,
      "step": 14000
    },
    {
      "epoch": 0.7909666157538731,
      "grad_norm": 3.129357099533081,
      "learning_rate": 1.4864577216676454e-05,
      "loss": 2.3137,
      "step": 14500
    },
    {
      "epoch": 0.8182413266419376,
      "grad_norm": 3.2217562198638916,
      "learning_rate": 1.4681444509688784e-05,
      "loss": 2.2899,
      "step": 15000
    },
    {
      "epoch": 0.8455160375300022,
      "grad_norm": 3.6490132808685303,
      "learning_rate": 1.4497944803288314e-05,
      "loss": 2.3015,
      "step": 15500
    },
    {
      "epoch": 0.8727907484180668,
      "grad_norm": 3.53132700920105,
      "learning_rate": 1.4314445096887846e-05,
      "loss": 2.2479,
      "step": 16000
    },
    {
      "epoch": 0.9000654593061314,
      "grad_norm": 2.9956769943237305,
      "learning_rate": 1.4130945390487376e-05,
      "loss": 2.2878,
      "step": 16500
    },
    {
      "epoch": 0.9273401701941959,
      "grad_norm": 3.027482271194458,
      "learning_rate": 1.3947445684086906e-05,
      "loss": 2.2656,
      "step": 17000
    },
    {
      "epoch": 0.9546148810822606,
      "grad_norm": 2.76747465133667,
      "learning_rate": 1.3763945977686436e-05,
      "loss": 2.3085,
      "step": 17500
    },
    {
      "epoch": 0.9818895919703251,
      "grad_norm": 3.367910385131836,
      "learning_rate": 1.3580446271285968e-05,
      "loss": 2.2726,
      "step": 18000
    },
    {
      "epoch": 1.0091643028583897,
      "grad_norm": 3.19476056098938,
      "learning_rate": 1.3396946564885498e-05,
      "loss": 2.2459,
      "step": 18500
    },
    {
      "epoch": 1.0364390137464543,
      "grad_norm": 2.947768211364746,
      "learning_rate": 1.3213446858485028e-05,
      "loss": 2.2225,
      "step": 19000
    },
    {
      "epoch": 1.063713724634519,
      "grad_norm": 3.5084404945373535,
      "learning_rate": 1.3030314151497358e-05,
      "loss": 2.2222,
      "step": 19500
    },
    {
      "epoch": 1.0909884355225834,
      "grad_norm": 3.5933291912078857,
      "learning_rate": 1.284681444509689e-05,
      "loss": 2.2169,
      "step": 20000
    },
    {
      "epoch": 1.118263146410648,
      "grad_norm": 3.1414079666137695,
      "learning_rate": 1.266331473869642e-05,
      "loss": 2.2104,
      "step": 20500
    },
    {
      "epoch": 1.1455378572987127,
      "grad_norm": 3.1075944900512695,
      "learning_rate": 1.247981503229595e-05,
      "loss": 2.1874,
      "step": 21000
    },
    {
      "epoch": 1.1728125681867771,
      "grad_norm": 3.304654598236084,
      "learning_rate": 1.229668232530828e-05,
      "loss": 2.2282,
      "step": 21500
    },
    {
      "epoch": 1.2000872790748418,
      "grad_norm": 2.92022705078125,
      "learning_rate": 1.211318261890781e-05,
      "loss": 2.2337,
      "step": 22000
    },
    {
      "epoch": 1.2273619899629065,
      "grad_norm": 3.450542449951172,
      "learning_rate": 1.1929682912507342e-05,
      "loss": 2.1784,
      "step": 22500
    },
    {
      "epoch": 1.254636700850971,
      "grad_norm": 2.8238916397094727,
      "learning_rate": 1.1746183206106872e-05,
      "loss": 2.2287,
      "step": 23000
    },
    {
      "epoch": 1.2819114117390356,
      "grad_norm": 2.935835123062134,
      "learning_rate": 1.1563050499119202e-05,
      "loss": 2.2062,
      "step": 23500
    },
    {
      "epoch": 1.3091861226271002,
      "grad_norm": 2.8934576511383057,
      "learning_rate": 1.1379550792718732e-05,
      "loss": 2.1836,
      "step": 24000
    },
    {
      "epoch": 1.3364608335151646,
      "grad_norm": 2.884962797164917,
      "learning_rate": 1.1196051086318263e-05,
      "loss": 2.1632,
      "step": 24500
    },
    {
      "epoch": 1.3637355444032293,
      "grad_norm": 3.163294792175293,
      "learning_rate": 1.1012551379917793e-05,
      "loss": 2.1715,
      "step": 25000
    },
    {
      "epoch": 1.391010255291294,
      "grad_norm": 3.31972599029541,
      "learning_rate": 1.0829051673517323e-05,
      "loss": 2.1867,
      "step": 25500
    },
    {
      "epoch": 1.4182849661793586,
      "grad_norm": 2.9578099250793457,
      "learning_rate": 1.0645918966529653e-05,
      "loss": 2.1851,
      "step": 26000
    },
    {
      "epoch": 1.445559677067423,
      "grad_norm": 3.1862030029296875,
      "learning_rate": 1.0462419260129184e-05,
      "loss": 2.1805,
      "step": 26500
    },
    {
      "epoch": 1.4728343879554877,
      "grad_norm": 3.1671643257141113,
      "learning_rate": 1.0278919553728715e-05,
      "loss": 2.1129,
      "step": 27000
    },
    {
      "epoch": 1.5001090988435521,
      "grad_norm": 2.691154718399048,
      "learning_rate": 1.0095786846741047e-05,
      "loss": 2.1386,
      "step": 27500
    },
    {
      "epoch": 1.5273838097316168,
      "grad_norm": 3.2530946731567383,
      "learning_rate": 9.912287140340577e-06,
      "loss": 2.2013,
      "step": 28000
    },
    {
      "epoch": 1.5546585206196815,
      "grad_norm": 2.7970876693725586,
      "learning_rate": 9.728787433940107e-06,
      "loss": 2.1913,
      "step": 28500
    },
    {
      "epoch": 1.5819332315077461,
      "grad_norm": 3.6721036434173584,
      "learning_rate": 9.545287727539637e-06,
      "loss": 2.1596,
      "step": 29000
    },
    {
      "epoch": 1.6092079423958106,
      "grad_norm": 3.152144432067871,
      "learning_rate": 9.361788021139167e-06,
      "loss": 2.1486,
      "step": 29500
    },
    {
      "epoch": 1.6364826532838752,
      "grad_norm": 3.174539089202881,
      "learning_rate": 9.178288314738697e-06,
      "loss": 2.169,
      "step": 30000
    },
    {
      "epoch": 1.6637573641719396,
      "grad_norm": 2.7281394004821777,
      "learning_rate": 8.994788608338227e-06,
      "loss": 2.1478,
      "step": 30500
    },
    {
      "epoch": 1.6910320750600043,
      "grad_norm": 3.2886815071105957,
      "learning_rate": 8.811288901937757e-06,
      "loss": 2.1372,
      "step": 31000
    },
    {
      "epoch": 1.718306785948069,
      "grad_norm": 2.9683258533477783,
      "learning_rate": 8.628156194950089e-06,
      "loss": 2.1335,
      "step": 31500
    },
    {
      "epoch": 1.7455814968361336,
      "grad_norm": 2.619202136993408,
      "learning_rate": 8.444656488549619e-06,
      "loss": 2.162,
      "step": 32000
    },
    {
      "epoch": 1.7728562077241983,
      "grad_norm": 2.4466633796691895,
      "learning_rate": 8.261156782149149e-06,
      "loss": 2.1595,
      "step": 32500
    },
    {
      "epoch": 1.8001309186122627,
      "grad_norm": 3.0279412269592285,
      "learning_rate": 8.077657075748679e-06,
      "loss": 2.1199,
      "step": 33000
    },
    {
      "epoch": 1.8274056295003271,
      "grad_norm": 2.4983465671539307,
      "learning_rate": 7.89415736934821e-06,
      "loss": 2.1368,
      "step": 33500
    },
    {
      "epoch": 1.8546803403883918,
      "grad_norm": 3.1389667987823486,
      "learning_rate": 7.71102466236054e-06,
      "loss": 2.1801,
      "step": 34000
    },
    {
      "epoch": 1.8819550512764565,
      "grad_norm": 3.049088716506958,
      "learning_rate": 7.5275249559600715e-06,
      "loss": 2.151,
      "step": 34500
    },
    {
      "epoch": 1.9092297621645211,
      "grad_norm": 2.899380922317505,
      "learning_rate": 7.344025249559601e-06,
      "loss": 2.1363,
      "step": 35000
    },
    {
      "epoch": 1.9365044730525858,
      "grad_norm": 3.1051132678985596,
      "learning_rate": 7.1605255431591316e-06,
      "loss": 2.1412,
      "step": 35500
    },
    {
      "epoch": 1.9637791839406502,
      "grad_norm": 2.9140512943267822,
      "learning_rate": 6.9773928361714625e-06,
      "loss": 2.1659,
      "step": 36000
    },
    {
      "epoch": 1.9910538948287149,
      "grad_norm": 2.7826218605041504,
      "learning_rate": 6.793893129770993e-06,
      "loss": 2.1352,
      "step": 36500
    }
  ],
  "logging_steps": 500,
  "max_steps": 54996,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1833,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9578965893120000.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}