{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 47724,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03143072667840081,
      "grad_norm": 12.45950698852539,
      "learning_rate": 4.947615455535999e-05,
      "loss": 3.2597,
      "step": 500
    },
    {
      "epoch": 0.06286145335680161,
      "grad_norm": 10.082228660583496,
      "learning_rate": 4.895230911071997e-05,
      "loss": 2.8456,
      "step": 1000
    },
    {
      "epoch": 0.09429218003520241,
      "grad_norm": 11.17715835571289,
      "learning_rate": 4.842846366607996e-05,
      "loss": 2.7087,
      "step": 1500
    },
    {
      "epoch": 0.12572290671360323,
      "grad_norm": 13.012718200683594,
      "learning_rate": 4.790461822143995e-05,
      "loss": 2.5935,
      "step": 2000
    },
    {
      "epoch": 0.15715363339200403,
      "grad_norm": 11.075544357299805,
      "learning_rate": 4.738077277679994e-05,
      "loss": 2.4862,
      "step": 2500
    },
    {
      "epoch": 0.18858436007040483,
      "grad_norm": 13.529458045959473,
      "learning_rate": 4.685692733215992e-05,
      "loss": 2.4535,
      "step": 3000
    },
    {
      "epoch": 0.22001508674880563,
      "grad_norm": 10.073129653930664,
      "learning_rate": 4.633308188751991e-05,
      "loss": 2.346,
      "step": 3500
    },
    {
      "epoch": 0.25144581342720645,
      "grad_norm": 11.378911018371582,
      "learning_rate": 4.580923644287989e-05,
      "loss": 2.3565,
      "step": 4000
    },
    {
      "epoch": 0.28287654010560725,
      "grad_norm": 10.559321403503418,
      "learning_rate": 4.528539099823988e-05,
      "loss": 2.2937,
      "step": 4500
    },
    {
      "epoch": 0.31430726678400805,
      "grad_norm": 10.023640632629395,
      "learning_rate": 4.476154555359987e-05,
      "loss": 2.2474,
      "step": 5000
    },
    {
      "epoch": 0.34573799346240885,
      "grad_norm": 9.890427589416504,
      "learning_rate": 4.423770010895986e-05,
      "loss": 2.1995,
      "step": 5500
    },
    {
      "epoch": 0.37716872014080965,
      "grad_norm": 11.694580078125,
      "learning_rate": 4.3713854664319845e-05,
      "loss": 2.2092,
      "step": 6000
    },
    {
      "epoch": 0.40859944681921045,
      "grad_norm": 10.519980430603027,
      "learning_rate": 4.319000921967983e-05,
      "loss": 2.1238,
      "step": 6500
    },
    {
      "epoch": 0.44003017349761125,
      "grad_norm": 9.350607872009277,
      "learning_rate": 4.2666163775039816e-05,
      "loss": 2.1271,
      "step": 7000
    },
    {
      "epoch": 0.47146090017601205,
      "grad_norm": 10.53451919555664,
      "learning_rate": 4.21423183303998e-05,
      "loss": 2.167,
      "step": 7500
    },
    {
      "epoch": 0.5028916268544129,
      "grad_norm": 11.468996047973633,
      "learning_rate": 4.1618472885759786e-05,
      "loss": 2.1129,
      "step": 8000
    },
    {
      "epoch": 0.5343223535328137,
      "grad_norm": 8.491691589355469,
      "learning_rate": 4.1094627441119775e-05,
      "loss": 2.0646,
      "step": 8500
    },
    {
      "epoch": 0.5657530802112145,
      "grad_norm": 20.68869400024414,
      "learning_rate": 4.0570781996479764e-05,
      "loss": 2.091,
      "step": 9000
    },
    {
      "epoch": 0.5971838068896153,
      "grad_norm": 8.865952491760254,
      "learning_rate": 4.0046936551839745e-05,
      "loss": 2.0602,
      "step": 9500
    },
    {
      "epoch": 0.6286145335680161,
      "grad_norm": 12.874096870422363,
      "learning_rate": 3.9523091107199734e-05,
      "loss": 2.0064,
      "step": 10000
    },
    {
      "epoch": 0.6600452602464169,
      "grad_norm": 9.40224838256836,
      "learning_rate": 3.8999245662559716e-05,
      "loss": 2.0428,
      "step": 10500
    },
    {
      "epoch": 0.6914759869248177,
      "grad_norm": 10.570882797241211,
      "learning_rate": 3.8475400217919705e-05,
      "loss": 2.0254,
      "step": 11000
    },
    {
      "epoch": 0.7229067136032185,
      "grad_norm": 12.117438316345215,
      "learning_rate": 3.795155477327969e-05,
      "loss": 2.003,
      "step": 11500
    },
    {
      "epoch": 0.7543374402816193,
      "grad_norm": 11.80452823638916,
      "learning_rate": 3.742770932863968e-05,
      "loss": 1.9471,
      "step": 12000
    },
    {
      "epoch": 0.7857681669600202,
      "grad_norm": 9.42381477355957,
      "learning_rate": 3.690386388399967e-05,
      "loss": 1.92,
      "step": 12500
    },
    {
      "epoch": 0.8171988936384209,
      "grad_norm": 14.402335166931152,
      "learning_rate": 3.638001843935965e-05,
      "loss": 1.939,
      "step": 13000
    },
    {
      "epoch": 0.8486296203168218,
      "grad_norm": 7.873473644256592,
      "learning_rate": 3.585617299471964e-05,
      "loss": 1.9271,
      "step": 13500
    },
    {
      "epoch": 0.8800603469952225,
      "grad_norm": 8.49427604675293,
      "learning_rate": 3.533232755007962e-05,
      "loss": 1.962,
      "step": 14000
    },
    {
      "epoch": 0.9114910736736234,
      "grad_norm": 9.598845481872559,
      "learning_rate": 3.480848210543961e-05,
      "loss": 1.8921,
      "step": 14500
    },
    {
      "epoch": 0.9429218003520241,
      "grad_norm": 10.971700668334961,
      "learning_rate": 3.42846366607996e-05,
      "loss": 1.8815,
      "step": 15000
    },
    {
      "epoch": 0.974352527030425,
      "grad_norm": 8.383946418762207,
      "learning_rate": 3.376079121615959e-05,
      "loss": 1.8714,
      "step": 15500
    },
    {
      "epoch": 1.0057832537088258,
      "grad_norm": 12.830138206481934,
      "learning_rate": 3.323694577151957e-05,
      "loss": 1.8797,
      "step": 16000
    },
    {
      "epoch": 1.0372139803872265,
      "grad_norm": 11.94987964630127,
      "learning_rate": 3.271310032687956e-05,
      "loss": 1.8339,
      "step": 16500
    },
    {
      "epoch": 1.0686447070656273,
      "grad_norm": 9.588113784790039,
      "learning_rate": 3.218925488223955e-05,
      "loss": 1.7749,
      "step": 17000
    },
    {
      "epoch": 1.1000754337440282,
      "grad_norm": 10.824664115905762,
      "learning_rate": 3.166540943759953e-05,
      "loss": 1.8145,
      "step": 17500
    },
    {
      "epoch": 1.131506160422429,
      "grad_norm": 11.03293228149414,
      "learning_rate": 3.114156399295952e-05,
      "loss": 1.7979,
      "step": 18000
    },
    {
      "epoch": 1.1629368871008299,
      "grad_norm": 8.919577598571777,
      "learning_rate": 3.061771854831951e-05,
      "loss": 1.8023,
      "step": 18500
    },
    {
      "epoch": 1.1943676137792305,
      "grad_norm": 7.814962863922119,
      "learning_rate": 3.0093873103679492e-05,
      "loss": 1.7587,
      "step": 19000
    },
    {
      "epoch": 1.2257983404576314,
      "grad_norm": 7.930530071258545,
      "learning_rate": 2.9570027659039477e-05,
      "loss": 1.7956,
      "step": 19500
    },
    {
      "epoch": 1.2572290671360322,
      "grad_norm": 12.548938751220703,
      "learning_rate": 2.9046182214399466e-05,
      "loss": 1.7925,
      "step": 20000
    },
    {
      "epoch": 1.2886597938144329,
      "grad_norm": 9.993666648864746,
      "learning_rate": 2.852233676975945e-05,
      "loss": 1.7609,
      "step": 20500
    },
    {
      "epoch": 1.3200905204928337,
      "grad_norm": 15.507658958435059,
      "learning_rate": 2.7998491325119436e-05,
      "loss": 1.7692,
      "step": 21000
    },
    {
      "epoch": 1.3515212471712346,
      "grad_norm": 11.69993782043457,
      "learning_rate": 2.7474645880479428e-05,
      "loss": 1.7253,
      "step": 21500
    },
    {
      "epoch": 1.3829519738496354,
      "grad_norm": 10.937728881835938,
      "learning_rate": 2.6950800435839413e-05,
      "loss": 1.732,
      "step": 22000
    },
    {
      "epoch": 1.4143827005280363,
      "grad_norm": 8.787737846374512,
      "learning_rate": 2.64269549911994e-05,
      "loss": 1.7335,
      "step": 22500
    },
    {
      "epoch": 1.4458134272064371,
      "grad_norm": 6.934820652008057,
      "learning_rate": 2.5903109546559384e-05,
      "loss": 1.6903,
      "step": 23000
    },
    {
      "epoch": 1.4772441538848378,
      "grad_norm": 9.452235221862793,
      "learning_rate": 2.537926410191937e-05,
      "loss": 1.7124,
      "step": 23500
    },
    {
      "epoch": 1.5086748805632386,
      "grad_norm": 7.406110763549805,
      "learning_rate": 2.4855418657279358e-05,
      "loss": 1.6978,
      "step": 24000
    },
    {
      "epoch": 1.5401056072416395,
      "grad_norm": 8.471650123596191,
      "learning_rate": 2.4331573212639343e-05,
      "loss": 1.6761,
      "step": 24500
    },
    {
      "epoch": 1.57153633392004,
      "grad_norm": 8.462206840515137,
      "learning_rate": 2.380772776799933e-05,
      "loss": 1.723,
      "step": 25000
    },
    {
      "epoch": 1.602967060598441,
      "grad_norm": 10.058784484863281,
      "learning_rate": 2.3283882323359317e-05,
      "loss": 1.687,
      "step": 25500
    },
    {
      "epoch": 1.6343977872768418,
      "grad_norm": 8.863463401794434,
      "learning_rate": 2.2760036878719305e-05,
      "loss": 1.6844,
      "step": 26000
    },
    {
      "epoch": 1.6658285139552427,
      "grad_norm": 10.471988677978516,
      "learning_rate": 2.223619143407929e-05,
      "loss": 1.7151,
      "step": 26500
    },
    {
      "epoch": 1.6972592406336435,
      "grad_norm": 9.061573028564453,
      "learning_rate": 2.1712345989439276e-05,
      "loss": 1.5916,
      "step": 27000
    },
    {
      "epoch": 1.7286899673120444,
      "grad_norm": 8.41381549835205,
      "learning_rate": 2.1188500544799264e-05,
      "loss": 1.6388,
      "step": 27500
    },
    {
      "epoch": 1.760120693990445,
      "grad_norm": 9.17684268951416,
      "learning_rate": 2.066465510015925e-05,
      "loss": 1.7045,
      "step": 28000
    },
    {
      "epoch": 1.7915514206688459,
      "grad_norm": 10.983583450317383,
      "learning_rate": 2.0140809655519235e-05,
      "loss": 1.6749,
      "step": 28500
    },
    {
      "epoch": 1.8229821473472465,
      "grad_norm": 8.048336029052734,
      "learning_rate": 1.9616964210879224e-05,
      "loss": 1.6347,
      "step": 29000
    },
    {
      "epoch": 1.8544128740256474,
      "grad_norm": 9.640363693237305,
      "learning_rate": 1.909311876623921e-05,
      "loss": 1.6424,
      "step": 29500
    },
    {
      "epoch": 1.8858436007040482,
      "grad_norm": 10.264333724975586,
      "learning_rate": 1.8569273321599194e-05,
      "loss": 1.5919,
      "step": 30000
    },
    {
      "epoch": 1.917274327382449,
      "grad_norm": 10.100754737854004,
      "learning_rate": 1.8045427876959183e-05,
      "loss": 1.6477,
      "step": 30500
    },
    {
      "epoch": 1.94870505406085,
      "grad_norm": 7.901167392730713,
      "learning_rate": 1.752158243231917e-05,
      "loss": 1.636,
      "step": 31000
    },
    {
      "epoch": 1.9801357807392508,
      "grad_norm": 10.34730339050293,
      "learning_rate": 1.6997736987679156e-05,
      "loss": 1.639,
      "step": 31500
    },
    {
      "epoch": 2.0115665074176516,
      "grad_norm": 9.00157642364502,
      "learning_rate": 1.6473891543039145e-05,
      "loss": 1.6027,
      "step": 32000
    },
    {
      "epoch": 2.0429972340960525,
      "grad_norm": 8.343816757202148,
      "learning_rate": 1.595004609839913e-05,
      "loss": 1.5728,
      "step": 32500
    },
    {
      "epoch": 2.074427960774453,
      "grad_norm": 9.560211181640625,
      "learning_rate": 1.5426200653759116e-05,
      "loss": 1.5924,
      "step": 33000
    },
    {
      "epoch": 2.1058586874528538,
      "grad_norm": 9.660225868225098,
      "learning_rate": 1.49023552091191e-05,
      "loss": 1.5557,
      "step": 33500
    },
    {
      "epoch": 2.1372894141312546,
      "grad_norm": 10.864439964294434,
      "learning_rate": 1.437850976447909e-05,
      "loss": 1.5636,
      "step": 34000
    },
    {
      "epoch": 2.1687201408096555,
      "grad_norm": 9.609240531921387,
      "learning_rate": 1.3854664319839075e-05,
      "loss": 1.5643,
      "step": 34500
    },
    {
      "epoch": 2.2001508674880563,
      "grad_norm": 9.839557647705078,
      "learning_rate": 1.3330818875199062e-05,
      "loss": 1.5742,
      "step": 35000
    },
    {
      "epoch": 2.231581594166457,
      "grad_norm": 9.829071998596191,
      "learning_rate": 1.280697343055905e-05,
      "loss": 1.551,
      "step": 35500
    },
    {
      "epoch": 2.263012320844858,
      "grad_norm": 8.181549072265625,
      "learning_rate": 1.2283127985919035e-05,
      "loss": 1.5712,
      "step": 36000
    },
    {
      "epoch": 2.294443047523259,
      "grad_norm": 7.759120464324951,
      "learning_rate": 1.1759282541279022e-05,
      "loss": 1.5383,
      "step": 36500
    },
    {
      "epoch": 2.3258737742016597,
      "grad_norm": 8.53459644317627,
      "learning_rate": 1.1235437096639008e-05,
      "loss": 1.5472,
      "step": 37000
    },
    {
      "epoch": 2.35730450088006,
      "grad_norm": 9.845288276672363,
      "learning_rate": 1.0711591651998994e-05,
      "loss": 1.5299,
      "step": 37500
    },
    {
      "epoch": 2.388735227558461,
      "grad_norm": 9.097940444946289,
      "learning_rate": 1.0187746207358981e-05,
      "loss": 1.5513,
      "step": 38000
    },
    {
      "epoch": 2.420165954236862,
      "grad_norm": 10.140700340270996,
      "learning_rate": 9.663900762718968e-06,
      "loss": 1.5314,
      "step": 38500
    },
    {
      "epoch": 2.4515966809152627,
      "grad_norm": 8.499448776245117,
      "learning_rate": 9.140055318078955e-06,
      "loss": 1.5381,
      "step": 39000
    },
    {
      "epoch": 2.4830274075936636,
      "grad_norm": 10.239831924438477,
      "learning_rate": 8.61620987343894e-06,
      "loss": 1.5338,
      "step": 39500
    },
    {
      "epoch": 2.5144581342720644,
      "grad_norm": 12.466906547546387,
      "learning_rate": 8.092364428798927e-06,
      "loss": 1.5361,
      "step": 40000
    },
    {
      "epoch": 2.5458888609504653,
      "grad_norm": 5.410431385040283,
      "learning_rate": 7.568518984158913e-06,
      "loss": 1.5506,
      "step": 40500
    },
    {
      "epoch": 2.5773195876288657,
      "grad_norm": 10.963549613952637,
      "learning_rate": 7.0446735395189e-06,
      "loss": 1.5383,
      "step": 41000
    },
    {
      "epoch": 2.608750314307267,
      "grad_norm": 9.227997779846191,
      "learning_rate": 6.520828094878888e-06,
      "loss": 1.5007,
      "step": 41500
    },
    {
      "epoch": 2.6401810409856674,
      "grad_norm": 9.551142692565918,
      "learning_rate": 5.996982650238874e-06,
      "loss": 1.5002,
      "step": 42000
    },
    {
      "epoch": 2.6716117676640683,
      "grad_norm": 9.357813835144043,
      "learning_rate": 5.47313720559886e-06,
      "loss": 1.521,
      "step": 42500
    },
    {
      "epoch": 2.703042494342469,
      "grad_norm": 9.11950969696045,
      "learning_rate": 4.949291760958847e-06,
      "loss": 1.5389,
      "step": 43000
    },
    {
      "epoch": 2.73447322102087,
      "grad_norm": 10.955974578857422,
      "learning_rate": 4.425446316318833e-06,
      "loss": 1.5481,
      "step": 43500
    },
    {
      "epoch": 2.765903947699271,
      "grad_norm": 9.225030899047852,
      "learning_rate": 3.90160087167882e-06,
      "loss": 1.5181,
      "step": 44000
    },
    {
      "epoch": 2.7973346743776717,
      "grad_norm": 10.103915214538574,
      "learning_rate": 3.377755427038807e-06,
      "loss": 1.5013,
      "step": 44500
    },
    {
      "epoch": 2.8287654010560725,
      "grad_norm": 8.871588706970215,
      "learning_rate": 2.853909982398793e-06,
      "loss": 1.5107,
      "step": 45000
    },
    {
      "epoch": 2.860196127734473,
      "grad_norm": 8.05070686340332,
      "learning_rate": 2.3300645377587797e-06,
      "loss": 1.4924,
      "step": 45500
    },
    {
      "epoch": 2.8916268544128743,
      "grad_norm": 10.415308952331543,
      "learning_rate": 1.8062190931187664e-06,
      "loss": 1.4606,
      "step": 46000
    },
    {
      "epoch": 2.9230575810912747,
      "grad_norm": 8.8274507522583,
      "learning_rate": 1.2823736484787529e-06,
      "loss": 1.4867,
      "step": 46500
    },
    {
      "epoch": 2.9544883077696755,
      "grad_norm": 7.306589126586914,
      "learning_rate": 7.585282038387395e-07,
      "loss": 1.4848,
      "step": 47000
    },
    {
      "epoch": 2.9859190344480764,
      "grad_norm": 9.284761428833008,
      "learning_rate": 2.34682759198726e-07,
      "loss": 1.4519,
      "step": 47500
    },
    {
      "epoch": 3.0,
      "step": 47724,
      "total_flos": 2.514865055648973e+16,
      "train_loss": 1.8108088482396418,
      "train_runtime": 10130.0955,
      "train_samples_per_second": 9.422,
      "train_steps_per_second": 4.711
    }
  ],
  "logging_steps": 500,
  "max_steps": 47724,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.514865055648973e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}