{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 436,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02298190175237001,
      "grad_norm": 96.69097137451172,
      "learning_rate": 4.5454545454545457e-07,
      "loss": 3.5384,
      "step": 5
    },
    {
      "epoch": 0.04596380350474002,
      "grad_norm": 86.741455078125,
      "learning_rate": 1.0227272727272729e-06,
      "loss": 3.0854,
      "step": 10
    },
    {
      "epoch": 0.06894570525711002,
      "grad_norm": 19.997196197509766,
      "learning_rate": 1.590909090909091e-06,
      "loss": 1.6651,
      "step": 15
    },
    {
      "epoch": 0.09192760700948004,
      "grad_norm": 11.091686248779297,
      "learning_rate": 2.1590909090909092e-06,
      "loss": 1.0304,
      "step": 20
    },
    {
      "epoch": 0.11490950876185005,
      "grad_norm": 7.87872838973999,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 0.6555,
      "step": 25
    },
    {
      "epoch": 0.13789141051422005,
      "grad_norm": 7.603336334228516,
      "learning_rate": 3.2954545454545456e-06,
      "loss": 0.4236,
      "step": 30
    },
    {
      "epoch": 0.16087331226659007,
      "grad_norm": 5.494614601135254,
      "learning_rate": 3.863636363636364e-06,
      "loss": 0.2925,
      "step": 35
    },
    {
      "epoch": 0.18385521401896007,
      "grad_norm": 5.1593194007873535,
      "learning_rate": 4.4318181818181824e-06,
      "loss": 0.2794,
      "step": 40
    },
    {
      "epoch": 0.20683711577133007,
      "grad_norm": 8.133462905883789,
      "learning_rate": 5e-06,
      "loss": 0.2366,
      "step": 45
    },
    {
      "epoch": 0.2298190175237001,
      "grad_norm": 5.586953639984131,
      "learning_rate": 4.9979931287121e-06,
      "loss": 0.229,
      "step": 50
    },
    {
      "epoch": 0.25280091927607007,
      "grad_norm": 4.7029266357421875,
      "learning_rate": 4.9919757368742895e-06,
      "loss": 0.1635,
      "step": 55
    },
    {
      "epoch": 0.2757828210284401,
      "grad_norm": 3.571770429611206,
      "learning_rate": 4.981957485391297e-06,
      "loss": 0.1463,
      "step": 60
    },
    {
      "epoch": 0.2987647227808101,
      "grad_norm": 3.206266164779663,
      "learning_rate": 4.967954458536126e-06,
      "loss": 0.1275,
      "step": 65
    },
    {
      "epoch": 0.32174662453318015,
      "grad_norm": 9.526619911193848,
      "learning_rate": 4.949989138126809e-06,
      "loss": 0.1459,
      "step": 70
    },
    {
      "epoch": 0.3447285262855501,
      "grad_norm": 3.7345776557922363,
      "learning_rate": 4.92809036743191e-06,
      "loss": 0.1433,
      "step": 75
    },
    {
      "epoch": 0.36771042803792015,
      "grad_norm": 2.7031843662261963,
      "learning_rate": 4.9022933048627496e-06,
      "loss": 0.1221,
      "step": 80
    },
    {
      "epoch": 0.39069232979029017,
      "grad_norm": 2.0503172874450684,
      "learning_rate": 4.872639367526672e-06,
      "loss": 0.085,
      "step": 85
    },
    {
      "epoch": 0.41367423154266014,
      "grad_norm": 1.5840656757354736,
      "learning_rate": 4.839176164732009e-06,
      "loss": 0.0968,
      "step": 90
    },
    {
      "epoch": 0.43665613329503017,
      "grad_norm": 1.4990031719207764,
      "learning_rate": 4.8019574215514705e-06,
      "loss": 0.0835,
      "step": 95
    },
    {
      "epoch": 0.4596380350474002,
      "grad_norm": 3.086148977279663,
      "learning_rate": 4.761042892566707e-06,
      "loss": 0.1249,
      "step": 100
    },
    {
      "epoch": 0.48261993679977017,
      "grad_norm": 1.3277251720428467,
      "learning_rate": 4.716498265932501e-06,
      "loss": 0.0705,
      "step": 105
    },
    {
      "epoch": 0.5056018385521401,
      "grad_norm": 2.104665517807007,
      "learning_rate": 4.668395057914627e-06,
      "loss": 0.0982,
      "step": 110
    },
    {
      "epoch": 0.5285837403045102,
      "grad_norm": 2.916086196899414,
      "learning_rate": 4.6168104980707105e-06,
      "loss": 0.0816,
      "step": 115
    },
    {
      "epoch": 0.5515656420568802,
      "grad_norm": 2.0353832244873047,
      "learning_rate": 4.56182740525839e-06,
      "loss": 0.0993,
      "step": 120
    },
    {
      "epoch": 0.5745475438092502,
      "grad_norm": 2.497854709625244,
      "learning_rate": 4.5035340546698915e-06,
      "loss": 0.1192,
      "step": 125
    },
    {
      "epoch": 0.5975294455616202,
      "grad_norm": 3.4010169506073,
      "learning_rate": 4.442024036106476e-06,
      "loss": 0.1433,
      "step": 130
    },
    {
      "epoch": 0.6205113473139903,
      "grad_norm": 2.062610149383545,
      "learning_rate": 4.3773961037202784e-06,
      "loss": 0.0784,
      "step": 135
    },
    {
      "epoch": 0.6434932490663603,
      "grad_norm": 1.2456505298614502,
      "learning_rate": 4.309754017464823e-06,
      "loss": 0.109,
      "step": 140
    },
    {
      "epoch": 0.6664751508187302,
      "grad_norm": 1.6870588064193726,
      "learning_rate": 4.239206376508716e-06,
      "loss": 0.0864,
      "step": 145
    },
    {
      "epoch": 0.6894570525711002,
      "grad_norm": 2.379024028778076,
      "learning_rate": 4.1658664448800105e-06,
      "loss": 0.0895,
      "step": 150
    },
    {
      "epoch": 0.7124389543234703,
      "grad_norm": 1.9115976095199585,
      "learning_rate": 4.089851969621138e-06,
      "loss": 0.0764,
      "step": 155
    },
    {
      "epoch": 0.7354208560758403,
      "grad_norm": 1.4540350437164307,
      "learning_rate": 4.0112849917463905e-06,
      "loss": 0.0719,
      "step": 160
    },
    {
      "epoch": 0.7584027578282103,
      "grad_norm": 2.1586649417877197,
      "learning_rate": 3.930291650305424e-06,
      "loss": 0.1243,
      "step": 165
    },
    {
      "epoch": 0.7813846595805803,
      "grad_norm": 2.2373526096343994,
      "learning_rate": 3.847001979867401e-06,
      "loss": 0.1274,
      "step": 170
    },
    {
      "epoch": 0.8043665613329503,
      "grad_norm": 2.466942548751831,
      "learning_rate": 3.761549701750865e-06,
      "loss": 0.0962,
      "step": 175
    },
    {
      "epoch": 0.8273484630853203,
      "grad_norm": 1.5502763986587524,
      "learning_rate": 3.6740720093345646e-06,
      "loss": 0.0872,
      "step": 180
    },
    {
      "epoch": 0.8503303648376903,
      "grad_norm": 2.0481746196746826,
      "learning_rate": 3.5847093477938955e-06,
      "loss": 0.1002,
      "step": 185
    },
    {
      "epoch": 0.8733122665900603,
      "grad_norm": 2.5169246196746826,
      "learning_rate": 3.4936051886165825e-06,
      "loss": 0.0878,
      "step": 190
    },
    {
      "epoch": 0.8962941683424304,
      "grad_norm": 1.0514692068099976,
      "learning_rate": 3.400905799259634e-06,
      "loss": 0.0779,
      "step": 195
    },
    {
      "epoch": 0.9192760700948004,
      "grad_norm": 1.217383623123169,
      "learning_rate": 3.3067600083173752e-06,
      "loss": 0.0839,
      "step": 200
    },
    {
      "epoch": 0.9422579718471703,
      "grad_norm": 0.7383676171302795,
      "learning_rate": 3.2113189665775812e-06,
      "loss": 0.0763,
      "step": 205
    },
    {
      "epoch": 0.9652398735995403,
      "grad_norm": 2.306929349899292,
      "learning_rate": 3.1147359043493365e-06,
      "loss": 0.0759,
      "step": 210
    },
    {
      "epoch": 0.9882217753519104,
      "grad_norm": 1.173736333847046,
      "learning_rate": 3.0171658854522274e-06,
      "loss": 0.0698,
      "step": 215
    },
    {
      "epoch": 1.009192760700948,
      "grad_norm": 1.4674370288848877,
      "learning_rate": 2.9187655582618413e-06,
      "loss": 0.064,
      "step": 220
    },
    {
      "epoch": 1.032174662453318,
      "grad_norm": 2.086500644683838,
      "learning_rate": 2.8196929042112652e-06,
      "loss": 0.0552,
      "step": 225
    },
    {
      "epoch": 1.055156564205688,
      "grad_norm": 0.8511289358139038,
      "learning_rate": 2.720106984152364e-06,
      "loss": 0.0525,
      "step": 230
    },
    {
      "epoch": 1.078138465958058,
      "grad_norm": 1.8033392429351807,
      "learning_rate": 2.620167682984052e-06,
      "loss": 0.063,
      "step": 235
    },
    {
      "epoch": 1.101120367710428,
      "grad_norm": 1.5028774738311768,
      "learning_rate": 2.5200354529575693e-06,
      "loss": 0.0544,
      "step": 240
    },
    {
      "epoch": 1.124102269462798,
      "grad_norm": 0.8938283324241638,
      "learning_rate": 2.4198710560708623e-06,
      "loss": 0.0763,
      "step": 245
    },
    {
      "epoch": 1.147084171215168,
      "grad_norm": 0.780461847782135,
      "learning_rate": 2.3198353059656763e-06,
      "loss": 0.0573,
      "step": 250
    },
    {
      "epoch": 1.170066072967538,
      "grad_norm": 0.74661785364151,
      "learning_rate": 2.2200888097417308e-06,
      "loss": 0.0471,
      "step": 255
    },
    {
      "epoch": 1.193047974719908,
      "grad_norm": 0.7056024074554443,
      "learning_rate": 2.120791710102499e-06,
      "loss": 0.0547,
      "step": 260
    },
    {
      "epoch": 1.2160298764722781,
      "grad_norm": 1.277025818824768,
      "learning_rate": 2.02210342824657e-06,
      "loss": 0.0569,
      "step": 265
    },
    {
      "epoch": 1.2390117782246481,
      "grad_norm": 0.9659903645515442,
      "learning_rate": 1.924182407917391e-06,
      "loss": 0.0566,
      "step": 270
    },
    {
      "epoch": 1.2619936799770182,
      "grad_norm": 0.9990319609642029,
      "learning_rate": 1.827185861022308e-06,
      "loss": 0.0614,
      "step": 275
    },
    {
      "epoch": 1.284975581729388,
      "grad_norm": 0.835111677646637,
      "learning_rate": 1.7312695152293112e-06,
      "loss": 0.0488,
      "step": 280
    },
    {
      "epoch": 1.3079574834817582,
      "grad_norm": 1.5359407663345337,
      "learning_rate": 1.6365873639467314e-06,
      "loss": 0.0606,
      "step": 285
    },
    {
      "epoch": 1.330939385234128,
      "grad_norm": 1.4668985605239868,
      "learning_rate": 1.5432914190872757e-06,
      "loss": 0.0569,
      "step": 290
    },
    {
      "epoch": 1.353921286986498,
      "grad_norm": 1.5932097434997559,
      "learning_rate": 1.4515314670133582e-06,
      "loss": 0.0543,
      "step": 295
    },
    {
      "epoch": 1.376903188738868,
      "grad_norm": 2.2497987747192383,
      "learning_rate": 1.3614548280555351e-06,
      "loss": 0.065,
      "step": 300
    },
    {
      "epoch": 1.399885090491238,
      "grad_norm": 1.6525359153747559,
      "learning_rate": 1.2732061199901563e-06,
      "loss": 0.0597,
      "step": 305
    },
    {
      "epoch": 1.4228669922436081,
      "grad_norm": 1.8272337913513184,
      "learning_rate": 1.1869270258559477e-06,
      "loss": 0.0479,
      "step": 310
    },
    {
      "epoch": 1.4458488939959782,
      "grad_norm": 1.1793828010559082,
      "learning_rate": 1.1027560664823208e-06,
      "loss": 0.0589,
      "step": 315
    },
    {
      "epoch": 1.4688307957483482,
      "grad_norm": 1.8448723554611206,
      "learning_rate": 1.020828378094592e-06,
      "loss": 0.0597,
      "step": 320
    },
    {
      "epoch": 1.4918126975007182,
      "grad_norm": 1.3766109943389893,
      "learning_rate": 9.412754953531664e-07,
      "loss": 0.0569,
      "step": 325
    },
    {
      "epoch": 1.5147945992530882,
      "grad_norm": 0.7711452841758728,
      "learning_rate": 8.642251401750395e-07,
      "loss": 0.0595,
      "step": 330
    },
    {
      "epoch": 1.537776501005458,
      "grad_norm": 1.5899217128753662,
      "learning_rate": 7.898010166766348e-07,
      "loss": 0.0592,
      "step": 335
    },
    {
      "epoch": 1.5607584027578283,
      "grad_norm": 0.8444715738296509,
      "learning_rate": 7.181226125672142e-07,
      "loss": 0.0504,
      "step": 340
    },
    {
      "epoch": 1.583740304510198,
      "grad_norm": 0.7058942914009094,
      "learning_rate": 6.493050073117115e-07,
      "loss": 0.0522,
      "step": 345
    },
    {
      "epoch": 1.6067222062625683,
      "grad_norm": 2.4318106174468994,
      "learning_rate": 5.83458687371e-07,
      "loss": 0.0597,
      "step": 350
    },
    {
      "epoch": 1.6297041080149381,
      "grad_norm": 1.4847302436828613,
      "learning_rate": 5.20689368816201e-07,
      "loss": 0.0536,
      "step": 355
    },
    {
      "epoch": 1.6526860097673084,
      "grad_norm": 0.6792570948600769,
      "learning_rate": 4.610978276018496e-07,
      "loss": 0.051,
      "step": 360
    },
    {
      "epoch": 1.6756679115196782,
      "grad_norm": 1.1941633224487305,
      "learning_rate": 4.0477973777039854e-07,
      "loss": 0.0489,
      "step": 365
    },
    {
      "epoch": 1.6986498132720482,
      "grad_norm": 1.1347832679748535,
      "learning_rate": 3.518255178478247e-07,
      "loss": 0.0438,
      "step": 370
    },
    {
      "epoch": 1.7216317150244183,
      "grad_norm": 0.7465569376945496,
      "learning_rate": 3.02320185676957e-07,
      "loss": 0.053,
      "step": 375
    },
    {
      "epoch": 1.7446136167767883,
      "grad_norm": 0.7349793910980225,
      "learning_rate": 2.563432219215814e-07,
      "loss": 0.0455,
      "step": 380
    },
    {
      "epoch": 1.7675955185291583,
      "grad_norm": 1.828784465789795,
      "learning_rate": 2.1396844246046904e-07,
      "loss": 0.0506,
      "step": 385
    },
    {
      "epoch": 1.7905774202815283,
      "grad_norm": 1.2421413660049438,
      "learning_rate": 1.7526387987620602e-07,
      "loss": 0.0444,
      "step": 390
    },
    {
      "epoch": 1.8135593220338984,
      "grad_norm": 0.9201889634132385,
      "learning_rate": 1.4029167422908107e-07,
      "loss": 0.0545,
      "step": 395
    },
    {
      "epoch": 1.8365412237862682,
      "grad_norm": 0.8820809125900269,
      "learning_rate": 1.0910797329140466e-07,
      "loss": 0.054,
      "step": 400
    },
    {
      "epoch": 1.8595231255386384,
      "grad_norm": 1.2385250329971313,
      "learning_rate": 8.176284240242638e-08,
      "loss": 0.0563,
      "step": 405
    },
    {
      "epoch": 1.8825050272910082,
      "grad_norm": 0.8065256476402283,
      "learning_rate": 5.830018408858246e-08,
      "loss": 0.0468,
      "step": 410
    },
    {
      "epoch": 1.9054869290433785,
      "grad_norm": 1.8467743396759033,
      "learning_rate": 3.8757667578119e-08,
      "loss": 0.0549,
      "step": 415
    },
    {
      "epoch": 1.9284688307957483,
      "grad_norm": 0.7403564453125,
      "learning_rate": 2.31666683232587e-08,
      "loss": 0.0559,
      "step": 420
    },
    {
      "epoch": 1.9514507325481185,
      "grad_norm": 0.8806583881378174,
      "learning_rate": 1.1552217627004426e-08,
      "loss": 0.047,
      "step": 425
    },
    {
      "epoch": 1.9744326343004883,
      "grad_norm": 0.663443922996521,
      "learning_rate": 3.932962455458489e-09,
      "loss": 0.0504,
      "step": 430
    },
    {
      "epoch": 1.9974145360528583,
      "grad_norm": 1.0851346254348755,
      "learning_rate": 3.2113550017198734e-10,
      "loss": 0.0525,
      "step": 435
    },
    {
      "epoch": 2.0,
      "step": 436,
      "total_flos": 1.0467157525694669e+18,
      "train_loss": 0.19755241828500678,
      "train_runtime": 9435.8558,
      "train_samples_per_second": 2.951,
      "train_steps_per_second": 0.046
    }
  ],
  "logging_steps": 5,
  "max_steps": 436,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0467157525694669e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}