{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 14109,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0425260472039124,
      "grad_norm": 4.5295610427856445,
      "learning_rate": 4.929123254660146e-05,
      "loss": 3.196,
      "step": 200
    },
    {
      "epoch": 0.0850520944078248,
      "grad_norm": 3.153637170791626,
      "learning_rate": 4.858246509320292e-05,
      "loss": 2.0357,
      "step": 400
    },
    {
      "epoch": 0.1275781416117372,
      "grad_norm": 2.833491802215576,
      "learning_rate": 4.7873697639804386e-05,
      "loss": 1.9407,
      "step": 600
    },
    {
      "epoch": 0.1701041888156496,
      "grad_norm": 3.3437952995300293,
      "learning_rate": 4.7164930186405845e-05,
      "loss": 1.8486,
      "step": 800
    },
    {
      "epoch": 0.212630236019562,
      "grad_norm": 2.2692761421203613,
      "learning_rate": 4.64561627330073e-05,
      "loss": 1.8338,
      "step": 1000
    },
    {
      "epoch": 0.2551562832234744,
      "grad_norm": 9.267280578613281,
      "learning_rate": 4.574739527960876e-05,
      "loss": 1.8259,
      "step": 1200
    },
    {
      "epoch": 0.2976823304273868,
      "grad_norm": 2.5437700748443604,
      "learning_rate": 4.503862782621022e-05,
      "loss": 1.7474,
      "step": 1400
    },
    {
      "epoch": 0.3402083776312992,
      "grad_norm": 3.454941511154175,
      "learning_rate": 4.432986037281168e-05,
      "loss": 1.7185,
      "step": 1600
    },
    {
      "epoch": 0.3827344248352116,
      "grad_norm": 2.709567070007324,
      "learning_rate": 4.3621092919413145e-05,
      "loss": 1.7468,
      "step": 1800
    },
    {
      "epoch": 0.425260472039124,
      "grad_norm": 2.4916956424713135,
      "learning_rate": 4.2912325466014604e-05,
      "loss": 1.7277,
      "step": 2000
    },
    {
      "epoch": 0.4677865192430364,
      "grad_norm": 2.3599932193756104,
      "learning_rate": 4.220355801261606e-05,
      "loss": 1.6742,
      "step": 2200
    },
    {
      "epoch": 0.5103125664469488,
      "grad_norm": 2.518554449081421,
      "learning_rate": 4.149479055921752e-05,
      "loss": 1.6232,
      "step": 2400
    },
    {
      "epoch": 0.5528386136508612,
      "grad_norm": 2.202854633331299,
      "learning_rate": 4.078602310581898e-05,
      "loss": 1.6431,
      "step": 2600
    },
    {
      "epoch": 0.5953646608547736,
      "grad_norm": 2.0834789276123047,
      "learning_rate": 4.007725565242044e-05,
      "loss": 1.6275,
      "step": 2800
    },
    {
      "epoch": 0.637890708058686,
      "grad_norm": 1.998606562614441,
      "learning_rate": 3.9368488199021905e-05,
      "loss": 1.6399,
      "step": 3000
    },
    {
      "epoch": 0.6804167552625984,
      "grad_norm": 2.2863314151763916,
      "learning_rate": 3.8659720745623364e-05,
      "loss": 1.631,
      "step": 3200
    },
    {
      "epoch": 0.7229428024665108,
      "grad_norm": 2.4040656089782715,
      "learning_rate": 3.795095329222482e-05,
      "loss": 1.5873,
      "step": 3400
    },
    {
      "epoch": 0.7654688496704232,
      "grad_norm": 2.490069627761841,
      "learning_rate": 3.724218583882628e-05,
      "loss": 1.6236,
      "step": 3600
    },
    {
      "epoch": 0.8079948968743356,
      "grad_norm": 2.4205753803253174,
      "learning_rate": 3.653341838542774e-05,
      "loss": 1.57,
      "step": 3800
    },
    {
      "epoch": 0.850520944078248,
      "grad_norm": 3.0913541316986084,
      "learning_rate": 3.58246509320292e-05,
      "loss": 1.5535,
      "step": 4000
    },
    {
      "epoch": 0.8930469912821604,
      "grad_norm": 2.7764530181884766,
      "learning_rate": 3.5115883478630665e-05,
      "loss": 1.6216,
      "step": 4200
    },
    {
      "epoch": 0.9355730384860728,
      "grad_norm": 2.6188108921051025,
      "learning_rate": 3.4407116025232124e-05,
      "loss": 1.5819,
      "step": 4400
    },
    {
      "epoch": 0.9780990856899852,
      "grad_norm": 2.5243659019470215,
      "learning_rate": 3.369834857183358e-05,
      "loss": 1.5277,
      "step": 4600
    },
    {
      "epoch": 1.0206251328938976,
      "grad_norm": 2.1120309829711914,
      "learning_rate": 3.298958111843504e-05,
      "loss": 1.5052,
      "step": 4800
    },
    {
      "epoch": 1.06315118009781,
      "grad_norm": 2.532235622406006,
      "learning_rate": 3.22808136650365e-05,
      "loss": 1.4648,
      "step": 5000
    },
    {
      "epoch": 1.1056772273017224,
      "grad_norm": 2.924020767211914,
      "learning_rate": 3.157204621163796e-05,
      "loss": 1.4526,
      "step": 5200
    },
    {
      "epoch": 1.1482032745056348,
      "grad_norm": 3.1454060077667236,
      "learning_rate": 3.0863278758239425e-05,
      "loss": 1.4659,
      "step": 5400
    },
    {
      "epoch": 1.1907293217095472,
      "grad_norm": 2.3973867893218994,
      "learning_rate": 3.0154511304840884e-05,
      "loss": 1.4622,
      "step": 5600
    },
    {
      "epoch": 1.2332553689134595,
      "grad_norm": 2.5608162879943848,
      "learning_rate": 2.9445743851442343e-05,
      "loss": 1.4919,
      "step": 5800
    },
    {
      "epoch": 1.275781416117372,
      "grad_norm": 1.8345551490783691,
      "learning_rate": 2.87369763980438e-05,
      "loss": 1.4644,
      "step": 6000
    },
    {
      "epoch": 1.3183074633212843,
      "grad_norm": 2.1043267250061035,
      "learning_rate": 2.8028208944645264e-05,
      "loss": 1.4662,
      "step": 6200
    },
    {
      "epoch": 1.3608335105251967,
      "grad_norm": 1.9993553161621094,
      "learning_rate": 2.7319441491246722e-05,
      "loss": 1.4558,
      "step": 6400
    },
    {
      "epoch": 1.4033595577291091,
      "grad_norm": 2.4795050621032715,
      "learning_rate": 2.661067403784818e-05,
      "loss": 1.4381,
      "step": 6600
    },
    {
      "epoch": 1.4458856049330215,
      "grad_norm": 2.0630886554718018,
      "learning_rate": 2.5901906584449644e-05,
      "loss": 1.4244,
      "step": 6800
    },
    {
      "epoch": 1.488411652136934,
      "grad_norm": 2.489701509475708,
      "learning_rate": 2.5193139131051102e-05,
      "loss": 1.4263,
      "step": 7000
    },
    {
      "epoch": 1.5309376993408463,
      "grad_norm": 2.2150392532348633,
      "learning_rate": 2.448437167765256e-05,
      "loss": 1.4437,
      "step": 7200
    },
    {
      "epoch": 1.5734637465447587,
      "grad_norm": 2.3309507369995117,
      "learning_rate": 2.3775604224254023e-05,
      "loss": 1.437,
      "step": 7400
    },
    {
      "epoch": 1.6159897937486711,
      "grad_norm": 2.4965591430664062,
      "learning_rate": 2.3066836770855482e-05,
      "loss": 1.4509,
      "step": 7600
    },
    {
      "epoch": 1.6585158409525835,
      "grad_norm": 2.379479169845581,
      "learning_rate": 2.235806931745694e-05,
      "loss": 1.3942,
      "step": 7800
    },
    {
      "epoch": 1.701041888156496,
      "grad_norm": 2.083732843399048,
      "learning_rate": 2.1649301864058403e-05,
      "loss": 1.3781,
      "step": 8000
    },
    {
      "epoch": 1.7435679353604083,
      "grad_norm": 2.258941411972046,
      "learning_rate": 2.0940534410659862e-05,
      "loss": 1.3947,
      "step": 8200
    },
    {
      "epoch": 1.7860939825643207,
      "grad_norm": 2.069312334060669,
      "learning_rate": 2.023176695726132e-05,
      "loss": 1.448,
      "step": 8400
    },
    {
      "epoch": 1.8286200297682331,
      "grad_norm": 2.0143864154815674,
      "learning_rate": 1.9522999503862783e-05,
      "loss": 1.387,
      "step": 8600
    },
    {
      "epoch": 1.8711460769721455,
      "grad_norm": 2.7671608924865723,
      "learning_rate": 1.8814232050464242e-05,
      "loss": 1.4122,
      "step": 8800
    },
    {
      "epoch": 1.913672124176058,
      "grad_norm": 2.4909543991088867,
      "learning_rate": 1.8105464597065704e-05,
      "loss": 1.4364,
      "step": 9000
    },
    {
      "epoch": 1.9561981713799703,
      "grad_norm": 2.471282720565796,
      "learning_rate": 1.7396697143667166e-05,
      "loss": 1.4177,
      "step": 9200
    },
    {
      "epoch": 1.9987242185838827,
      "grad_norm": 3.3425190448760986,
      "learning_rate": 1.6687929690268625e-05,
      "loss": 1.4231,
      "step": 9400
    },
    {
      "epoch": 2.041250265787795,
      "grad_norm": 2.7562530040740967,
      "learning_rate": 1.5979162236870084e-05,
      "loss": 1.3759,
      "step": 9600
    },
    {
      "epoch": 2.0837763129917075,
      "grad_norm": 2.857525587081909,
      "learning_rate": 1.5270394783471546e-05,
      "loss": 1.3467,
      "step": 9800
    },
    {
      "epoch": 2.12630236019562,
      "grad_norm": 2.1692278385162354,
      "learning_rate": 1.4561627330073005e-05,
      "loss": 1.3397,
      "step": 10000
    },
    {
      "epoch": 2.1688284073995323,
      "grad_norm": 2.8106327056884766,
      "learning_rate": 1.3852859876674466e-05,
      "loss": 1.3203,
      "step": 10200
    },
    {
      "epoch": 2.2113544546034447,
      "grad_norm": 2.509658098220825,
      "learning_rate": 1.3144092423275925e-05,
      "loss": 1.3108,
      "step": 10400
    },
    {
      "epoch": 2.253880501807357,
      "grad_norm": 2.2616686820983887,
      "learning_rate": 1.2435324969877383e-05,
      "loss": 1.3394,
      "step": 10600
    },
    {
      "epoch": 2.2964065490112695,
      "grad_norm": 2.6901018619537354,
      "learning_rate": 1.1726557516478844e-05,
      "loss": 1.3267,
      "step": 10800
    },
    {
      "epoch": 2.338932596215182,
      "grad_norm": 3.918454647064209,
      "learning_rate": 1.1017790063080303e-05,
      "loss": 1.3223,
      "step": 11000
    },
    {
      "epoch": 2.3814586434190943,
      "grad_norm": 2.4760806560516357,
      "learning_rate": 1.0309022609681763e-05,
      "loss": 1.3549,
      "step": 11200
    },
    {
      "epoch": 2.4239846906230067,
      "grad_norm": 2.3667821884155273,
      "learning_rate": 9.600255156283224e-06,
      "loss": 1.3398,
      "step": 11400
    },
    {
      "epoch": 2.466510737826919,
      "grad_norm": 2.600486993789673,
      "learning_rate": 8.891487702884683e-06,
      "loss": 1.3195,
      "step": 11600
    },
    {
      "epoch": 2.5090367850308315,
      "grad_norm": 2.0121264457702637,
      "learning_rate": 8.182720249486143e-06,
      "loss": 1.3308,
      "step": 11800
    },
    {
      "epoch": 2.551562832234744,
      "grad_norm": 2.0753839015960693,
      "learning_rate": 7.473952796087604e-06,
      "loss": 1.3607,
      "step": 12000
    },
    {
      "epoch": 2.5940888794386563,
      "grad_norm": 2.37463641166687,
      "learning_rate": 6.7651853426890634e-06,
      "loss": 1.3247,
      "step": 12200
    },
    {
      "epoch": 2.6366149266425687,
      "grad_norm": 2.016892910003662,
      "learning_rate": 6.056417889290524e-06,
      "loss": 1.2809,
      "step": 12400
    },
    {
      "epoch": 2.679140973846481,
      "grad_norm": 2.9185993671417236,
      "learning_rate": 5.3476504358919845e-06,
      "loss": 1.3227,
      "step": 12600
    },
    {
      "epoch": 2.7216670210503935,
      "grad_norm": 3.1016061305999756,
      "learning_rate": 4.638882982493444e-06,
      "loss": 1.3224,
      "step": 12800
    },
    {
      "epoch": 2.764193068254306,
      "grad_norm": 1.890191912651062,
      "learning_rate": 3.930115529094904e-06,
      "loss": 1.3527,
      "step": 13000
    },
    {
      "epoch": 2.8067191154582183,
      "grad_norm": 2.897411346435547,
      "learning_rate": 3.221348075696364e-06,
      "loss": 1.3423,
      "step": 13200
    },
    {
      "epoch": 2.8492451626621307,
      "grad_norm": 2.7592458724975586,
      "learning_rate": 2.512580622297824e-06,
      "loss": 1.3389,
      "step": 13400
    },
    {
      "epoch": 2.891771209866043,
      "grad_norm": 2.054417610168457,
      "learning_rate": 1.8038131688992842e-06,
      "loss": 1.3143,
      "step": 13600
    },
    {
      "epoch": 2.9342972570699555,
      "grad_norm": 2.7590255737304688,
      "learning_rate": 1.0950457155007443e-06,
      "loss": 1.3156,
      "step": 13800
    },
    {
      "epoch": 2.976823304273868,
      "grad_norm": 2.5642201900482178,
      "learning_rate": 3.862782621022043e-07,
      "loss": 1.3214,
      "step": 14000
    }
  ],
  "logging_steps": 200,
  "max_steps": 14109,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3685981372416000.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}