{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 4452,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.022461814914645103,
      "grad_norm": 28.0,
      "learning_rate": 1.977987421383648e-05,
      "loss": 0.906695556640625,
      "step": 50
    },
    {
      "epoch": 0.044923629829290206,
      "grad_norm": 22.125,
      "learning_rate": 1.955525606469003e-05,
      "loss": 0.678603515625,
      "step": 100
    },
    {
      "epoch": 0.0673854447439353,
      "grad_norm": 7.28125,
      "learning_rate": 1.9330637915543576e-05,
      "loss": 0.454794921875,
      "step": 150
    },
    {
      "epoch": 0.08984725965858041,
      "grad_norm": 13.1875,
      "learning_rate": 1.9106019766397127e-05,
      "loss": 0.497265625,
      "step": 200
    },
    {
      "epoch": 0.11230907457322552,
      "grad_norm": 8.0625,
      "learning_rate": 1.8881401617250677e-05,
      "loss": 0.489990234375,
      "step": 250
    },
    {
      "epoch": 0.1347708894878706,
      "grad_norm": 9.3125,
      "learning_rate": 1.8656783468104224e-05,
      "loss": 0.427421875,
      "step": 300
    },
    {
      "epoch": 0.15723270440251572,
      "grad_norm": 9.0,
      "learning_rate": 1.843216531895777e-05,
      "loss": 0.4686279296875,
      "step": 350
    },
    {
      "epoch": 0.17969451931716082,
      "grad_norm": 9.4375,
      "learning_rate": 1.8207547169811322e-05,
      "loss": 0.41837890625,
      "step": 400
    },
    {
      "epoch": 0.20215633423180593,
      "grad_norm": 9.9375,
      "learning_rate": 1.7982929020664873e-05,
      "loss": 0.419453125,
      "step": 450
    },
    {
      "epoch": 0.22461814914645103,
      "grad_norm": 15.875,
      "learning_rate": 1.775831087151842e-05,
      "loss": 0.3905078125,
      "step": 500
    },
    {
      "epoch": 0.24707996406109614,
      "grad_norm": 8.5625,
      "learning_rate": 1.753369272237197e-05,
      "loss": 0.43294921875,
      "step": 550
    },
    {
      "epoch": 0.2695417789757412,
      "grad_norm": 9.5,
      "learning_rate": 1.7309074573225517e-05,
      "loss": 0.41015625,
      "step": 600
    },
    {
      "epoch": 0.2920035938903863,
      "grad_norm": 7.125,
      "learning_rate": 1.7084456424079068e-05,
      "loss": 0.4262890625,
      "step": 650
    },
    {
      "epoch": 0.31446540880503143,
      "grad_norm": 15.3125,
      "learning_rate": 1.6859838274932615e-05,
      "loss": 0.39517578125,
      "step": 700
    },
    {
      "epoch": 0.33692722371967654,
      "grad_norm": 10.375,
      "learning_rate": 1.6635220125786166e-05,
      "loss": 0.389140625,
      "step": 750
    },
    {
      "epoch": 0.35938903863432164,
      "grad_norm": 8.75,
      "learning_rate": 1.6410601976639713e-05,
      "loss": 0.4083984375,
      "step": 800
    },
    {
      "epoch": 0.38185085354896675,
      "grad_norm": 11.0625,
      "learning_rate": 1.6185983827493263e-05,
      "loss": 0.4368359375,
      "step": 850
    },
    {
      "epoch": 0.40431266846361186,
      "grad_norm": 5.71875,
      "learning_rate": 1.596136567834681e-05,
      "loss": 0.36828125,
      "step": 900
    },
    {
      "epoch": 0.42677448337825696,
      "grad_norm": 53.0,
      "learning_rate": 1.573674752920036e-05,
      "loss": 0.3913671875,
      "step": 950
    },
    {
      "epoch": 0.44923629829290207,
      "grad_norm": 11.25,
      "learning_rate": 1.5512129380053908e-05,
      "loss": 0.38828125,
      "step": 1000
    },
    {
      "epoch": 0.4716981132075472,
      "grad_norm": 10.1875,
      "learning_rate": 1.528751123090746e-05,
      "loss": 0.40328125,
      "step": 1050
    },
    {
      "epoch": 0.4941599281221923,
      "grad_norm": 10.0,
      "learning_rate": 1.5062893081761007e-05,
      "loss": 0.389921875,
      "step": 1100
    },
    {
      "epoch": 0.5166217430368374,
      "grad_norm": 14.875,
      "learning_rate": 1.4838274932614556e-05,
      "loss": 0.434111328125,
      "step": 1150
    },
    {
      "epoch": 0.5390835579514824,
      "grad_norm": 9.5,
      "learning_rate": 1.4613656783468107e-05,
      "loss": 0.34328125,
      "step": 1200
    },
    {
      "epoch": 0.5615453728661276,
      "grad_norm": 9.375,
      "learning_rate": 1.4389038634321656e-05,
      "loss": 0.37349609375,
      "step": 1250
    },
    {
      "epoch": 0.5840071877807727,
      "grad_norm": 12.5,
      "learning_rate": 1.4164420485175203e-05,
      "loss": 0.3348828125,
      "step": 1300
    },
    {
      "epoch": 0.6064690026954178,
      "grad_norm": 18.875,
      "learning_rate": 1.3939802336028752e-05,
      "loss": 0.351328125,
      "step": 1350
    },
    {
      "epoch": 0.6289308176100629,
      "grad_norm": 8.875,
      "learning_rate": 1.3715184186882302e-05,
      "loss": 0.340703125,
      "step": 1400
    },
    {
      "epoch": 0.651392632524708,
      "grad_norm": 22.875,
      "learning_rate": 1.3490566037735851e-05,
      "loss": 0.357861328125,
      "step": 1450
    },
    {
      "epoch": 0.6738544474393531,
      "grad_norm": 9.375,
      "learning_rate": 1.3265947888589398e-05,
      "loss": 0.3385546875,
      "step": 1500
    },
    {
      "epoch": 0.6963162623539982,
      "grad_norm": 10.0,
      "learning_rate": 1.3041329739442947e-05,
      "loss": 0.3638671875,
      "step": 1550
    },
    {
      "epoch": 0.7187780772686433,
      "grad_norm": 41.75,
      "learning_rate": 1.2816711590296497e-05,
      "loss": 0.38921875,
      "step": 1600
    },
    {
      "epoch": 0.7412398921832885,
      "grad_norm": 13.1875,
      "learning_rate": 1.2592093441150046e-05,
      "loss": 0.3930078125,
      "step": 1650
    },
    {
      "epoch": 0.7637017070979335,
      "grad_norm": 11.5625,
      "learning_rate": 1.2367475292003595e-05,
      "loss": 0.39505859375,
      "step": 1700
    },
    {
      "epoch": 0.7861635220125787,
      "grad_norm": 12.0625,
      "learning_rate": 1.2142857142857142e-05,
      "loss": 0.3996875,
      "step": 1750
    },
    {
      "epoch": 0.8086253369272237,
      "grad_norm": 49.5,
      "learning_rate": 1.1918238993710693e-05,
      "loss": 0.3456640625,
      "step": 1800
    },
    {
      "epoch": 0.8310871518418689,
      "grad_norm": 20.75,
      "learning_rate": 1.1693620844564242e-05,
      "loss": 0.34216796875,
      "step": 1850
    },
    {
      "epoch": 0.8535489667565139,
      "grad_norm": 6.375,
      "learning_rate": 1.146900269541779e-05,
      "loss": 0.3304296875,
      "step": 1900
    },
    {
      "epoch": 0.876010781671159,
      "grad_norm": 11.0625,
      "learning_rate": 1.124438454627134e-05,
      "loss": 0.3637109375,
      "step": 1950
    },
    {
      "epoch": 0.8984725965858041,
      "grad_norm": 10.75,
      "learning_rate": 1.101976639712489e-05,
      "loss": 0.3279296875,
      "step": 2000
    },
    {
      "epoch": 0.9209344115004492,
      "grad_norm": 11.4375,
      "learning_rate": 1.0795148247978437e-05,
      "loss": 0.31841796875,
      "step": 2050
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 14.125,
      "learning_rate": 1.0570530098831986e-05,
      "loss": 0.39021484375,
      "step": 2100
    },
    {
      "epoch": 0.9658580413297394,
      "grad_norm": 14.75,
      "learning_rate": 1.0345911949685536e-05,
      "loss": 0.3907421875,
      "step": 2150
    },
    {
      "epoch": 0.9883198562443846,
      "grad_norm": 15.3125,
      "learning_rate": 1.0121293800539085e-05,
      "loss": 0.4207421875,
      "step": 2200
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.4922337234020233,
      "eval_runtime": 1.5467,
      "eval_samples_per_second": 218.526,
      "eval_steps_per_second": 7.112,
      "step": 2226
    },
    {
      "epoch": 1.0107816711590296,
      "grad_norm": 7.75,
      "learning_rate": 9.896675651392634e-06,
      "loss": 0.3505859375,
      "step": 2250
    },
    {
      "epoch": 1.0332434860736748,
      "grad_norm": 10.875,
      "learning_rate": 9.672057502246183e-06,
      "loss": 0.356328125,
      "step": 2300
    },
    {
      "epoch": 1.05570530098832,
      "grad_norm": 9.25,
      "learning_rate": 9.447439353099732e-06,
      "loss": 0.34708984375,
      "step": 2350
    },
    {
      "epoch": 1.0781671159029649,
      "grad_norm": 13.5625,
      "learning_rate": 9.22282120395328e-06,
      "loss": 0.39890625,
      "step": 2400
    },
    {
      "epoch": 1.10062893081761,
      "grad_norm": 10.375,
      "learning_rate": 8.99820305480683e-06,
      "loss": 0.3733984375,
      "step": 2450
    },
    {
      "epoch": 1.1230907457322552,
      "grad_norm": 14.5625,
      "learning_rate": 8.773584905660378e-06,
      "loss": 0.39263671875,
      "step": 2500
    },
    {
      "epoch": 1.1455525606469004,
      "grad_norm": 10.1875,
      "learning_rate": 8.548966756513927e-06,
      "loss": 0.34439453125,
      "step": 2550
    },
    {
      "epoch": 1.1680143755615453,
      "grad_norm": 12.0,
      "learning_rate": 8.324348607367476e-06,
      "loss": 0.3821484375,
      "step": 2600
    },
    {
      "epoch": 1.1904761904761905,
      "grad_norm": 15.3125,
      "learning_rate": 8.099730458221025e-06,
      "loss": 0.357578125,
      "step": 2650
    },
    {
      "epoch": 1.2129380053908356,
      "grad_norm": 11.5,
      "learning_rate": 7.875112309074573e-06,
      "loss": 0.348828125,
      "step": 2700
    },
    {
      "epoch": 1.2353998203054806,
      "grad_norm": 10.4375,
      "learning_rate": 7.650494159928122e-06,
      "loss": 0.35625,
      "step": 2750
    },
    {
      "epoch": 1.2578616352201257,
      "grad_norm": 9.9375,
      "learning_rate": 7.425876010781672e-06,
      "loss": 0.321953125,
      "step": 2800
    },
    {
      "epoch": 1.280323450134771,
      "grad_norm": 14.3125,
      "learning_rate": 7.20125786163522e-06,
      "loss": 0.33,
      "step": 2850
    },
    {
      "epoch": 1.302785265049416,
      "grad_norm": 11.375,
      "learning_rate": 6.97663971248877e-06,
      "loss": 0.38716796875,
      "step": 2900
    },
    {
      "epoch": 1.3252470799640612,
      "grad_norm": 6.65625,
      "learning_rate": 6.752021563342318e-06,
      "loss": 0.386923828125,
      "step": 2950
    },
    {
      "epoch": 1.3477088948787062,
      "grad_norm": 10.5,
      "learning_rate": 6.527403414195867e-06,
      "loss": 0.347265625,
      "step": 3000
    },
    {
      "epoch": 1.3701707097933513,
      "grad_norm": 9.9375,
      "learning_rate": 6.302785265049417e-06,
      "loss": 0.37748046875,
      "step": 3050
    },
    {
      "epoch": 1.3926325247079965,
      "grad_norm": 14.3125,
      "learning_rate": 6.078167115902965e-06,
      "loss": 0.3586328125,
      "step": 3100
    },
    {
      "epoch": 1.4150943396226414,
      "grad_norm": 17.375,
      "learning_rate": 5.853548966756515e-06,
      "loss": 0.3994140625,
      "step": 3150
    },
    {
      "epoch": 1.4375561545372866,
      "grad_norm": 10.125,
      "learning_rate": 5.6289308176100635e-06,
      "loss": 0.3518359375,
      "step": 3200
    },
    {
      "epoch": 1.4600179694519317,
      "grad_norm": 10.0,
      "learning_rate": 5.404312668463612e-06,
      "loss": 0.3810546875,
      "step": 3250
    },
    {
      "epoch": 1.482479784366577,
      "grad_norm": 12.375,
      "learning_rate": 5.179694519317161e-06,
      "loss": 0.3007421875,
      "step": 3300
    },
    {
      "epoch": 1.504941599281222,
      "grad_norm": 12.0,
      "learning_rate": 4.95507637017071e-06,
      "loss": 0.377421875,
      "step": 3350
    },
    {
      "epoch": 1.527403414195867,
      "grad_norm": 12.0,
      "learning_rate": 4.730458221024259e-06,
      "loss": 0.339921875,
      "step": 3400
    },
    {
      "epoch": 1.5498652291105122,
      "grad_norm": 20.125,
      "learning_rate": 4.5058400718778085e-06,
      "loss": 0.381875,
      "step": 3450
    },
    {
      "epoch": 1.5723270440251573,
      "grad_norm": 11.875,
      "learning_rate": 4.281221922731357e-06,
      "loss": 0.35994140625,
      "step": 3500
    },
    {
      "epoch": 1.5947888589398023,
      "grad_norm": 13.5,
      "learning_rate": 4.056603773584906e-06,
      "loss": 0.3158203125,
      "step": 3550
    },
    {
      "epoch": 1.6172506738544474,
      "grad_norm": 10.3125,
      "learning_rate": 3.831985624438455e-06,
      "loss": 0.3405078125,
      "step": 3600
    },
    {
      "epoch": 1.6397124887690926,
      "grad_norm": 14.8125,
      "learning_rate": 3.607367475292004e-06,
      "loss": 0.3773046875,
      "step": 3650
    },
    {
      "epoch": 1.6621743036837375,
      "grad_norm": 9.5625,
      "learning_rate": 3.382749326145553e-06,
      "loss": 0.36173828125,
      "step": 3700
    },
    {
      "epoch": 1.684636118598383,
      "grad_norm": 16.25,
      "learning_rate": 3.158131176999102e-06,
      "loss": 0.388466796875,
      "step": 3750
    },
    {
      "epoch": 1.7070979335130279,
      "grad_norm": 11.375,
      "learning_rate": 2.9335130278526508e-06,
      "loss": 0.346953125,
      "step": 3800
    },
    {
      "epoch": 1.7295597484276728,
      "grad_norm": 11.3125,
      "learning_rate": 2.7088948787061996e-06,
      "loss": 0.3610546875,
      "step": 3850
    },
    {
      "epoch": 1.7520215633423182,
      "grad_norm": 7.4375,
      "learning_rate": 2.4842767295597484e-06,
      "loss": 0.3298046875,
      "step": 3900
    },
    {
      "epoch": 1.7744833782569631,
      "grad_norm": 14.5625,
      "learning_rate": 2.2596585804132977e-06,
      "loss": 0.3954296875,
      "step": 3950
    },
    {
      "epoch": 1.7969451931716083,
      "grad_norm": 8.0625,
      "learning_rate": 2.0350404312668465e-06,
      "loss": 0.35046875,
      "step": 4000
    },
    {
      "epoch": 1.8194070080862534,
      "grad_norm": 13.125,
      "learning_rate": 1.8104222821203954e-06,
      "loss": 0.3076953125,
      "step": 4050
    },
    {
      "epoch": 1.8418688230008984,
      "grad_norm": 7.28125,
      "learning_rate": 1.5858041329739444e-06,
      "loss": 0.3510546875,
      "step": 4100
    },
    {
      "epoch": 1.8643306379155435,
      "grad_norm": 11.1875,
      "learning_rate": 1.3611859838274934e-06,
      "loss": 0.3405078125,
      "step": 4150
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 10.8125,
      "learning_rate": 1.1365678346810423e-06,
      "loss": 0.36060546875,
      "step": 4200
    },
    {
      "epoch": 1.9092542677448336,
      "grad_norm": 6.71875,
      "learning_rate": 9.119496855345912e-07,
      "loss": 0.3446484375,
      "step": 4250
    },
    {
      "epoch": 1.931716082659479,
      "grad_norm": 0.0,
      "learning_rate": 6.873315363881403e-07,
      "loss": 0.3423828125,
      "step": 4300
    },
    {
      "epoch": 1.954177897574124,
      "grad_norm": 8.5,
      "learning_rate": 4.627133872416892e-07,
      "loss": 0.3666015625,
      "step": 4350
    },
    {
      "epoch": 1.9766397124887691,
      "grad_norm": 13.75,
      "learning_rate": 2.3809523809523811e-07,
      "loss": 0.319375,
      "step": 4400
    },
    {
      "epoch": 1.9991015274034143,
      "grad_norm": 15.375,
      "learning_rate": 1.3477088948787064e-08,
      "loss": 0.3132421875,
      "step": 4450
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.48594674468040466,
      "eval_runtime": 1.5235,
      "eval_samples_per_second": 221.859,
      "eval_steps_per_second": 7.22,
      "step": 4452
    }
  ],
  "logging_steps": 50,
  "max_steps": 4452,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}