{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 4731,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03170577045022194,
      "grad_norm": 1.268200397491455,
      "learning_rate": 4.94715704924963e-05,
      "loss": 3.4652,
      "step": 50
    },
    {
      "epoch": 0.06341154090044387,
      "grad_norm": 0.5000430345535278,
      "learning_rate": 4.8943140984992603e-05,
      "loss": 0.7541,
      "step": 100
    },
    {
      "epoch": 0.09511731135066583,
      "grad_norm": 0.5261800289154053,
      "learning_rate": 4.8414711477488904e-05,
      "loss": 0.644,
      "step": 150
    },
    {
      "epoch": 0.12682308180088775,
      "grad_norm": 0.4932628571987152,
      "learning_rate": 4.788628196998521e-05,
      "loss": 0.6921,
      "step": 200
    },
    {
      "epoch": 0.15852885225110971,
      "grad_norm": 0.42867496609687805,
      "learning_rate": 4.7357852462481505e-05,
      "loss": 0.5897,
      "step": 250
    },
    {
      "epoch": 0.19023462270133165,
      "grad_norm": 0.8850355744361877,
      "learning_rate": 4.6829422954977806e-05,
      "loss": 0.6959,
      "step": 300
    },
    {
      "epoch": 0.2219403931515536,
      "grad_norm": 0.5738270878791809,
      "learning_rate": 4.630099344747411e-05,
      "loss": 0.754,
      "step": 350
    },
    {
      "epoch": 0.2536461636017755,
      "grad_norm": 0.2961026728153229,
      "learning_rate": 4.577256393997041e-05,
      "loss": 0.6201,
      "step": 400
    },
    {
      "epoch": 0.2853519340519975,
      "grad_norm": 0.35886821150779724,
      "learning_rate": 4.5244134432466714e-05,
      "loss": 0.5422,
      "step": 450
    },
    {
      "epoch": 0.31705770450221943,
      "grad_norm": 0.45828256011009216,
      "learning_rate": 4.4715704924963015e-05,
      "loss": 0.6895,
      "step": 500
    },
    {
      "epoch": 0.34876347495244137,
      "grad_norm": 0.40253156423568726,
      "learning_rate": 4.4187275417459315e-05,
      "loss": 0.5722,
      "step": 550
    },
    {
      "epoch": 0.3804692454026633,
      "grad_norm": 0.55234295129776,
      "learning_rate": 4.3658845909955616e-05,
      "loss": 0.6205,
      "step": 600
    },
    {
      "epoch": 0.41217501585288524,
      "grad_norm": 0.38063228130340576,
      "learning_rate": 4.3130416402451916e-05,
      "loss": 0.6161,
      "step": 650
    },
    {
      "epoch": 0.4438807863031072,
      "grad_norm": 0.8910940289497375,
      "learning_rate": 4.260198689494822e-05,
      "loss": 0.6756,
      "step": 700
    },
    {
      "epoch": 0.4755865567533291,
      "grad_norm": 0.569128692150116,
      "learning_rate": 4.207355738744452e-05,
      "loss": 0.5942,
      "step": 750
    },
    {
      "epoch": 0.507292327203551,
      "grad_norm": 0.713485062122345,
      "learning_rate": 4.154512787994082e-05,
      "loss": 0.5872,
      "step": 800
    },
    {
      "epoch": 0.5389980976537729,
      "grad_norm": 0.547140896320343,
      "learning_rate": 4.101669837243712e-05,
      "loss": 0.6482,
      "step": 850
    },
    {
      "epoch": 0.570703868103995,
      "grad_norm": 0.8154194355010986,
      "learning_rate": 4.048826886493342e-05,
      "loss": 0.5775,
      "step": 900
    },
    {
      "epoch": 0.6024096385542169,
      "grad_norm": 0.37084701657295227,
      "learning_rate": 3.995983935742972e-05,
      "loss": 0.5649,
      "step": 950
    },
    {
      "epoch": 0.6341154090044389,
      "grad_norm": 0.6335868835449219,
      "learning_rate": 3.943140984992602e-05,
      "loss": 0.5654,
      "step": 1000
    },
    {
      "epoch": 0.6658211794546608,
      "grad_norm": 0.3464125990867615,
      "learning_rate": 3.890298034242232e-05,
      "loss": 0.5236,
      "step": 1050
    },
    {
      "epoch": 0.6975269499048827,
      "grad_norm": 0.3801315128803253,
      "learning_rate": 3.837455083491862e-05,
      "loss": 0.5876,
      "step": 1100
    },
    {
      "epoch": 0.7292327203551047,
      "grad_norm": 0.489006370306015,
      "learning_rate": 3.784612132741492e-05,
      "loss": 0.5639,
      "step": 1150
    },
    {
      "epoch": 0.7609384908053266,
      "grad_norm": 0.9908373951911926,
      "learning_rate": 3.731769181991123e-05,
      "loss": 0.6626,
      "step": 1200
    },
    {
      "epoch": 0.7926442612555485,
      "grad_norm": 0.39679229259490967,
      "learning_rate": 3.678926231240752e-05,
      "loss": 0.6259,
      "step": 1250
    },
    {
      "epoch": 0.8243500317057705,
      "grad_norm": 0.3946998715400696,
      "learning_rate": 3.626083280490383e-05,
      "loss": 0.573,
      "step": 1300
    },
    {
      "epoch": 0.8560558021559924,
      "grad_norm": 0.41675782203674316,
      "learning_rate": 3.573240329740013e-05,
      "loss": 0.6568,
      "step": 1350
    },
    {
      "epoch": 0.8877615726062144,
      "grad_norm": 0.2899385988712311,
      "learning_rate": 3.5203973789896424e-05,
      "loss": 0.5031,
      "step": 1400
    },
    {
      "epoch": 0.9194673430564363,
      "grad_norm": 0.8684162497520447,
      "learning_rate": 3.467554428239273e-05,
      "loss": 0.5836,
      "step": 1450
    },
    {
      "epoch": 0.9511731135066582,
      "grad_norm": 0.9390568137168884,
      "learning_rate": 3.414711477488903e-05,
      "loss": 0.5832,
      "step": 1500
    },
    {
      "epoch": 0.9828788839568802,
      "grad_norm": 0.6241053342819214,
      "learning_rate": 3.361868526738533e-05,
      "loss": 0.6668,
      "step": 1550
    },
    {
      "epoch": 1.014584654407102,
      "grad_norm": 0.43029874563217163,
      "learning_rate": 3.309025575988163e-05,
      "loss": 0.5516,
      "step": 1600
    },
    {
      "epoch": 1.046290424857324,
      "grad_norm": 0.3394105136394501,
      "learning_rate": 3.2561826252377934e-05,
      "loss": 0.5909,
      "step": 1650
    },
    {
      "epoch": 1.0779961953075459,
      "grad_norm": 0.76275634765625,
      "learning_rate": 3.2033396744874234e-05,
      "loss": 0.5535,
      "step": 1700
    },
    {
      "epoch": 1.1097019657577678,
      "grad_norm": 0.6357107162475586,
      "learning_rate": 3.1504967237370535e-05,
      "loss": 0.5632,
      "step": 1750
    },
    {
      "epoch": 1.1414077362079897,
      "grad_norm": 0.3675929009914398,
      "learning_rate": 3.097653772986684e-05,
      "loss": 0.5744,
      "step": 1800
    },
    {
      "epoch": 1.1731135066582117,
      "grad_norm": 0.3702389895915985,
      "learning_rate": 3.0448108222363136e-05,
      "loss": 0.6319,
      "step": 1850
    },
    {
      "epoch": 1.2048192771084336,
      "grad_norm": 0.43757855892181396,
      "learning_rate": 2.991967871485944e-05,
      "loss": 0.5595,
      "step": 1900
    },
    {
      "epoch": 1.2365250475586556,
      "grad_norm": 0.40859177708625793,
      "learning_rate": 2.939124920735574e-05,
      "loss": 0.5012,
      "step": 1950
    },
    {
      "epoch": 1.2682308180088775,
      "grad_norm": 0.8433944582939148,
      "learning_rate": 2.886281969985204e-05,
      "loss": 0.6965,
      "step": 2000
    },
    {
      "epoch": 1.2999365884590994,
      "grad_norm": 0.2780860662460327,
      "learning_rate": 2.833439019234834e-05,
      "loss": 0.6852,
      "step": 2050
    },
    {
      "epoch": 1.3316423589093214,
      "grad_norm": 0.4954093098640442,
      "learning_rate": 2.7805960684844646e-05,
      "loss": 0.5412,
      "step": 2100
    },
    {
      "epoch": 1.3633481293595433,
      "grad_norm": 0.9253292083740234,
      "learning_rate": 2.7277531177340943e-05,
      "loss": 0.5013,
      "step": 2150
    },
    {
      "epoch": 1.3950538998097652,
      "grad_norm": 0.5632855892181396,
      "learning_rate": 2.6749101669837247e-05,
      "loss": 0.5343,
      "step": 2200
    },
    {
      "epoch": 1.4267596702599874,
      "grad_norm": 0.5401332974433899,
      "learning_rate": 2.6220672162333547e-05,
      "loss": 0.4798,
      "step": 2250
    },
    {
      "epoch": 1.4584654407102093,
      "grad_norm": 0.6006923913955688,
      "learning_rate": 2.5692242654829844e-05,
      "loss": 0.5151,
      "step": 2300
    },
    {
      "epoch": 1.4901712111604313,
      "grad_norm": 1.1926701068878174,
      "learning_rate": 2.5163813147326148e-05,
      "loss": 0.4518,
      "step": 2350
    },
    {
      "epoch": 1.521876981610653,
      "grad_norm": 0.37329962849617004,
      "learning_rate": 2.463538363982245e-05,
      "loss": 0.5288,
      "step": 2400
    },
    {
      "epoch": 1.553582752060875,
      "grad_norm": 0.49744054675102234,
      "learning_rate": 2.410695413231875e-05,
      "loss": 0.5408,
      "step": 2450
    },
    {
      "epoch": 1.5852885225110969,
      "grad_norm": 0.4050131142139435,
      "learning_rate": 2.357852462481505e-05,
      "loss": 0.5719,
      "step": 2500
    },
    {
      "epoch": 1.6169942929613188,
      "grad_norm": 0.34927618503570557,
      "learning_rate": 2.305009511731135e-05,
      "loss": 0.4688,
      "step": 2550
    },
    {
      "epoch": 1.6487000634115407,
      "grad_norm": 0.3404935896396637,
      "learning_rate": 2.252166560980765e-05,
      "loss": 0.5272,
      "step": 2600
    },
    {
      "epoch": 1.6804058338617627,
      "grad_norm": 0.3460475504398346,
      "learning_rate": 2.1993236102303955e-05,
      "loss": 0.632,
      "step": 2650
    },
    {
      "epoch": 1.7121116043119848,
      "grad_norm": 0.47324368357658386,
      "learning_rate": 2.1464806594800255e-05,
      "loss": 0.5488,
      "step": 2700
    },
    {
      "epoch": 1.7438173747622068,
      "grad_norm": 0.7189714908599854,
      "learning_rate": 2.0936377087296556e-05,
      "loss": 0.5807,
      "step": 2750
    },
    {
      "epoch": 1.7755231452124287,
      "grad_norm": 0.7753428816795349,
      "learning_rate": 2.0407947579792857e-05,
      "loss": 0.5803,
      "step": 2800
    },
    {
      "epoch": 1.8072289156626506,
      "grad_norm": 0.32366201281547546,
      "learning_rate": 1.9879518072289157e-05,
      "loss": 0.5606,
      "step": 2850
    },
    {
      "epoch": 1.8389346861128726,
      "grad_norm": 0.42843538522720337,
      "learning_rate": 1.9351088564785458e-05,
      "loss": 0.4822,
      "step": 2900
    },
    {
      "epoch": 1.8706404565630945,
      "grad_norm": 0.8468919992446899,
      "learning_rate": 1.882265905728176e-05,
      "loss": 0.7272,
      "step": 2950
    },
    {
      "epoch": 1.9023462270133165,
      "grad_norm": 1.0852563381195068,
      "learning_rate": 1.8294229549778062e-05,
      "loss": 0.5527,
      "step": 3000
    },
    {
      "epoch": 1.9340519974635384,
      "grad_norm": 0.41115501523017883,
      "learning_rate": 1.776580004227436e-05,
      "loss": 0.5216,
      "step": 3050
    },
    {
      "epoch": 1.9657577679137603,
      "grad_norm": 0.4568431079387665,
      "learning_rate": 1.7237370534770663e-05,
      "loss": 0.5098,
      "step": 3100
    },
    {
      "epoch": 1.9974635383639823,
      "grad_norm": 0.3544299602508545,
      "learning_rate": 1.6708941027266964e-05,
      "loss": 0.584,
      "step": 3150
    },
    {
      "epoch": 2.029169308814204,
      "grad_norm": 0.8790750503540039,
      "learning_rate": 1.6180511519763264e-05,
      "loss": 0.43,
      "step": 3200
    },
    {
      "epoch": 2.060875079264426,
      "grad_norm": 0.4061146080493927,
      "learning_rate": 1.5652082012259565e-05,
      "loss": 0.4727,
      "step": 3250
    },
    {
      "epoch": 2.092580849714648,
      "grad_norm": 0.4029136002063751,
      "learning_rate": 1.5123652504755867e-05,
      "loss": 0.4701,
      "step": 3300
    },
    {
      "epoch": 2.12428662016487,
      "grad_norm": 0.7732370495796204,
      "learning_rate": 1.4595222997252168e-05,
      "loss": 0.4954,
      "step": 3350
    },
    {
      "epoch": 2.1559923906150917,
      "grad_norm": 0.7356324791908264,
      "learning_rate": 1.4066793489748466e-05,
      "loss": 0.6128,
      "step": 3400
    },
    {
      "epoch": 2.187698161065314,
      "grad_norm": 0.8712078332901001,
      "learning_rate": 1.353836398224477e-05,
      "loss": 0.6239,
      "step": 3450
    },
    {
      "epoch": 2.2194039315155356,
      "grad_norm": 0.44863361120224,
      "learning_rate": 1.3009934474741071e-05,
      "loss": 0.5606,
      "step": 3500
    },
    {
      "epoch": 2.251109701965758,
      "grad_norm": 0.9408974051475525,
      "learning_rate": 1.2481504967237371e-05,
      "loss": 0.514,
      "step": 3550
    },
    {
      "epoch": 2.2828154724159795,
      "grad_norm": 0.3982991874217987,
      "learning_rate": 1.1953075459733672e-05,
      "loss": 0.47,
      "step": 3600
    },
    {
      "epoch": 2.314521242866202,
      "grad_norm": 0.4143155813217163,
      "learning_rate": 1.1424645952229973e-05,
      "loss": 0.5677,
      "step": 3650
    },
    {
      "epoch": 2.3462270133164234,
      "grad_norm": 0.31523430347442627,
      "learning_rate": 1.0896216444726275e-05,
      "loss": 0.5579,
      "step": 3700
    },
    {
      "epoch": 2.3779327837666457,
      "grad_norm": 0.41940435767173767,
      "learning_rate": 1.0367786937222575e-05,
      "loss": 0.5204,
      "step": 3750
    },
    {
      "epoch": 2.4096385542168672,
      "grad_norm": 0.7688931226730347,
      "learning_rate": 9.839357429718876e-06,
      "loss": 0.5144,
      "step": 3800
    },
    {
      "epoch": 2.4413443246670896,
      "grad_norm": 0.3608909845352173,
      "learning_rate": 9.310927922215176e-06,
      "loss": 0.4891,
      "step": 3850
    },
    {
      "epoch": 2.473050095117311,
      "grad_norm": 0.47781404852867126,
      "learning_rate": 8.782498414711479e-06,
      "loss": 0.4892,
      "step": 3900
    },
    {
      "epoch": 2.5047558655675335,
      "grad_norm": 0.3204675316810608,
      "learning_rate": 8.254068907207779e-06,
      "loss": 0.5816,
      "step": 3950
    },
    {
      "epoch": 2.536461636017755,
      "grad_norm": 0.806425929069519,
      "learning_rate": 7.72563939970408e-06,
      "loss": 0.5725,
      "step": 4000
    },
    {
      "epoch": 2.5681674064679774,
      "grad_norm": 0.8784732818603516,
      "learning_rate": 7.197209892200381e-06,
      "loss": 0.63,
      "step": 4050
    },
    {
      "epoch": 2.599873176918199,
      "grad_norm": 0.26710501313209534,
      "learning_rate": 6.6687803846966825e-06,
      "loss": 0.4984,
      "step": 4100
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 0.46410611271858215,
      "learning_rate": 6.140350877192982e-06,
      "loss": 0.5894,
      "step": 4150
    },
    {
      "epoch": 2.6632847178186427,
      "grad_norm": 0.39962610602378845,
      "learning_rate": 5.6119213696892836e-06,
      "loss": 0.5719,
      "step": 4200
    },
    {
      "epoch": 2.694990488268865,
      "grad_norm": 0.42166388034820557,
      "learning_rate": 5.083491862185585e-06,
      "loss": 0.611,
      "step": 4250
    },
    {
      "epoch": 2.7266962587190866,
      "grad_norm": 0.3567692041397095,
      "learning_rate": 4.5550623546818855e-06,
      "loss": 0.536,
      "step": 4300
    },
    {
      "epoch": 2.758402029169309,
      "grad_norm": 0.7851996421813965,
      "learning_rate": 4.026632847178187e-06,
      "loss": 0.5243,
      "step": 4350
    },
    {
      "epoch": 2.7901077996195305,
      "grad_norm": 0.49054574966430664,
      "learning_rate": 3.498203339674488e-06,
      "loss": 0.4414,
      "step": 4400
    },
    {
      "epoch": 2.821813570069753,
      "grad_norm": 0.3000207841396332,
      "learning_rate": 2.9697738321707884e-06,
      "loss": 0.4498,
      "step": 4450
    },
    {
      "epoch": 2.853519340519975,
      "grad_norm": 0.7114095091819763,
      "learning_rate": 2.4413443246670898e-06,
      "loss": 0.5504,
      "step": 4500
    },
    {
      "epoch": 2.8852251109701967,
      "grad_norm": 0.9974462985992432,
      "learning_rate": 1.9129148171633903e-06,
      "loss": 0.6545,
      "step": 4550
    },
    {
      "epoch": 2.9169308814204187,
      "grad_norm": 0.339517205953598,
      "learning_rate": 1.3844853096596915e-06,
      "loss": 0.482,
      "step": 4600
    },
    {
      "epoch": 2.9486366518706406,
      "grad_norm": 0.3462202548980713,
      "learning_rate": 8.560558021559924e-07,
      "loss": 0.5029,
      "step": 4650
    },
    {
      "epoch": 2.9803424223208625,
      "grad_norm": 0.6948489546775818,
      "learning_rate": 3.276262946522934e-07,
      "loss": 0.4969,
      "step": 4700
    }
  ],
  "logging_steps": 50,
  "max_steps": 4731,
  "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": 2471997908975616.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}