{
  "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": 347.7311096191406,
      "learning_rate": 4.94715704924963e-05,
      "loss": 8.5747,
      "step": 50
    },
    {
      "epoch": 0.06341154090044387,
      "grad_norm": 216.6991424560547,
      "learning_rate": 4.8943140984992603e-05,
      "loss": 3.9346,
      "step": 100
    },
    {
      "epoch": 0.09511731135066583,
      "grad_norm": 1.8576736450195312,
      "learning_rate": 4.8414711477488904e-05,
      "loss": 0.7161,
      "step": 150
    },
    {
      "epoch": 0.12682308180088775,
      "grad_norm": 1.3204137086868286,
      "learning_rate": 4.788628196998521e-05,
      "loss": 0.6464,
      "step": 200
    },
    {
      "epoch": 0.15852885225110971,
      "grad_norm": 1.0963375568389893,
      "learning_rate": 4.7357852462481505e-05,
      "loss": 0.5442,
      "step": 250
    },
    {
      "epoch": 0.19023462270133165,
      "grad_norm": 1.6606075763702393,
      "learning_rate": 4.6829422954977806e-05,
      "loss": 0.6393,
      "step": 300
    },
    {
      "epoch": 0.2219403931515536,
      "grad_norm": 0.8779278993606567,
      "learning_rate": 4.630099344747411e-05,
      "loss": 0.689,
      "step": 350
    },
    {
      "epoch": 0.2536461636017755,
      "grad_norm": 0.6462757587432861,
      "learning_rate": 4.577256393997041e-05,
      "loss": 0.5693,
      "step": 400
    },
    {
      "epoch": 0.2853519340519975,
      "grad_norm": 0.670758843421936,
      "learning_rate": 4.5244134432466714e-05,
      "loss": 0.4993,
      "step": 450
    },
    {
      "epoch": 0.31705770450221943,
      "grad_norm": 0.7589445114135742,
      "learning_rate": 4.4715704924963015e-05,
      "loss": 0.6332,
      "step": 500
    },
    {
      "epoch": 0.34876347495244137,
      "grad_norm": 0.7136730551719666,
      "learning_rate": 4.4187275417459315e-05,
      "loss": 0.5269,
      "step": 550
    },
    {
      "epoch": 0.3804692454026633,
      "grad_norm": 0.9714166522026062,
      "learning_rate": 4.3658845909955616e-05,
      "loss": 0.5706,
      "step": 600
    },
    {
      "epoch": 0.41217501585288524,
      "grad_norm": 0.7672790884971619,
      "learning_rate": 4.3130416402451916e-05,
      "loss": 0.5689,
      "step": 650
    },
    {
      "epoch": 0.4438807863031072,
      "grad_norm": 1.4292829036712646,
      "learning_rate": 4.260198689494822e-05,
      "loss": 0.6206,
      "step": 700
    },
    {
      "epoch": 0.4755865567533291,
      "grad_norm": 0.8618677258491516,
      "learning_rate": 4.207355738744452e-05,
      "loss": 0.5413,
      "step": 750
    },
    {
      "epoch": 0.507292327203551,
      "grad_norm": 0.9496092796325684,
      "learning_rate": 4.154512787994082e-05,
      "loss": 0.5418,
      "step": 800
    },
    {
      "epoch": 0.5389980976537729,
      "grad_norm": 0.8666266202926636,
      "learning_rate": 4.101669837243712e-05,
      "loss": 0.597,
      "step": 850
    },
    {
      "epoch": 0.570703868103995,
      "grad_norm": 1.0330841541290283,
      "learning_rate": 4.048826886493342e-05,
      "loss": 0.5309,
      "step": 900
    },
    {
      "epoch": 0.6024096385542169,
      "grad_norm": 0.6087735295295715,
      "learning_rate": 3.995983935742972e-05,
      "loss": 0.5225,
      "step": 950
    },
    {
      "epoch": 0.6341154090044389,
      "grad_norm": 0.9573830962181091,
      "learning_rate": 3.943140984992602e-05,
      "loss": 0.5227,
      "step": 1000
    },
    {
      "epoch": 0.6658211794546608,
      "grad_norm": 0.6024531126022339,
      "learning_rate": 3.890298034242232e-05,
      "loss": 0.4838,
      "step": 1050
    },
    {
      "epoch": 0.6975269499048827,
      "grad_norm": 0.5939881801605225,
      "learning_rate": 3.837455083491862e-05,
      "loss": 0.541,
      "step": 1100
    },
    {
      "epoch": 0.7292327203551047,
      "grad_norm": 0.6684996485710144,
      "learning_rate": 3.784612132741492e-05,
      "loss": 0.5198,
      "step": 1150
    },
    {
      "epoch": 0.7609384908053266,
      "grad_norm": 1.1875889301300049,
      "learning_rate": 3.731769181991123e-05,
      "loss": 0.6105,
      "step": 1200
    },
    {
      "epoch": 0.7926442612555485,
      "grad_norm": 0.570648729801178,
      "learning_rate": 3.678926231240752e-05,
      "loss": 0.5735,
      "step": 1250
    },
    {
      "epoch": 0.8243500317057705,
      "grad_norm": 0.6439939141273499,
      "learning_rate": 3.626083280490383e-05,
      "loss": 0.5283,
      "step": 1300
    },
    {
      "epoch": 0.8560558021559924,
      "grad_norm": 0.6673012375831604,
      "learning_rate": 3.573240329740013e-05,
      "loss": 0.6057,
      "step": 1350
    },
    {
      "epoch": 0.8877615726062144,
      "grad_norm": 0.4656541645526886,
      "learning_rate": 3.5203973789896424e-05,
      "loss": 0.4664,
      "step": 1400
    },
    {
      "epoch": 0.9194673430564363,
      "grad_norm": 1.075217366218567,
      "learning_rate": 3.467554428239273e-05,
      "loss": 0.5336,
      "step": 1450
    },
    {
      "epoch": 0.9511731135066582,
      "grad_norm": 1.2251460552215576,
      "learning_rate": 3.414711477488903e-05,
      "loss": 0.5354,
      "step": 1500
    },
    {
      "epoch": 0.9828788839568802,
      "grad_norm": 0.822003185749054,
      "learning_rate": 3.361868526738533e-05,
      "loss": 0.6133,
      "step": 1550
    },
    {
      "epoch": 1.014584654407102,
      "grad_norm": 0.5614562630653381,
      "learning_rate": 3.309025575988163e-05,
      "loss": 0.5059,
      "step": 1600
    },
    {
      "epoch": 1.046290424857324,
      "grad_norm": 0.5171016454696655,
      "learning_rate": 3.2561826252377934e-05,
      "loss": 0.5285,
      "step": 1650
    },
    {
      "epoch": 1.0779961953075459,
      "grad_norm": 0.9495669007301331,
      "learning_rate": 3.2033396744874234e-05,
      "loss": 0.4955,
      "step": 1700
    },
    {
      "epoch": 1.1097019657577678,
      "grad_norm": 0.9576842784881592,
      "learning_rate": 3.1504967237370535e-05,
      "loss": 0.5107,
      "step": 1750
    },
    {
      "epoch": 1.1414077362079897,
      "grad_norm": 0.6107915639877319,
      "learning_rate": 3.097653772986684e-05,
      "loss": 0.519,
      "step": 1800
    },
    {
      "epoch": 1.1731135066582117,
      "grad_norm": 0.5931509137153625,
      "learning_rate": 3.0448108222363136e-05,
      "loss": 0.5699,
      "step": 1850
    },
    {
      "epoch": 1.2048192771084336,
      "grad_norm": 0.6647824048995972,
      "learning_rate": 2.991967871485944e-05,
      "loss": 0.5022,
      "step": 1900
    },
    {
      "epoch": 1.2365250475586556,
      "grad_norm": 0.5938781499862671,
      "learning_rate": 2.939124920735574e-05,
      "loss": 0.4507,
      "step": 1950
    },
    {
      "epoch": 1.2682308180088775,
      "grad_norm": 1.1466686725616455,
      "learning_rate": 2.886281969985204e-05,
      "loss": 0.6235,
      "step": 2000
    },
    {
      "epoch": 1.2999365884590994,
      "grad_norm": 0.4420027434825897,
      "learning_rate": 2.833439019234834e-05,
      "loss": 0.6154,
      "step": 2050
    },
    {
      "epoch": 1.3316423589093214,
      "grad_norm": 0.746532142162323,
      "learning_rate": 2.7805960684844646e-05,
      "loss": 0.4898,
      "step": 2100
    },
    {
      "epoch": 1.3633481293595433,
      "grad_norm": 1.0959316492080688,
      "learning_rate": 2.7277531177340943e-05,
      "loss": 0.4544,
      "step": 2150
    },
    {
      "epoch": 1.3950538998097652,
      "grad_norm": 0.8176191449165344,
      "learning_rate": 2.6749101669837247e-05,
      "loss": 0.481,
      "step": 2200
    },
    {
      "epoch": 1.4267596702599874,
      "grad_norm": 0.8100579380989075,
      "learning_rate": 2.6220672162333547e-05,
      "loss": 0.4333,
      "step": 2250
    },
    {
      "epoch": 1.4584654407102093,
      "grad_norm": 0.8650724291801453,
      "learning_rate": 2.5692242654829844e-05,
      "loss": 0.4629,
      "step": 2300
    },
    {
      "epoch": 1.4901712111604313,
      "grad_norm": 1.6807165145874023,
      "learning_rate": 2.5163813147326148e-05,
      "loss": 0.4126,
      "step": 2350
    },
    {
      "epoch": 1.521876981610653,
      "grad_norm": 0.6398534774780273,
      "learning_rate": 2.463538363982245e-05,
      "loss": 0.4776,
      "step": 2400
    },
    {
      "epoch": 1.553582752060875,
      "grad_norm": 0.7640373110771179,
      "learning_rate": 2.410695413231875e-05,
      "loss": 0.4901,
      "step": 2450
    },
    {
      "epoch": 1.5852885225110969,
      "grad_norm": 0.644173800945282,
      "learning_rate": 2.357852462481505e-05,
      "loss": 0.5133,
      "step": 2500
    },
    {
      "epoch": 1.6169942929613188,
      "grad_norm": 0.577743649482727,
      "learning_rate": 2.305009511731135e-05,
      "loss": 0.4221,
      "step": 2550
    },
    {
      "epoch": 1.6487000634115407,
      "grad_norm": 0.6014630198478699,
      "learning_rate": 2.252166560980765e-05,
      "loss": 0.4766,
      "step": 2600
    },
    {
      "epoch": 1.6804058338617627,
      "grad_norm": 0.6592232584953308,
      "learning_rate": 2.1993236102303955e-05,
      "loss": 0.5726,
      "step": 2650
    },
    {
      "epoch": 1.7121116043119848,
      "grad_norm": 0.7719835042953491,
      "learning_rate": 2.1464806594800255e-05,
      "loss": 0.4941,
      "step": 2700
    },
    {
      "epoch": 1.7438173747622068,
      "grad_norm": 0.9424024224281311,
      "learning_rate": 2.0936377087296556e-05,
      "loss": 0.5236,
      "step": 2750
    },
    {
      "epoch": 1.7755231452124287,
      "grad_norm": 1.0417604446411133,
      "learning_rate": 2.0407947579792857e-05,
      "loss": 0.5269,
      "step": 2800
    },
    {
      "epoch": 1.8072289156626506,
      "grad_norm": 0.42674678564071655,
      "learning_rate": 1.9879518072289157e-05,
      "loss": 0.5067,
      "step": 2850
    },
    {
      "epoch": 1.8389346861128726,
      "grad_norm": 0.6428393721580505,
      "learning_rate": 1.9351088564785458e-05,
      "loss": 0.4381,
      "step": 2900
    },
    {
      "epoch": 1.8706404565630945,
      "grad_norm": 0.936777651309967,
      "learning_rate": 1.882265905728176e-05,
      "loss": 0.6561,
      "step": 2950
    },
    {
      "epoch": 1.9023462270133165,
      "grad_norm": 1.3489336967468262,
      "learning_rate": 1.8294229549778062e-05,
      "loss": 0.4983,
      "step": 3000
    },
    {
      "epoch": 1.9340519974635384,
      "grad_norm": 0.6496655344963074,
      "learning_rate": 1.776580004227436e-05,
      "loss": 0.4759,
      "step": 3050
    },
    {
      "epoch": 1.9657577679137603,
      "grad_norm": 0.712348997592926,
      "learning_rate": 1.7237370534770663e-05,
      "loss": 0.4621,
      "step": 3100
    },
    {
      "epoch": 1.9974635383639823,
      "grad_norm": 0.5334410667419434,
      "learning_rate": 1.6708941027266964e-05,
      "loss": 0.5246,
      "step": 3150
    },
    {
      "epoch": 2.029169308814204,
      "grad_norm": 0.9799864888191223,
      "learning_rate": 1.6180511519763264e-05,
      "loss": 0.388,
      "step": 3200
    },
    {
      "epoch": 2.060875079264426,
      "grad_norm": 0.6191227436065674,
      "learning_rate": 1.5652082012259565e-05,
      "loss": 0.422,
      "step": 3250
    },
    {
      "epoch": 2.092580849714648,
      "grad_norm": 0.6415786743164062,
      "learning_rate": 1.5123652504755867e-05,
      "loss": 0.4213,
      "step": 3300
    },
    {
      "epoch": 2.12428662016487,
      "grad_norm": 1.024706244468689,
      "learning_rate": 1.4595222997252168e-05,
      "loss": 0.4378,
      "step": 3350
    },
    {
      "epoch": 2.1559923906150917,
      "grad_norm": 1.0147749185562134,
      "learning_rate": 1.4066793489748466e-05,
      "loss": 0.5441,
      "step": 3400
    },
    {
      "epoch": 2.187698161065314,
      "grad_norm": 1.1020487546920776,
      "learning_rate": 1.353836398224477e-05,
      "loss": 0.554,
      "step": 3450
    },
    {
      "epoch": 2.2194039315155356,
      "grad_norm": 0.7216790318489075,
      "learning_rate": 1.3009934474741071e-05,
      "loss": 0.4983,
      "step": 3500
    },
    {
      "epoch": 2.251109701965758,
      "grad_norm": 1.2225050926208496,
      "learning_rate": 1.2481504967237371e-05,
      "loss": 0.4586,
      "step": 3550
    },
    {
      "epoch": 2.2828154724159795,
      "grad_norm": 0.7050449252128601,
      "learning_rate": 1.1953075459733672e-05,
      "loss": 0.4235,
      "step": 3600
    },
    {
      "epoch": 2.314521242866202,
      "grad_norm": 0.6529106497764587,
      "learning_rate": 1.1424645952229973e-05,
      "loss": 0.5061,
      "step": 3650
    },
    {
      "epoch": 2.3462270133164234,
      "grad_norm": 0.5396713018417358,
      "learning_rate": 1.0896216444726275e-05,
      "loss": 0.4983,
      "step": 3700
    },
    {
      "epoch": 2.3779327837666457,
      "grad_norm": 0.6897116899490356,
      "learning_rate": 1.0367786937222575e-05,
      "loss": 0.4658,
      "step": 3750
    },
    {
      "epoch": 2.4096385542168672,
      "grad_norm": 1.0474814176559448,
      "learning_rate": 9.839357429718876e-06,
      "loss": 0.4578,
      "step": 3800
    },
    {
      "epoch": 2.4413443246670896,
      "grad_norm": 0.5905832648277283,
      "learning_rate": 9.310927922215176e-06,
      "loss": 0.4349,
      "step": 3850
    },
    {
      "epoch": 2.473050095117311,
      "grad_norm": 0.6709778904914856,
      "learning_rate": 8.782498414711479e-06,
      "loss": 0.4351,
      "step": 3900
    },
    {
      "epoch": 2.5047558655675335,
      "grad_norm": 0.5383151173591614,
      "learning_rate": 8.254068907207779e-06,
      "loss": 0.52,
      "step": 3950
    },
    {
      "epoch": 2.536461636017755,
      "grad_norm": 0.9183222055435181,
      "learning_rate": 7.72563939970408e-06,
      "loss": 0.5099,
      "step": 4000
    },
    {
      "epoch": 2.5681674064679774,
      "grad_norm": 1.0442452430725098,
      "learning_rate": 7.197209892200381e-06,
      "loss": 0.561,
      "step": 4050
    },
    {
      "epoch": 2.599873176918199,
      "grad_norm": 0.4895930886268616,
      "learning_rate": 6.6687803846966825e-06,
      "loss": 0.4435,
      "step": 4100
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 0.6229114532470703,
      "learning_rate": 6.140350877192982e-06,
      "loss": 0.5262,
      "step": 4150
    },
    {
      "epoch": 2.6632847178186427,
      "grad_norm": 0.647319495677948,
      "learning_rate": 5.6119213696892836e-06,
      "loss": 0.5116,
      "step": 4200
    },
    {
      "epoch": 2.694990488268865,
      "grad_norm": 0.6666443347930908,
      "learning_rate": 5.083491862185585e-06,
      "loss": 0.5408,
      "step": 4250
    },
    {
      "epoch": 2.7266962587190866,
      "grad_norm": 0.6282554268836975,
      "learning_rate": 4.5550623546818855e-06,
      "loss": 0.4801,
      "step": 4300
    },
    {
      "epoch": 2.758402029169309,
      "grad_norm": 1.0603163242340088,
      "learning_rate": 4.026632847178187e-06,
      "loss": 0.4662,
      "step": 4350
    },
    {
      "epoch": 2.7901077996195305,
      "grad_norm": 0.7217985391616821,
      "learning_rate": 3.498203339674488e-06,
      "loss": 0.3947,
      "step": 4400
    },
    {
      "epoch": 2.821813570069753,
      "grad_norm": 0.49021148681640625,
      "learning_rate": 2.9697738321707884e-06,
      "loss": 0.4047,
      "step": 4450
    },
    {
      "epoch": 2.853519340519975,
      "grad_norm": 0.9363583922386169,
      "learning_rate": 2.4413443246670898e-06,
      "loss": 0.4912,
      "step": 4500
    },
    {
      "epoch": 2.8852251109701967,
      "grad_norm": 0.9978638291358948,
      "learning_rate": 1.9129148171633903e-06,
      "loss": 0.579,
      "step": 4550
    },
    {
      "epoch": 2.9169308814204187,
      "grad_norm": 0.6237664222717285,
      "learning_rate": 1.3844853096596915e-06,
      "loss": 0.4324,
      "step": 4600
    },
    {
      "epoch": 2.9486366518706406,
      "grad_norm": 0.6290932893753052,
      "learning_rate": 8.560558021559924e-07,
      "loss": 0.4495,
      "step": 4650
    },
    {
      "epoch": 2.9803424223208625,
      "grad_norm": 0.8747038245201111,
      "learning_rate": 3.276262946522934e-07,
      "loss": 0.44,
      "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": 4943906537472000.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}