{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.01068979463122881,
  "eval_steps": 500,
  "global_step": 12000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 173.0,
      "learning_rate": 4.4539461963299484e-07,
      "loss": 2.153,
      "step": 100
    },
    {
      "epoch": 0.0,
      "grad_norm": 40.5,
      "learning_rate": 8.907892392659897e-07,
      "loss": 2.398,
      "step": 200
    },
    {
      "epoch": 0.0,
      "grad_norm": 223.0,
      "learning_rate": 1.3361838588989846e-06,
      "loss": 1.8912,
      "step": 300
    },
    {
      "epoch": 0.0,
      "grad_norm": 95.5,
      "learning_rate": 1.7815784785319793e-06,
      "loss": 1.6403,
      "step": 400
    },
    {
      "epoch": 0.0,
      "grad_norm": 600.0,
      "learning_rate": 2.226973098164974e-06,
      "loss": 1.4652,
      "step": 500
    },
    {
      "epoch": 0.0,
      "grad_norm": 51.75,
      "learning_rate": 2.672367717797969e-06,
      "loss": 1.2489,
      "step": 600
    },
    {
      "epoch": 0.0,
      "grad_norm": 290.0,
      "learning_rate": 3.1177623374309637e-06,
      "loss": 1.2421,
      "step": 700
    },
    {
      "epoch": 0.0,
      "grad_norm": 27.75,
      "learning_rate": 3.5631569570639587e-06,
      "loss": 1.2188,
      "step": 800
    },
    {
      "epoch": 0.0,
      "grad_norm": 13.0,
      "learning_rate": 4.008551576696954e-06,
      "loss": 1.123,
      "step": 900
    },
    {
      "epoch": 0.0,
      "grad_norm": 39.75,
      "learning_rate": 4.453946196329948e-06,
      "loss": 1.0388,
      "step": 1000
    },
    {
      "epoch": 0.0,
      "grad_norm": 21.375,
      "learning_rate": 4.899340815962943e-06,
      "loss": 1.3034,
      "step": 1100
    },
    {
      "epoch": 0.0,
      "grad_norm": 36.75,
      "learning_rate": 5.344735435595938e-06,
      "loss": 1.2272,
      "step": 1200
    },
    {
      "epoch": 0.0,
      "grad_norm": 71.5,
      "learning_rate": 5.790130055228933e-06,
      "loss": 1.1825,
      "step": 1300
    },
    {
      "epoch": 0.0,
      "grad_norm": 54.5,
      "learning_rate": 6.235524674861927e-06,
      "loss": 1.3466,
      "step": 1400
    },
    {
      "epoch": 0.0,
      "grad_norm": 169.0,
      "learning_rate": 6.680919294494922e-06,
      "loss": 1.0543,
      "step": 1500
    },
    {
      "epoch": 0.0,
      "grad_norm": 47.75,
      "learning_rate": 7.126313914127917e-06,
      "loss": 1.1117,
      "step": 1600
    },
    {
      "epoch": 0.0,
      "grad_norm": 9.5,
      "learning_rate": 7.571708533760913e-06,
      "loss": 1.1458,
      "step": 1700
    },
    {
      "epoch": 0.0,
      "grad_norm": 119.0,
      "learning_rate": 8.017103153393907e-06,
      "loss": 1.1823,
      "step": 1800
    },
    {
      "epoch": 0.0,
      "grad_norm": 31.0,
      "learning_rate": 8.462497773026902e-06,
      "loss": 1.2171,
      "step": 1900
    },
    {
      "epoch": 0.0,
      "grad_norm": 57.25,
      "learning_rate": 8.907892392659896e-06,
      "loss": 1.16,
      "step": 2000
    },
    {
      "epoch": 0.0,
      "grad_norm": 77.0,
      "learning_rate": 9.353287012292893e-06,
      "loss": 1.1243,
      "step": 2100
    },
    {
      "epoch": 0.0,
      "grad_norm": 31.5,
      "learning_rate": 9.798681631925886e-06,
      "loss": 1.1218,
      "step": 2200
    },
    {
      "epoch": 0.0,
      "grad_norm": 98.0,
      "learning_rate": 1.0244076251558882e-05,
      "loss": 1.1527,
      "step": 2300
    },
    {
      "epoch": 0.0,
      "grad_norm": 32.25,
      "learning_rate": 1.0689470871191876e-05,
      "loss": 1.1042,
      "step": 2400
    },
    {
      "epoch": 0.0,
      "grad_norm": 29.0,
      "learning_rate": 1.1134865490824871e-05,
      "loss": 1.0131,
      "step": 2500
    },
    {
      "epoch": 0.0,
      "grad_norm": 50.0,
      "learning_rate": 1.1580260110457866e-05,
      "loss": 1.2544,
      "step": 2600
    },
    {
      "epoch": 0.0,
      "grad_norm": 20.625,
      "learning_rate": 1.2025654730090862e-05,
      "loss": 1.1475,
      "step": 2700
    },
    {
      "epoch": 0.0,
      "grad_norm": 23.375,
      "learning_rate": 1.2471049349723855e-05,
      "loss": 1.2345,
      "step": 2800
    },
    {
      "epoch": 0.0,
      "grad_norm": 39.5,
      "learning_rate": 1.2916443969356851e-05,
      "loss": 1.269,
      "step": 2900
    },
    {
      "epoch": 0.0,
      "grad_norm": 30.5,
      "learning_rate": 1.3361838588989844e-05,
      "loss": 1.1995,
      "step": 3000
    },
    {
      "epoch": 0.0,
      "grad_norm": 17.875,
      "learning_rate": 1.3807233208622842e-05,
      "loss": 1.1367,
      "step": 3100
    },
    {
      "epoch": 0.0,
      "grad_norm": 255.0,
      "learning_rate": 1.4252627828255835e-05,
      "loss": 1.1993,
      "step": 3200
    },
    {
      "epoch": 0.0,
      "grad_norm": 73.5,
      "learning_rate": 1.469802244788883e-05,
      "loss": 1.039,
      "step": 3300
    },
    {
      "epoch": 0.0,
      "grad_norm": 106.5,
      "learning_rate": 1.5143417067521826e-05,
      "loss": 1.2538,
      "step": 3400
    },
    {
      "epoch": 0.0,
      "grad_norm": 153.0,
      "learning_rate": 1.558881168715482e-05,
      "loss": 1.156,
      "step": 3500
    },
    {
      "epoch": 0.0,
      "grad_norm": 111.0,
      "learning_rate": 1.6034206306787815e-05,
      "loss": 1.2418,
      "step": 3600
    },
    {
      "epoch": 0.0,
      "grad_norm": 20.25,
      "learning_rate": 1.647960092642081e-05,
      "loss": 1.0018,
      "step": 3700
    },
    {
      "epoch": 0.0,
      "grad_norm": 68.0,
      "learning_rate": 1.6924995546053804e-05,
      "loss": 1.101,
      "step": 3800
    },
    {
      "epoch": 0.0,
      "grad_norm": 79.5,
      "learning_rate": 1.7370390165686802e-05,
      "loss": 1.1538,
      "step": 3900
    },
    {
      "epoch": 0.0,
      "grad_norm": 26.5,
      "learning_rate": 1.7815784785319793e-05,
      "loss": 1.176,
      "step": 4000
    },
    {
      "epoch": 0.0,
      "grad_norm": 96.5,
      "learning_rate": 1.8261179404952788e-05,
      "loss": 0.9909,
      "step": 4100
    },
    {
      "epoch": 0.0,
      "grad_norm": 45.75,
      "learning_rate": 1.8706574024585786e-05,
      "loss": 1.1521,
      "step": 4200
    },
    {
      "epoch": 0.0,
      "grad_norm": 42.0,
      "learning_rate": 1.915196864421878e-05,
      "loss": 1.1438,
      "step": 4300
    },
    {
      "epoch": 0.0,
      "grad_norm": 16.125,
      "learning_rate": 1.959736326385177e-05,
      "loss": 1.1887,
      "step": 4400
    },
    {
      "epoch": 0.0,
      "grad_norm": 38.75,
      "learning_rate": 2.004275788348477e-05,
      "loss": 1.2946,
      "step": 4500
    },
    {
      "epoch": 0.0,
      "grad_norm": 99.5,
      "learning_rate": 2.0488152503117764e-05,
      "loss": 1.2574,
      "step": 4600
    },
    {
      "epoch": 0.0,
      "grad_norm": 65.5,
      "learning_rate": 2.093354712275076e-05,
      "loss": 1.1517,
      "step": 4700
    },
    {
      "epoch": 0.0,
      "grad_norm": 79.5,
      "learning_rate": 2.1378941742383753e-05,
      "loss": 1.187,
      "step": 4800
    },
    {
      "epoch": 0.0,
      "grad_norm": 162.0,
      "learning_rate": 2.1824336362016748e-05,
      "loss": 0.9974,
      "step": 4900
    },
    {
      "epoch": 0.0,
      "grad_norm": 48.5,
      "learning_rate": 2.2269730981649742e-05,
      "loss": 1.2278,
      "step": 5000
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.09423828125,
      "learning_rate": 2.271512560128274e-05,
      "loss": 1.2128,
      "step": 5100
    },
    {
      "epoch": 0.0,
      "grad_norm": 34.75,
      "learning_rate": 2.316052022091573e-05,
      "loss": 1.1712,
      "step": 5200
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.01483154296875,
      "learning_rate": 2.3605914840548726e-05,
      "loss": 0.9784,
      "step": 5300
    },
    {
      "epoch": 0.0,
      "grad_norm": 89.0,
      "learning_rate": 2.4051309460181724e-05,
      "loss": 1.1382,
      "step": 5400
    },
    {
      "epoch": 0.0,
      "grad_norm": 59.5,
      "learning_rate": 2.449670407981472e-05,
      "loss": 1.291,
      "step": 5500
    },
    {
      "epoch": 0.0,
      "grad_norm": 29.5,
      "learning_rate": 2.494209869944771e-05,
      "loss": 1.1317,
      "step": 5600
    },
    {
      "epoch": 0.01,
      "grad_norm": 24.125,
      "learning_rate": 2.5387493319080707e-05,
      "loss": 1.1689,
      "step": 5700
    },
    {
      "epoch": 0.01,
      "grad_norm": 34.0,
      "learning_rate": 2.5832887938713702e-05,
      "loss": 1.2427,
      "step": 5800
    },
    {
      "epoch": 0.01,
      "grad_norm": 34.5,
      "learning_rate": 2.6278282558346697e-05,
      "loss": 1.353,
      "step": 5900
    },
    {
      "epoch": 0.01,
      "grad_norm": 32.25,
      "learning_rate": 2.6723677177979688e-05,
      "loss": 1.1593,
      "step": 6000
    },
    {
      "epoch": 0.01,
      "grad_norm": 33.75,
      "learning_rate": 2.716907179761269e-05,
      "loss": 1.1391,
      "step": 6100
    },
    {
      "epoch": 0.01,
      "grad_norm": 82.0,
      "learning_rate": 2.7614466417245684e-05,
      "loss": 1.1999,
      "step": 6200
    },
    {
      "epoch": 0.01,
      "grad_norm": 25.375,
      "learning_rate": 2.8059861036878675e-05,
      "loss": 1.0865,
      "step": 6300
    },
    {
      "epoch": 0.01,
      "grad_norm": 34.25,
      "learning_rate": 2.850525565651167e-05,
      "loss": 1.0761,
      "step": 6400
    },
    {
      "epoch": 0.01,
      "grad_norm": 44.0,
      "learning_rate": 2.8950650276144664e-05,
      "loss": 1.1187,
      "step": 6500
    },
    {
      "epoch": 0.01,
      "grad_norm": 38.25,
      "learning_rate": 2.939604489577766e-05,
      "loss": 1.0824,
      "step": 6600
    },
    {
      "epoch": 0.01,
      "grad_norm": 49.75,
      "learning_rate": 2.9841439515410657e-05,
      "loss": 1.0826,
      "step": 6700
    },
    {
      "epoch": 0.01,
      "grad_norm": 47.0,
      "learning_rate": 3.028683413504365e-05,
      "loss": 1.1387,
      "step": 6800
    },
    {
      "epoch": 0.01,
      "grad_norm": 61.0,
      "learning_rate": 3.073222875467664e-05,
      "loss": 1.1661,
      "step": 6900
    },
    {
      "epoch": 0.01,
      "grad_norm": 31.625,
      "learning_rate": 3.117762337430964e-05,
      "loss": 1.2022,
      "step": 7000
    },
    {
      "epoch": 0.01,
      "grad_norm": 38.75,
      "learning_rate": 3.162301799394263e-05,
      "loss": 1.2461,
      "step": 7100
    },
    {
      "epoch": 0.01,
      "grad_norm": 28.5,
      "learning_rate": 3.206841261357563e-05,
      "loss": 1.2583,
      "step": 7200
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.0,
      "learning_rate": 3.251380723320863e-05,
      "loss": 1.1094,
      "step": 7300
    },
    {
      "epoch": 0.01,
      "grad_norm": 68.0,
      "learning_rate": 3.295920185284162e-05,
      "loss": 1.2244,
      "step": 7400
    },
    {
      "epoch": 0.01,
      "grad_norm": 42.0,
      "learning_rate": 3.3404596472474617e-05,
      "loss": 1.1486,
      "step": 7500
    },
    {
      "epoch": 0.01,
      "grad_norm": 344.0,
      "learning_rate": 3.384999109210761e-05,
      "loss": 1.2049,
      "step": 7600
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.6875,
      "learning_rate": 3.42953857117406e-05,
      "loss": 1.0226,
      "step": 7700
    },
    {
      "epoch": 0.01,
      "grad_norm": 43.75,
      "learning_rate": 3.4740780331373604e-05,
      "loss": 1.2662,
      "step": 7800
    },
    {
      "epoch": 0.01,
      "grad_norm": 51.5,
      "learning_rate": 3.5186174951006595e-05,
      "loss": 1.1029,
      "step": 7900
    },
    {
      "epoch": 0.01,
      "grad_norm": 36.25,
      "learning_rate": 3.5631569570639586e-05,
      "loss": 1.1623,
      "step": 8000
    },
    {
      "epoch": 0.01,
      "grad_norm": 32.75,
      "learning_rate": 3.6076964190272584e-05,
      "loss": 1.0682,
      "step": 8100
    },
    {
      "epoch": 0.01,
      "grad_norm": 57.75,
      "learning_rate": 3.6522358809905575e-05,
      "loss": 1.1232,
      "step": 8200
    },
    {
      "epoch": 0.01,
      "grad_norm": 27.125,
      "learning_rate": 3.696775342953857e-05,
      "loss": 1.1474,
      "step": 8300
    },
    {
      "epoch": 0.01,
      "grad_norm": 93.0,
      "learning_rate": 3.741314804917157e-05,
      "loss": 1.0526,
      "step": 8400
    },
    {
      "epoch": 0.01,
      "grad_norm": 20.125,
      "learning_rate": 3.785854266880456e-05,
      "loss": 1.2,
      "step": 8500
    },
    {
      "epoch": 0.01,
      "grad_norm": 46.75,
      "learning_rate": 3.830393728843756e-05,
      "loss": 1.1716,
      "step": 8600
    },
    {
      "epoch": 0.01,
      "grad_norm": 54.75,
      "learning_rate": 3.874933190807055e-05,
      "loss": 1.042,
      "step": 8700
    },
    {
      "epoch": 0.01,
      "grad_norm": 51.0,
      "learning_rate": 3.919472652770354e-05,
      "loss": 1.1756,
      "step": 8800
    },
    {
      "epoch": 0.01,
      "grad_norm": 48.25,
      "learning_rate": 3.964012114733654e-05,
      "loss": 1.2597,
      "step": 8900
    },
    {
      "epoch": 0.01,
      "grad_norm": 14.75,
      "learning_rate": 4.008551576696954e-05,
      "loss": 1.0449,
      "step": 9000
    },
    {
      "epoch": 0.01,
      "grad_norm": 18.25,
      "learning_rate": 4.0530910386602536e-05,
      "loss": 1.2163,
      "step": 9100
    },
    {
      "epoch": 0.01,
      "grad_norm": 80.0,
      "learning_rate": 4.097630500623553e-05,
      "loss": 1.2302,
      "step": 9200
    },
    {
      "epoch": 0.01,
      "grad_norm": 32.0,
      "learning_rate": 4.142169962586852e-05,
      "loss": 1.1843,
      "step": 9300
    },
    {
      "epoch": 0.01,
      "grad_norm": 44.25,
      "learning_rate": 4.186709424550152e-05,
      "loss": 1.0384,
      "step": 9400
    },
    {
      "epoch": 0.01,
      "grad_norm": 66.5,
      "learning_rate": 4.231248886513451e-05,
      "loss": 1.0315,
      "step": 9500
    },
    {
      "epoch": 0.01,
      "grad_norm": 72.5,
      "learning_rate": 4.2757883484767506e-05,
      "loss": 1.1746,
      "step": 9600
    },
    {
      "epoch": 0.01,
      "grad_norm": 37.5,
      "learning_rate": 4.3203278104400504e-05,
      "loss": 1.2093,
      "step": 9700
    },
    {
      "epoch": 0.01,
      "grad_norm": 81.5,
      "learning_rate": 4.3648672724033495e-05,
      "loss": 1.3342,
      "step": 9800
    },
    {
      "epoch": 0.01,
      "grad_norm": 88.5,
      "learning_rate": 4.409406734366649e-05,
      "loss": 1.1162,
      "step": 9900
    },
    {
      "epoch": 0.01,
      "grad_norm": 356.0,
      "learning_rate": 4.4539461963299484e-05,
      "loss": 0.8431,
      "step": 10000
    },
    {
      "epoch": 0.01,
      "grad_norm": 75.5,
      "learning_rate": 4.498485658293248e-05,
      "loss": 1.15,
      "step": 10100
    },
    {
      "epoch": 0.01,
      "grad_norm": 51.0,
      "learning_rate": 4.543025120256548e-05,
      "loss": 1.1635,
      "step": 10200
    },
    {
      "epoch": 0.01,
      "grad_norm": 136.0,
      "learning_rate": 4.587564582219847e-05,
      "loss": 1.1175,
      "step": 10300
    },
    {
      "epoch": 0.01,
      "grad_norm": 180.0,
      "learning_rate": 4.632104044183146e-05,
      "loss": 1.0931,
      "step": 10400
    },
    {
      "epoch": 0.01,
      "grad_norm": 107.5,
      "learning_rate": 4.676643506146446e-05,
      "loss": 1.3118,
      "step": 10500
    },
    {
      "epoch": 0.01,
      "grad_norm": 34.0,
      "learning_rate": 4.721182968109745e-05,
      "loss": 1.0301,
      "step": 10600
    },
    {
      "epoch": 0.01,
      "grad_norm": 33.75,
      "learning_rate": 4.765722430073045e-05,
      "loss": 0.9912,
      "step": 10700
    },
    {
      "epoch": 0.01,
      "grad_norm": 48.5,
      "learning_rate": 4.810261892036345e-05,
      "loss": 1.1649,
      "step": 10800
    },
    {
      "epoch": 0.01,
      "grad_norm": 88.5,
      "learning_rate": 4.854801353999644e-05,
      "loss": 1.4056,
      "step": 10900
    },
    {
      "epoch": 0.01,
      "grad_norm": 143.0,
      "learning_rate": 4.899340815962944e-05,
      "loss": 1.1315,
      "step": 11000
    },
    {
      "epoch": 0.01,
      "grad_norm": 26.5,
      "learning_rate": 4.943880277926243e-05,
      "loss": 1.231,
      "step": 11100
    },
    {
      "epoch": 0.01,
      "grad_norm": 47.75,
      "learning_rate": 4.988419739889542e-05,
      "loss": 1.2519,
      "step": 11200
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.5625,
      "learning_rate": 4.999667068583872e-05,
      "loss": 1.1036,
      "step": 11300
    },
    {
      "epoch": 0.01,
      "grad_norm": 72.5,
      "learning_rate": 4.99921716126478e-05,
      "loss": 1.0886,
      "step": 11400
    },
    {
      "epoch": 0.01,
      "grad_norm": 102.5,
      "learning_rate": 4.998767253945688e-05,
      "loss": 1.1742,
      "step": 11500
    },
    {
      "epoch": 0.01,
      "grad_norm": 127.0,
      "learning_rate": 4.9983173466265954e-05,
      "loss": 1.2533,
      "step": 11600
    },
    {
      "epoch": 0.01,
      "grad_norm": 221.0,
      "learning_rate": 4.997867439307503e-05,
      "loss": 1.1356,
      "step": 11700
    },
    {
      "epoch": 0.01,
      "grad_norm": 132.0,
      "learning_rate": 4.997417531988411e-05,
      "loss": 1.0534,
      "step": 11800
    },
    {
      "epoch": 0.01,
      "grad_norm": 30.625,
      "learning_rate": 4.996967624669318e-05,
      "loss": 0.936,
      "step": 11900
    },
    {
      "epoch": 0.01,
      "grad_norm": 56.5,
      "learning_rate": 4.9965177173502256e-05,
      "loss": 1.0134,
      "step": 12000
    }
  ],
  "logging_steps": 100,
  "max_steps": 1122566,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 6000,
  "total_flos": 1.8994803365154816e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}