{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100000000,
  "global_step": 100000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 1.43193519115448,
      "learning_rate": 9.9003e-06,
      "loss": 1.0381,
      "step": 1000
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.9174104928970337,
      "learning_rate": 9.8003e-06,
      "loss": 1.0262,
      "step": 2000
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5657432079315186,
      "learning_rate": 9.700300000000001e-06,
      "loss": 1.024,
      "step": 3000
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.5156095027923584,
      "learning_rate": 9.600300000000002e-06,
      "loss": 1.0186,
      "step": 4000
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.6368006467819214,
      "learning_rate": 9.500300000000001e-06,
      "loss": 1.0183,
      "step": 5000
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.562856674194336,
      "learning_rate": 9.4003e-06,
      "loss": 1.0194,
      "step": 6000
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.552268147468567,
      "learning_rate": 9.300300000000001e-06,
      "loss": 1.0222,
      "step": 7000
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.5083566904067993,
      "learning_rate": 9.2003e-06,
      "loss": 1.0152,
      "step": 8000
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.4872832298278809,
      "learning_rate": 9.100300000000002e-06,
      "loss": 1.0183,
      "step": 9000
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.5950316190719604,
      "learning_rate": 9.000300000000001e-06,
      "loss": 1.0233,
      "step": 10000
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5159080028533936,
      "learning_rate": 8.900400000000002e-06,
      "loss": 1.0248,
      "step": 11000
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.7614279985427856,
      "learning_rate": 8.800400000000001e-06,
      "loss": 1.0197,
      "step": 12000
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.6475117206573486,
      "learning_rate": 8.7005e-06,
      "loss": 1.0232,
      "step": 13000
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.6129435300827026,
      "learning_rate": 8.600600000000002e-06,
      "loss": 1.027,
      "step": 14000
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.5888806581497192,
      "learning_rate": 8.500600000000001e-06,
      "loss": 1.0169,
      "step": 15000
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.8314592838287354,
      "learning_rate": 8.4007e-06,
      "loss": 1.0274,
      "step": 16000
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.8506577014923096,
      "learning_rate": 8.3007e-06,
      "loss": 1.0241,
      "step": 17000
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.5969544649124146,
      "learning_rate": 8.200800000000001e-06,
      "loss": 1.0268,
      "step": 18000
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.716140627861023,
      "learning_rate": 8.1008e-06,
      "loss": 1.028,
      "step": 19000
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.5300655364990234,
      "learning_rate": 8.000800000000001e-06,
      "loss": 1.0289,
      "step": 20000
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.258260488510132,
      "learning_rate": 7.900900000000001e-06,
      "loss": 1.0238,
      "step": 21000
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6213842630386353,
      "learning_rate": 7.801e-06,
      "loss": 1.0288,
      "step": 22000
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.9307907819747925,
      "learning_rate": 7.701000000000001e-06,
      "loss": 1.0301,
      "step": 23000
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.6679686307907104,
      "learning_rate": 7.601000000000001e-06,
      "loss": 1.0373,
      "step": 24000
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.8511743545532227,
      "learning_rate": 7.501000000000001e-06,
      "loss": 1.0294,
      "step": 25000
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.0238113403320312,
      "learning_rate": 7.401100000000001e-06,
      "loss": 1.0318,
      "step": 26000
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.9416135549545288,
      "learning_rate": 7.301100000000001e-06,
      "loss": 1.0285,
      "step": 27000
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.8442825078964233,
      "learning_rate": 7.201100000000001e-06,
      "loss": 1.0356,
      "step": 28000
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.7065600156784058,
      "learning_rate": 7.1011000000000005e-06,
      "loss": 1.0261,
      "step": 29000
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.6564326286315918,
      "learning_rate": 7.0012e-06,
      "loss": 1.029,
      "step": 30000
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.886237382888794,
      "learning_rate": 6.901300000000001e-06,
      "loss": 1.0321,
      "step": 31000
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.6222507953643799,
      "learning_rate": 6.8013000000000004e-06,
      "loss": 1.0359,
      "step": 32000
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.8137024641036987,
      "learning_rate": 6.7013000000000005e-06,
      "loss": 1.0423,
      "step": 33000
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.7678865194320679,
      "learning_rate": 6.601300000000001e-06,
      "loss": 1.0342,
      "step": 34000
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.7596269845962524,
      "learning_rate": 6.501400000000001e-06,
      "loss": 1.0359,
      "step": 35000
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.9185104370117188,
      "learning_rate": 6.401400000000001e-06,
      "loss": 1.0355,
      "step": 36000
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.8480957746505737,
      "learning_rate": 6.301400000000001e-06,
      "loss": 1.0405,
      "step": 37000
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.2403833866119385,
      "learning_rate": 6.2014000000000005e-06,
      "loss": 1.0366,
      "step": 38000
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.678298830986023,
      "learning_rate": 6.1015e-06,
      "loss": 1.0396,
      "step": 39000
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.8097726106643677,
      "learning_rate": 6.0015e-06,
      "loss": 1.032,
      "step": 40000
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.9577500820159912,
      "learning_rate": 5.9016e-06,
      "loss": 1.034,
      "step": 41000
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.1903908252716064,
      "learning_rate": 5.8016000000000005e-06,
      "loss": 1.0471,
      "step": 42000
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.9866857528686523,
      "learning_rate": 5.7017e-06,
      "loss": 1.0392,
      "step": 43000
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.159879207611084,
      "learning_rate": 5.6017e-06,
      "loss": 1.0386,
      "step": 44000
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.0432796478271484,
      "learning_rate": 5.5018e-06,
      "loss": 1.0388,
      "step": 45000
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.1784613132476807,
      "learning_rate": 5.4019e-06,
      "loss": 1.0482,
      "step": 46000
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.118239402770996,
      "learning_rate": 5.301900000000001e-06,
      "loss": 1.0478,
      "step": 47000
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.4094135761260986,
      "learning_rate": 5.201900000000001e-06,
      "loss": 1.0492,
      "step": 48000
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.2859714031219482,
      "learning_rate": 5.101900000000001e-06,
      "loss": 1.0456,
      "step": 49000
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.0180344581604004,
      "learning_rate": 5.0019e-06,
      "loss": 1.0467,
      "step": 50000
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.944109559059143,
      "learning_rate": 4.9019000000000005e-06,
      "loss": 1.0449,
      "step": 51000
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.5860376358032227,
      "learning_rate": 4.802100000000001e-06,
      "loss": 1.0429,
      "step": 52000
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.394335985183716,
      "learning_rate": 4.7021e-06,
      "loss": 1.0438,
      "step": 53000
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.9749242067337036,
      "learning_rate": 4.6021e-06,
      "loss": 1.0297,
      "step": 54000
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.190340518951416,
      "learning_rate": 4.5021000000000005e-06,
      "loss": 1.0371,
      "step": 55000
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.4714150428771973,
      "learning_rate": 4.4022e-06,
      "loss": 1.0383,
      "step": 56000
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.1997413635253906,
      "learning_rate": 4.3023e-06,
      "loss": 1.0388,
      "step": 57000
    },
    {
      "epoch": 0.58,
      "grad_norm": 2.3572850227355957,
      "learning_rate": 4.2023e-06,
      "loss": 1.0385,
      "step": 58000
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.820908546447754,
      "learning_rate": 4.1023000000000005e-06,
      "loss": 1.0463,
      "step": 59000
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.5761115550994873,
      "learning_rate": 4.0023e-06,
      "loss": 1.0319,
      "step": 60000
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.0554027557373047,
      "learning_rate": 3.902300000000001e-06,
      "loss": 1.0402,
      "step": 61000
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.0767500400543213,
      "learning_rate": 3.8023000000000004e-06,
      "loss": 1.045,
      "step": 62000
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.9431368112564087,
      "learning_rate": 3.7024000000000003e-06,
      "loss": 1.0367,
      "step": 63000
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6991102695465088,
      "learning_rate": 3.6025000000000002e-06,
      "loss": 1.031,
      "step": 64000
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.0003433227539062,
      "learning_rate": 3.5025000000000003e-06,
      "loss": 1.0336,
      "step": 65000
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.5492982864379883,
      "learning_rate": 3.4025000000000005e-06,
      "loss": 1.0363,
      "step": 66000
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.1553924083709717,
      "learning_rate": 3.3025e-06,
      "loss": 1.0301,
      "step": 67000
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.9345446825027466,
      "learning_rate": 3.2025000000000003e-06,
      "loss": 1.0322,
      "step": 68000
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.102257251739502,
      "learning_rate": 3.1026e-06,
      "loss": 1.0257,
      "step": 69000
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.8478143215179443,
      "learning_rate": 3.0026000000000007e-06,
      "loss": 1.0308,
      "step": 70000
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.9828852415084839,
      "learning_rate": 2.9026000000000004e-06,
      "loss": 1.0309,
      "step": 71000
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.127558708190918,
      "learning_rate": 2.8027000000000003e-06,
      "loss": 1.0334,
      "step": 72000
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.8984837532043457,
      "learning_rate": 2.7028000000000006e-06,
      "loss": 1.0338,
      "step": 73000
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.105077028274536,
      "learning_rate": 2.6028000000000003e-06,
      "loss": 1.0217,
      "step": 74000
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.1777162551879883,
      "learning_rate": 2.5028000000000004e-06,
      "loss": 1.0331,
      "step": 75000
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.2614808082580566,
      "learning_rate": 2.4028e-06,
      "loss": 1.0288,
      "step": 76000
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.123326063156128,
      "learning_rate": 2.3029e-06,
      "loss": 1.0311,
      "step": 77000
    },
    {
      "epoch": 0.78,
      "grad_norm": 2.9798057079315186,
      "learning_rate": 2.2029e-06,
      "loss": 1.0438,
      "step": 78000
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.481782913208008,
      "learning_rate": 2.1029000000000002e-06,
      "loss": 1.0308,
      "step": 79000
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.476494073867798,
      "learning_rate": 2.003e-06,
      "loss": 1.0315,
      "step": 80000
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.2931506633758545,
      "learning_rate": 1.9031000000000003e-06,
      "loss": 1.0232,
      "step": 81000
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.8126013278961182,
      "learning_rate": 1.8031000000000002e-06,
      "loss": 1.028,
      "step": 82000
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.0132505893707275,
      "learning_rate": 1.7031e-06,
      "loss": 1.0293,
      "step": 83000
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.030576467514038,
      "learning_rate": 1.6031000000000002e-06,
      "loss": 1.0266,
      "step": 84000
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.3928701877593994,
      "learning_rate": 1.5031e-06,
      "loss": 1.0303,
      "step": 85000
    },
    {
      "epoch": 0.86,
      "grad_norm": 2.091346502304077,
      "learning_rate": 1.4033000000000002e-06,
      "loss": 1.0303,
      "step": 86000
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.9686309099197388,
      "learning_rate": 1.3033e-06,
      "loss": 1.0339,
      "step": 87000
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.8884007930755615,
      "learning_rate": 1.2034000000000002e-06,
      "loss": 1.0289,
      "step": 88000
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.19309663772583,
      "learning_rate": 1.1034e-06,
      "loss": 1.0257,
      "step": 89000
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.507934808731079,
      "learning_rate": 1.0034e-06,
      "loss": 1.0228,
      "step": 90000
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.336470603942871,
      "learning_rate": 9.034000000000001e-07,
      "loss": 1.0232,
      "step": 91000
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.7181164026260376,
      "learning_rate": 8.034e-07,
      "loss": 1.0283,
      "step": 92000
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.9081435203552246,
      "learning_rate": 7.034e-07,
      "loss": 1.0235,
      "step": 93000
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.7745755910873413,
      "learning_rate": 6.034e-07,
      "loss": 1.0254,
      "step": 94000
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.0873942375183105,
      "learning_rate": 5.035e-07,
      "loss": 1.0323,
      "step": 95000
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.1039388179779053,
      "learning_rate": 4.035e-07,
      "loss": 1.0247,
      "step": 96000
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.9102365970611572,
      "learning_rate": 3.0360000000000005e-07,
      "loss": 1.0313,
      "step": 97000
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.2324767112731934,
      "learning_rate": 2.037e-07,
      "loss": 1.0193,
      "step": 98000
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.005739450454712,
      "learning_rate": 1.0370000000000002e-07,
      "loss": 1.0318,
      "step": 99000
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.0899102687835693,
      "learning_rate": 3.8e-09,
      "loss": 1.0217,
      "step": 100000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 20000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0909051256832e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}