{
  "best_metric": 0.14040808379650116,
  "best_model_checkpoint": "./results/checkpoint-51064",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 51064,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 1.986944487962818e-05,
      "loss": 0.2496,
      "step": 500
    },
    {
      "epoch": 0.04,
      "learning_rate": 1.9738889759256358e-05,
      "loss": 0.208,
      "step": 1000
    },
    {
      "epoch": 0.06,
      "learning_rate": 1.9608334638884537e-05,
      "loss": 0.1884,
      "step": 1500
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.9477779518512717e-05,
      "loss": 0.1955,
      "step": 2000
    },
    {
      "epoch": 0.1,
      "learning_rate": 1.9347224398140897e-05,
      "loss": 0.2005,
      "step": 2500
    },
    {
      "epoch": 0.12,
      "learning_rate": 1.9216669277769077e-05,
      "loss": 0.1625,
      "step": 3000
    },
    {
      "epoch": 0.14,
      "learning_rate": 1.9086114157397256e-05,
      "loss": 0.18,
      "step": 3500
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.8955559037025433e-05,
      "loss": 0.153,
      "step": 4000
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.8825003916653613e-05,
      "loss": 0.1996,
      "step": 4500
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.8694448796281792e-05,
      "loss": 0.1643,
      "step": 5000
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.856389367590997e-05,
      "loss": 0.2022,
      "step": 5500
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.843333855553815e-05,
      "loss": 0.1706,
      "step": 6000
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.8302783435166328e-05,
      "loss": 0.1895,
      "step": 6500
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.8172228314794508e-05,
      "loss": 0.1733,
      "step": 7000
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.8041673194422688e-05,
      "loss": 0.1938,
      "step": 7500
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.7911118074050867e-05,
      "loss": 0.1786,
      "step": 8000
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.7780562953679047e-05,
      "loss": 0.19,
      "step": 8500
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.7650007833307223e-05,
      "loss": 0.1908,
      "step": 9000
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.7519452712935403e-05,
      "loss": 0.1755,
      "step": 9500
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.7388897592563583e-05,
      "loss": 0.1536,
      "step": 10000
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.725834247219176e-05,
      "loss": 0.1574,
      "step": 10500
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.712778735181994e-05,
      "loss": 0.1472,
      "step": 11000
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.699723223144812e-05,
      "loss": 0.1936,
      "step": 11500
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.68666771110763e-05,
      "loss": 0.1746,
      "step": 12000
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.6736121990704478e-05,
      "loss": 0.1731,
      "step": 12500
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.6605566870332658e-05,
      "loss": 0.1927,
      "step": 13000
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.6475011749960834e-05,
      "loss": 0.1946,
      "step": 13500
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.6344456629589014e-05,
      "loss": 0.1636,
      "step": 14000
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.6213901509217194e-05,
      "loss": 0.2385,
      "step": 14500
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.608334638884537e-05,
      "loss": 0.2137,
      "step": 15000
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.595279126847355e-05,
      "loss": 0.1736,
      "step": 15500
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.582223614810173e-05,
      "loss": 0.2196,
      "step": 16000
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.569168102772991e-05,
      "loss": 0.1878,
      "step": 16500
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.556112590735809e-05,
      "loss": 0.1742,
      "step": 17000
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.543057078698627e-05,
      "loss": 0.1907,
      "step": 17500
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.5300015666614445e-05,
      "loss": 0.2048,
      "step": 18000
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.5169460546242625e-05,
      "loss": 0.1777,
      "step": 18500
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.5038905425870805e-05,
      "loss": 0.213,
      "step": 19000
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.4908350305498983e-05,
      "loss": 0.1941,
      "step": 19500
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.4777795185127163e-05,
      "loss": 0.1525,
      "step": 20000
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.4647240064755342e-05,
      "loss": 0.1638,
      "step": 20500
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.4516684944383519e-05,
      "loss": 0.174,
      "step": 21000
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.4386129824011698e-05,
      "loss": 0.1693,
      "step": 21500
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.4255574703639878e-05,
      "loss": 0.1926,
      "step": 22000
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.4125019583268056e-05,
      "loss": 0.1915,
      "step": 22500
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.3994464462896236e-05,
      "loss": 0.1671,
      "step": 23000
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.3863909342524416e-05,
      "loss": 0.1572,
      "step": 23500
    },
    {
      "epoch": 0.94,
      "learning_rate": 1.3733354222152594e-05,
      "loss": 0.1613,
      "step": 24000
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.3602799101780773e-05,
      "loss": 0.165,
      "step": 24500
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.3472243981408951e-05,
      "loss": 0.1739,
      "step": 25000
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.3341688861037131e-05,
      "loss": 0.1611,
      "step": 25500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.17089270055294037,
      "eval_runtime": 899.1592,
      "eval_samples_per_second": 35.494,
      "eval_steps_per_second": 7.099,
      "step": 25532
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.321113374066531e-05,
      "loss": 0.1776,
      "step": 26000
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.3080578620293489e-05,
      "loss": 0.1759,
      "step": 26500
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.2950023499921669e-05,
      "loss": 0.1544,
      "step": 27000
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.2819468379549847e-05,
      "loss": 0.1691,
      "step": 27500
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.2688913259178027e-05,
      "loss": 0.1866,
      "step": 28000
    },
    {
      "epoch": 1.12,
      "learning_rate": 1.2558358138806206e-05,
      "loss": 0.1833,
      "step": 28500
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.2427803018434383e-05,
      "loss": 0.1791,
      "step": 29000
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.2297247898062562e-05,
      "loss": 0.173,
      "step": 29500
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.2166692777690742e-05,
      "loss": 0.2536,
      "step": 30000
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.203613765731892e-05,
      "loss": 0.2705,
      "step": 30500
    },
    {
      "epoch": 1.21,
      "learning_rate": 1.19055825369471e-05,
      "loss": 0.2602,
      "step": 31000
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.177502741657528e-05,
      "loss": 0.1628,
      "step": 31500
    },
    {
      "epoch": 1.25,
      "learning_rate": 1.1644472296203458e-05,
      "loss": 0.249,
      "step": 32000
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.1513917175831637e-05,
      "loss": 0.2991,
      "step": 32500
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.1383362055459817e-05,
      "loss": 0.3618,
      "step": 33000
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.1252806935087994e-05,
      "loss": 0.3487,
      "step": 33500
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.1122251814716173e-05,
      "loss": 0.3179,
      "step": 34000
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.0991696694344353e-05,
      "loss": 0.3457,
      "step": 34500
    },
    {
      "epoch": 1.37,
      "learning_rate": 1.0861141573972531e-05,
      "loss": 0.3574,
      "step": 35000
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.073058645360071e-05,
      "loss": 0.3486,
      "step": 35500
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.060003133322889e-05,
      "loss": 0.2576,
      "step": 36000
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.0469476212857069e-05,
      "loss": 0.1753,
      "step": 36500
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.0338921092485248e-05,
      "loss": 0.2012,
      "step": 37000
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.0208365972113428e-05,
      "loss": 0.2095,
      "step": 37500
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.0077810851741604e-05,
      "loss": 0.1966,
      "step": 38000
    },
    {
      "epoch": 1.51,
      "learning_rate": 9.947255731369784e-06,
      "loss": 0.1492,
      "step": 38500
    },
    {
      "epoch": 1.53,
      "learning_rate": 9.816700610997964e-06,
      "loss": 0.1663,
      "step": 39000
    },
    {
      "epoch": 1.55,
      "learning_rate": 9.686145490626144e-06,
      "loss": 0.158,
      "step": 39500
    },
    {
      "epoch": 1.57,
      "learning_rate": 9.555590370254322e-06,
      "loss": 0.1635,
      "step": 40000
    },
    {
      "epoch": 1.59,
      "learning_rate": 9.425035249882501e-06,
      "loss": 0.1605,
      "step": 40500
    },
    {
      "epoch": 1.61,
      "learning_rate": 9.29448012951068e-06,
      "loss": 0.1479,
      "step": 41000
    },
    {
      "epoch": 1.63,
      "learning_rate": 9.16392500913886e-06,
      "loss": 0.1555,
      "step": 41500
    },
    {
      "epoch": 1.64,
      "learning_rate": 9.033369888767039e-06,
      "loss": 0.1657,
      "step": 42000
    },
    {
      "epoch": 1.66,
      "learning_rate": 8.902814768395217e-06,
      "loss": 0.1322,
      "step": 42500
    },
    {
      "epoch": 1.68,
      "learning_rate": 8.772259648023395e-06,
      "loss": 0.1441,
      "step": 43000
    },
    {
      "epoch": 1.7,
      "learning_rate": 8.641704527651575e-06,
      "loss": 0.1514,
      "step": 43500
    },
    {
      "epoch": 1.72,
      "learning_rate": 8.511149407279755e-06,
      "loss": 0.1556,
      "step": 44000
    },
    {
      "epoch": 1.74,
      "learning_rate": 8.380594286907934e-06,
      "loss": 0.2028,
      "step": 44500
    },
    {
      "epoch": 1.76,
      "learning_rate": 8.250039166536112e-06,
      "loss": 0.1538,
      "step": 45000
    },
    {
      "epoch": 1.78,
      "learning_rate": 8.11948404616429e-06,
      "loss": 0.1804,
      "step": 45500
    },
    {
      "epoch": 1.8,
      "learning_rate": 7.98892892579247e-06,
      "loss": 0.146,
      "step": 46000
    },
    {
      "epoch": 1.82,
      "learning_rate": 7.85837380542065e-06,
      "loss": 0.1538,
      "step": 46500
    },
    {
      "epoch": 1.84,
      "learning_rate": 7.727818685048828e-06,
      "loss": 0.1476,
      "step": 47000
    },
    {
      "epoch": 1.86,
      "learning_rate": 7.597263564677008e-06,
      "loss": 0.1766,
      "step": 47500
    },
    {
      "epoch": 1.88,
      "learning_rate": 7.4667084443051865e-06,
      "loss": 0.1807,
      "step": 48000
    },
    {
      "epoch": 1.9,
      "learning_rate": 7.336153323933365e-06,
      "loss": 0.1416,
      "step": 48500
    },
    {
      "epoch": 1.92,
      "learning_rate": 7.205598203561544e-06,
      "loss": 0.1453,
      "step": 49000
    },
    {
      "epoch": 1.94,
      "learning_rate": 7.075043083189723e-06,
      "loss": 0.1349,
      "step": 49500
    },
    {
      "epoch": 1.96,
      "learning_rate": 6.944487962817902e-06,
      "loss": 0.1852,
      "step": 50000
    },
    {
      "epoch": 1.98,
      "learning_rate": 6.813932842446082e-06,
      "loss": 0.1599,
      "step": 50500
    },
    {
      "epoch": 2.0,
      "learning_rate": 6.68337772207426e-06,
      "loss": 0.1184,
      "step": 51000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.14040808379650116,
      "eval_runtime": 896.823,
      "eval_samples_per_second": 35.587,
      "eval_steps_per_second": 7.117,
      "step": 51064
    }
  ],
  "logging_steps": 500,
  "max_steps": 76596,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 6.717540976607232e+16,
  "trial_name": null,
  "trial_params": null
}