{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 16.69195751138088,
  "eval_steps": 500,
  "global_step": 66000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.13,
      "learning_rate": 2.9776845487815763e-05,
      "loss": 0.1875,
      "step": 500
    },
    {
      "epoch": 0.25,
      "learning_rate": 2.955369097563153e-05,
      "loss": 0.1685,
      "step": 1000
    },
    {
      "epoch": 0.38,
      "learning_rate": 2.9330536463447292e-05,
      "loss": 0.1685,
      "step": 1500
    },
    {
      "epoch": 0.51,
      "learning_rate": 2.910738195126306e-05,
      "loss": 0.174,
      "step": 2000
    },
    {
      "epoch": 0.63,
      "learning_rate": 2.8884227439078818e-05,
      "loss": 0.1667,
      "step": 2500
    },
    {
      "epoch": 0.76,
      "learning_rate": 2.866107292689458e-05,
      "loss": 0.1677,
      "step": 3000
    },
    {
      "epoch": 0.89,
      "learning_rate": 2.8437918414710347e-05,
      "loss": 0.1655,
      "step": 3500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.43983822042467136,
      "eval_loss": 0.11440373957157135,
      "eval_runtime": 18.7487,
      "eval_samples_per_second": 52.75,
      "eval_steps_per_second": 2.24,
      "step": 3954
    },
    {
      "epoch": 1.01,
      "learning_rate": 2.821476390252611e-05,
      "loss": 0.1685,
      "step": 4000
    },
    {
      "epoch": 1.14,
      "learning_rate": 2.7991609390341872e-05,
      "loss": 0.1672,
      "step": 4500
    },
    {
      "epoch": 1.26,
      "learning_rate": 2.7768454878157638e-05,
      "loss": 0.1667,
      "step": 5000
    },
    {
      "epoch": 1.39,
      "learning_rate": 2.75453003659734e-05,
      "loss": 0.1734,
      "step": 5500
    },
    {
      "epoch": 1.52,
      "learning_rate": 2.7322145853789167e-05,
      "loss": 0.1704,
      "step": 6000
    },
    {
      "epoch": 1.64,
      "learning_rate": 2.709899134160493e-05,
      "loss": 0.1646,
      "step": 6500
    },
    {
      "epoch": 1.77,
      "learning_rate": 2.687583682942069e-05,
      "loss": 0.1686,
      "step": 7000
    },
    {
      "epoch": 1.9,
      "learning_rate": 2.6652682317236455e-05,
      "loss": 0.2201,
      "step": 7500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.38119312436804853,
      "eval_loss": 0.11238503456115723,
      "eval_runtime": 18.3646,
      "eval_samples_per_second": 53.854,
      "eval_steps_per_second": 2.287,
      "step": 7908
    },
    {
      "epoch": 2.02,
      "learning_rate": 2.6429527805052218e-05,
      "loss": 0.1645,
      "step": 8000
    },
    {
      "epoch": 2.15,
      "learning_rate": 2.6206373292867984e-05,
      "loss": 0.1697,
      "step": 8500
    },
    {
      "epoch": 2.28,
      "learning_rate": 2.5983218780683747e-05,
      "loss": 0.1695,
      "step": 9000
    },
    {
      "epoch": 2.4,
      "learning_rate": 2.576006426849951e-05,
      "loss": 0.167,
      "step": 9500
    },
    {
      "epoch": 2.53,
      "learning_rate": 2.5536909756315275e-05,
      "loss": 0.1717,
      "step": 10000
    },
    {
      "epoch": 2.66,
      "learning_rate": 2.5313755244131038e-05,
      "loss": 0.1694,
      "step": 10500
    },
    {
      "epoch": 2.78,
      "learning_rate": 2.50906007319468e-05,
      "loss": 0.1654,
      "step": 11000
    },
    {
      "epoch": 2.91,
      "learning_rate": 2.4867446219762564e-05,
      "loss": 0.1657,
      "step": 11500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.43983822042467136,
      "eval_loss": 0.11309783160686493,
      "eval_runtime": 18.1147,
      "eval_samples_per_second": 54.597,
      "eval_steps_per_second": 2.319,
      "step": 11862
    },
    {
      "epoch": 3.03,
      "learning_rate": 2.4644291707578326e-05,
      "loss": 0.1693,
      "step": 12000
    },
    {
      "epoch": 3.16,
      "learning_rate": 2.4421137195394092e-05,
      "loss": 0.1697,
      "step": 12500
    },
    {
      "epoch": 3.29,
      "learning_rate": 2.4197982683209855e-05,
      "loss": 0.1698,
      "step": 13000
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.3974828171025618e-05,
      "loss": 0.165,
      "step": 13500
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.3751673658841384e-05,
      "loss": 0.1664,
      "step": 14000
    },
    {
      "epoch": 3.67,
      "learning_rate": 2.3528519146657147e-05,
      "loss": 0.168,
      "step": 14500
    },
    {
      "epoch": 3.79,
      "learning_rate": 2.330536463447291e-05,
      "loss": 0.1639,
      "step": 15000
    },
    {
      "epoch": 3.92,
      "learning_rate": 2.3082210122288672e-05,
      "loss": 0.167,
      "step": 15500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4287158746208291,
      "eval_loss": 0.1112765222787857,
      "eval_runtime": 18.0277,
      "eval_samples_per_second": 54.86,
      "eval_steps_per_second": 2.33,
      "step": 15816
    },
    {
      "epoch": 4.05,
      "learning_rate": 2.2859055610104435e-05,
      "loss": 0.1693,
      "step": 16000
    },
    {
      "epoch": 4.17,
      "learning_rate": 2.26359010979202e-05,
      "loss": 0.1701,
      "step": 16500
    },
    {
      "epoch": 4.3,
      "learning_rate": 2.2412746585735964e-05,
      "loss": 0.1673,
      "step": 17000
    },
    {
      "epoch": 4.43,
      "learning_rate": 2.2189592073551726e-05,
      "loss": 0.1686,
      "step": 17500
    },
    {
      "epoch": 4.55,
      "learning_rate": 2.1966437561367493e-05,
      "loss": 0.1646,
      "step": 18000
    },
    {
      "epoch": 4.68,
      "learning_rate": 2.1743283049183255e-05,
      "loss": 0.1691,
      "step": 18500
    },
    {
      "epoch": 4.81,
      "learning_rate": 2.152012853699902e-05,
      "loss": 0.1674,
      "step": 19000
    },
    {
      "epoch": 4.93,
      "learning_rate": 2.1296974024814784e-05,
      "loss": 0.1669,
      "step": 19500
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4074823053589484,
      "eval_loss": 0.1131151020526886,
      "eval_runtime": 18.562,
      "eval_samples_per_second": 53.281,
      "eval_steps_per_second": 2.263,
      "step": 19770
    },
    {
      "epoch": 5.06,
      "learning_rate": 2.1073819512630543e-05,
      "loss": 0.1677,
      "step": 20000
    },
    {
      "epoch": 5.18,
      "learning_rate": 2.085066500044631e-05,
      "loss": 0.1659,
      "step": 20500
    },
    {
      "epoch": 5.31,
      "learning_rate": 2.0627510488262072e-05,
      "loss": 0.1658,
      "step": 21000
    },
    {
      "epoch": 5.44,
      "learning_rate": 2.0404355976077835e-05,
      "loss": 0.1694,
      "step": 21500
    },
    {
      "epoch": 5.56,
      "learning_rate": 2.01812014638936e-05,
      "loss": 0.1731,
      "step": 22000
    },
    {
      "epoch": 5.69,
      "learning_rate": 1.9958046951709364e-05,
      "loss": 0.1696,
      "step": 22500
    },
    {
      "epoch": 5.82,
      "learning_rate": 1.973489243952513e-05,
      "loss": 0.1669,
      "step": 23000
    },
    {
      "epoch": 5.94,
      "learning_rate": 1.9511737927340893e-05,
      "loss": 0.1648,
      "step": 23500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.43983822042467136,
      "eval_loss": 0.11177950352430344,
      "eval_runtime": 18.1533,
      "eval_samples_per_second": 54.48,
      "eval_steps_per_second": 2.314,
      "step": 23724
    },
    {
      "epoch": 6.07,
      "learning_rate": 1.9288583415156655e-05,
      "loss": 0.1673,
      "step": 24000
    },
    {
      "epoch": 6.2,
      "learning_rate": 1.9065428902972418e-05,
      "loss": 0.1679,
      "step": 24500
    },
    {
      "epoch": 6.32,
      "learning_rate": 1.884227439078818e-05,
      "loss": 0.1668,
      "step": 25000
    },
    {
      "epoch": 6.45,
      "learning_rate": 1.8619119878603947e-05,
      "loss": 0.1649,
      "step": 25500
    },
    {
      "epoch": 6.58,
      "learning_rate": 1.839596536641971e-05,
      "loss": 0.1701,
      "step": 26000
    },
    {
      "epoch": 6.7,
      "learning_rate": 1.8172810854235472e-05,
      "loss": 0.1683,
      "step": 26500
    },
    {
      "epoch": 6.83,
      "learning_rate": 1.794965634205124e-05,
      "loss": 0.1649,
      "step": 27000
    },
    {
      "epoch": 6.95,
      "learning_rate": 1.7726501829867e-05,
      "loss": 0.1686,
      "step": 27500
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.43983822042467136,
      "eval_loss": 0.11043604463338852,
      "eval_runtime": 17.9733,
      "eval_samples_per_second": 55.026,
      "eval_steps_per_second": 2.337,
      "step": 27678
    },
    {
      "epoch": 7.08,
      "learning_rate": 1.7503347317682764e-05,
      "loss": 0.1692,
      "step": 28000
    },
    {
      "epoch": 7.21,
      "learning_rate": 1.728019280549853e-05,
      "loss": 0.166,
      "step": 28500
    },
    {
      "epoch": 7.33,
      "learning_rate": 1.705703829331429e-05,
      "loss": 0.1682,
      "step": 29000
    },
    {
      "epoch": 7.46,
      "learning_rate": 1.6833883781130055e-05,
      "loss": 0.1668,
      "step": 29500
    },
    {
      "epoch": 7.59,
      "learning_rate": 1.6610729268945818e-05,
      "loss": 0.1661,
      "step": 30000
    },
    {
      "epoch": 7.71,
      "learning_rate": 1.638757475676158e-05,
      "loss": 0.1684,
      "step": 30500
    },
    {
      "epoch": 7.84,
      "learning_rate": 1.6164420244577347e-05,
      "loss": 0.1692,
      "step": 31000
    },
    {
      "epoch": 7.97,
      "learning_rate": 1.594126573239311e-05,
      "loss": 0.1659,
      "step": 31500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.4287158746208291,
      "eval_loss": 0.11213327944278717,
      "eval_runtime": 18.4703,
      "eval_samples_per_second": 53.546,
      "eval_steps_per_second": 2.274,
      "step": 31632
    },
    {
      "epoch": 8.09,
      "learning_rate": 1.5718111220208872e-05,
      "loss": 0.1675,
      "step": 32000
    },
    {
      "epoch": 8.22,
      "learning_rate": 1.549495670802464e-05,
      "loss": 0.1662,
      "step": 32500
    },
    {
      "epoch": 8.35,
      "learning_rate": 1.52718021958404e-05,
      "loss": 0.167,
      "step": 33000
    },
    {
      "epoch": 8.47,
      "learning_rate": 1.5048647683656162e-05,
      "loss": 0.1684,
      "step": 33500
    },
    {
      "epoch": 8.6,
      "learning_rate": 1.4825493171471928e-05,
      "loss": 0.1666,
      "step": 34000
    },
    {
      "epoch": 8.73,
      "learning_rate": 1.4602338659287691e-05,
      "loss": 0.1668,
      "step": 34500
    },
    {
      "epoch": 8.85,
      "learning_rate": 1.4379184147103455e-05,
      "loss": 0.1681,
      "step": 35000
    },
    {
      "epoch": 8.98,
      "learning_rate": 1.4156029634919218e-05,
      "loss": 0.1689,
      "step": 35500
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.4236602628918099,
      "eval_loss": 0.11145440489053726,
      "eval_runtime": 18.1816,
      "eval_samples_per_second": 54.396,
      "eval_steps_per_second": 2.31,
      "step": 35586
    },
    {
      "epoch": 9.1,
      "learning_rate": 1.3932875122734983e-05,
      "loss": 0.1662,
      "step": 36000
    },
    {
      "epoch": 9.23,
      "learning_rate": 1.3709720610550745e-05,
      "loss": 0.1663,
      "step": 36500
    },
    {
      "epoch": 9.36,
      "learning_rate": 1.348656609836651e-05,
      "loss": 0.1657,
      "step": 37000
    },
    {
      "epoch": 9.48,
      "learning_rate": 1.3263411586182272e-05,
      "loss": 0.1669,
      "step": 37500
    },
    {
      "epoch": 9.61,
      "learning_rate": 1.3040257073998037e-05,
      "loss": 0.1673,
      "step": 38000
    },
    {
      "epoch": 9.74,
      "learning_rate": 1.2817102561813801e-05,
      "loss": 0.1704,
      "step": 38500
    },
    {
      "epoch": 9.86,
      "learning_rate": 1.2593948049629564e-05,
      "loss": 0.1678,
      "step": 39000
    },
    {
      "epoch": 9.99,
      "learning_rate": 1.2370793537445327e-05,
      "loss": 0.1687,
      "step": 39500
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.43983822042467136,
      "eval_loss": 0.11185000836849213,
      "eval_runtime": 18.0459,
      "eval_samples_per_second": 54.805,
      "eval_steps_per_second": 2.327,
      "step": 39540
    },
    {
      "epoch": 10.12,
      "learning_rate": 1.2147639025261091e-05,
      "loss": 0.1683,
      "step": 40000
    },
    {
      "epoch": 10.24,
      "learning_rate": 1.1924484513076856e-05,
      "loss": 0.167,
      "step": 40500
    },
    {
      "epoch": 10.37,
      "learning_rate": 1.1701330000892618e-05,
      "loss": 0.1673,
      "step": 41000
    },
    {
      "epoch": 10.5,
      "learning_rate": 1.1478175488708381e-05,
      "loss": 0.1682,
      "step": 41500
    },
    {
      "epoch": 10.62,
      "learning_rate": 1.1255020976524145e-05,
      "loss": 0.167,
      "step": 42000
    },
    {
      "epoch": 10.75,
      "learning_rate": 1.103186646433991e-05,
      "loss": 0.1654,
      "step": 42500
    },
    {
      "epoch": 10.88,
      "learning_rate": 1.0808711952155672e-05,
      "loss": 0.164,
      "step": 43000
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.43983822042467136,
      "eval_loss": 0.11140052229166031,
      "eval_runtime": 18.3103,
      "eval_samples_per_second": 54.013,
      "eval_steps_per_second": 2.294,
      "step": 43494
    },
    {
      "epoch": 11.0,
      "learning_rate": 1.0585557439971437e-05,
      "loss": 0.17,
      "step": 43500
    },
    {
      "epoch": 11.13,
      "learning_rate": 1.03624029277872e-05,
      "loss": 0.1677,
      "step": 44000
    },
    {
      "epoch": 11.25,
      "learning_rate": 1.0139248415602964e-05,
      "loss": 0.1663,
      "step": 44500
    },
    {
      "epoch": 11.38,
      "learning_rate": 9.916093903418728e-06,
      "loss": 0.167,
      "step": 45000
    },
    {
      "epoch": 11.51,
      "learning_rate": 9.692939391234491e-06,
      "loss": 0.1679,
      "step": 45500
    },
    {
      "epoch": 11.63,
      "learning_rate": 9.469784879050254e-06,
      "loss": 0.1674,
      "step": 46000
    },
    {
      "epoch": 11.76,
      "learning_rate": 9.246630366866018e-06,
      "loss": 0.1619,
      "step": 46500
    },
    {
      "epoch": 11.89,
      "learning_rate": 9.023475854681783e-06,
      "loss": 0.1701,
      "step": 47000
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.43983822042467136,
      "eval_loss": 0.11146707087755203,
      "eval_runtime": 18.0716,
      "eval_samples_per_second": 54.727,
      "eval_steps_per_second": 2.324,
      "step": 47448
    },
    {
      "epoch": 12.01,
      "learning_rate": 8.800321342497545e-06,
      "loss": 0.167,
      "step": 47500
    },
    {
      "epoch": 12.14,
      "learning_rate": 8.577166830313308e-06,
      "loss": 0.1673,
      "step": 48000
    },
    {
      "epoch": 12.27,
      "learning_rate": 8.354012318129073e-06,
      "loss": 0.1691,
      "step": 48500
    },
    {
      "epoch": 12.39,
      "learning_rate": 8.130857805944837e-06,
      "loss": 0.17,
      "step": 49000
    },
    {
      "epoch": 12.52,
      "learning_rate": 7.907703293760601e-06,
      "loss": 0.1642,
      "step": 49500
    },
    {
      "epoch": 12.65,
      "learning_rate": 7.684548781576362e-06,
      "loss": 0.1713,
      "step": 50000
    },
    {
      "epoch": 12.77,
      "learning_rate": 7.461394269392127e-06,
      "loss": 0.1641,
      "step": 50500
    },
    {
      "epoch": 12.9,
      "learning_rate": 7.238239757207891e-06,
      "loss": 0.1659,
      "step": 51000
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.43983822042467136,
      "eval_loss": 0.11206455528736115,
      "eval_runtime": 18.1933,
      "eval_samples_per_second": 54.361,
      "eval_steps_per_second": 2.309,
      "step": 51402
    },
    {
      "epoch": 13.02,
      "learning_rate": 7.015085245023655e-06,
      "loss": 0.1672,
      "step": 51500
    },
    {
      "epoch": 13.15,
      "learning_rate": 6.791930732839418e-06,
      "loss": 0.164,
      "step": 52000
    },
    {
      "epoch": 13.28,
      "learning_rate": 6.568776220655182e-06,
      "loss": 0.1666,
      "step": 52500
    },
    {
      "epoch": 13.4,
      "learning_rate": 6.3456217084709455e-06,
      "loss": 0.1679,
      "step": 53000
    },
    {
      "epoch": 13.53,
      "learning_rate": 6.122467196286709e-06,
      "loss": 0.168,
      "step": 53500
    },
    {
      "epoch": 13.66,
      "learning_rate": 5.899312684102473e-06,
      "loss": 0.1662,
      "step": 54000
    },
    {
      "epoch": 13.78,
      "learning_rate": 5.676158171918236e-06,
      "loss": 0.1663,
      "step": 54500
    },
    {
      "epoch": 13.91,
      "learning_rate": 5.453003659734e-06,
      "loss": 0.1671,
      "step": 55000
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.43983822042467136,
      "eval_loss": 0.11148205399513245,
      "eval_runtime": 18.2266,
      "eval_samples_per_second": 54.261,
      "eval_steps_per_second": 2.304,
      "step": 55356
    },
    {
      "epoch": 14.04,
      "learning_rate": 5.229849147549764e-06,
      "loss": 0.1689,
      "step": 55500
    },
    {
      "epoch": 14.16,
      "learning_rate": 5.006694635365527e-06,
      "loss": 0.1677,
      "step": 56000
    },
    {
      "epoch": 14.29,
      "learning_rate": 4.783540123181291e-06,
      "loss": 0.1703,
      "step": 56500
    },
    {
      "epoch": 14.42,
      "learning_rate": 4.560385610997054e-06,
      "loss": 0.1659,
      "step": 57000
    },
    {
      "epoch": 14.54,
      "learning_rate": 4.3372310988128184e-06,
      "loss": 0.1673,
      "step": 57500
    },
    {
      "epoch": 14.67,
      "learning_rate": 4.114076586628581e-06,
      "loss": 0.1694,
      "step": 58000
    },
    {
      "epoch": 14.8,
      "learning_rate": 3.8909220744443456e-06,
      "loss": 0.1636,
      "step": 58500
    },
    {
      "epoch": 14.92,
      "learning_rate": 3.667767562260109e-06,
      "loss": 0.1662,
      "step": 59000
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.43983822042467136,
      "eval_loss": 0.11212924867868423,
      "eval_runtime": 17.8885,
      "eval_samples_per_second": 55.287,
      "eval_steps_per_second": 2.348,
      "step": 59310
    },
    {
      "epoch": 15.05,
      "learning_rate": 3.4446130500758727e-06,
      "loss": 0.168,
      "step": 59500
    },
    {
      "epoch": 15.17,
      "learning_rate": 3.2214585378916363e-06,
      "loss": 0.1684,
      "step": 60000
    },
    {
      "epoch": 15.3,
      "learning_rate": 2.9983040257074e-06,
      "loss": 0.1668,
      "step": 60500
    },
    {
      "epoch": 15.43,
      "learning_rate": 2.7751495135231634e-06,
      "loss": 0.1675,
      "step": 61000
    },
    {
      "epoch": 15.55,
      "learning_rate": 2.551995001338927e-06,
      "loss": 0.1639,
      "step": 61500
    },
    {
      "epoch": 15.68,
      "learning_rate": 2.3288404891546905e-06,
      "loss": 0.1645,
      "step": 62000
    },
    {
      "epoch": 15.81,
      "learning_rate": 2.105685976970454e-06,
      "loss": 0.1693,
      "step": 62500
    },
    {
      "epoch": 15.93,
      "learning_rate": 1.8825314647862179e-06,
      "loss": 0.1674,
      "step": 63000
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.43983822042467136,
      "eval_loss": 0.1121586263179779,
      "eval_runtime": 18.4647,
      "eval_samples_per_second": 53.562,
      "eval_steps_per_second": 2.275,
      "step": 63264
    },
    {
      "epoch": 16.06,
      "learning_rate": 1.6593769526019816e-06,
      "loss": 0.164,
      "step": 63500
    },
    {
      "epoch": 16.19,
      "learning_rate": 1.4362224404177452e-06,
      "loss": 0.1691,
      "step": 64000
    },
    {
      "epoch": 16.31,
      "learning_rate": 1.213067928233509e-06,
      "loss": 0.1674,
      "step": 64500
    },
    {
      "epoch": 16.44,
      "learning_rate": 9.899134160492725e-07,
      "loss": 0.1679,
      "step": 65000
    },
    {
      "epoch": 16.57,
      "learning_rate": 7.667589038650362e-07,
      "loss": 0.1698,
      "step": 65500
    },
    {
      "epoch": 16.69,
      "learning_rate": 5.436043916807998e-07,
      "loss": 0.168,
      "step": 66000
    }
  ],
  "logging_steps": 500,
  "max_steps": 67218,
  "num_train_epochs": 17,
  "save_steps": 1500,
  "total_flos": 9939187287216000.0,
  "trial_name": null,
  "trial_params": null
}