{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "global_step": 75115,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 1.9866870798109567e-05,
      "loss": 1.9359,
      "step": 500
    },
    {
      "epoch": 0.07,
      "learning_rate": 1.9733741596219132e-05,
      "loss": 1.7409,
      "step": 1000
    },
    {
      "epoch": 0.1,
      "learning_rate": 1.9600612394328697e-05,
      "loss": 1.6825,
      "step": 1500
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.9467483192438262e-05,
      "loss": 1.6341,
      "step": 2000
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.9334353990547827e-05,
      "loss": 1.5971,
      "step": 2500
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.9201224788657392e-05,
      "loss": 1.5629,
      "step": 3000
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.906809558676696e-05,
      "loss": 1.5576,
      "step": 3500
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.8934966384876526e-05,
      "loss": 1.5493,
      "step": 4000
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.8801837182986088e-05,
      "loss": 1.5461,
      "step": 4500
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.8668707981095656e-05,
      "loss": 1.5289,
      "step": 5000
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.853557877920522e-05,
      "loss": 1.4632,
      "step": 5500
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.8402449577314786e-05,
      "loss": 1.5016,
      "step": 6000
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.826932037542435e-05,
      "loss": 1.4624,
      "step": 6500
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.8136191173533916e-05,
      "loss": 1.4797,
      "step": 7000
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.800306197164348e-05,
      "loss": 1.52,
      "step": 7500
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.7869932769753047e-05,
      "loss": 1.4606,
      "step": 8000
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.773680356786261e-05,
      "loss": 1.471,
      "step": 8500
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.7603674365972177e-05,
      "loss": 1.4209,
      "step": 9000
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.7470545164081742e-05,
      "loss": 1.4318,
      "step": 9500
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.7337415962191307e-05,
      "loss": 1.4383,
      "step": 10000
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.7204286760300872e-05,
      "loss": 1.4414,
      "step": 10500
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.7071157558410437e-05,
      "loss": 1.4189,
      "step": 11000
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.6938028356520006e-05,
      "loss": 1.4017,
      "step": 11500
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.680489915462957e-05,
      "loss": 1.4162,
      "step": 12000
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.6671769952739136e-05,
      "loss": 1.4045,
      "step": 12500
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.65386407508487e-05,
      "loss": 1.3919,
      "step": 13000
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.6405511548958266e-05,
      "loss": 1.3825,
      "step": 13500
    },
    {
      "epoch": 0.93,
      "learning_rate": 1.627238234706783e-05,
      "loss": 1.3635,
      "step": 14000
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.6139253145177396e-05,
      "loss": 1.3731,
      "step": 14500
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.600612394328696e-05,
      "loss": 1.3736,
      "step": 15000
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.361890435218811,
      "eval_runtime": 194.0971,
      "eval_samples_per_second": 8.604,
      "eval_steps_per_second": 8.604,
      "step": 15023
    },
    {
      "epoch": 1.03,
      "learning_rate": 1.5872994741396526e-05,
      "loss": 1.1824,
      "step": 15500
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.573986553950609e-05,
      "loss": 1.1677,
      "step": 16000
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.5606736337615657e-05,
      "loss": 1.1428,
      "step": 16500
    },
    {
      "epoch": 1.13,
      "learning_rate": 1.547360713572522e-05,
      "loss": 1.1777,
      "step": 17000
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.5340477933834787e-05,
      "loss": 1.1826,
      "step": 17500
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.5207348731944354e-05,
      "loss": 1.1654,
      "step": 18000
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.507421953005392e-05,
      "loss": 1.1406,
      "step": 18500
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.4941090328163484e-05,
      "loss": 1.1789,
      "step": 19000
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.4807961126273049e-05,
      "loss": 1.2027,
      "step": 19500
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.4674831924382616e-05,
      "loss": 1.1533,
      "step": 20000
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.4541702722492179e-05,
      "loss": 1.1704,
      "step": 20500
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.4408573520601746e-05,
      "loss": 1.1783,
      "step": 21000
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.4275444318711311e-05,
      "loss": 1.1379,
      "step": 21500
    },
    {
      "epoch": 1.46,
      "learning_rate": 1.4142315116820874e-05,
      "loss": 1.1466,
      "step": 22000
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.4009185914930441e-05,
      "loss": 1.1507,
      "step": 22500
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.3876056713040006e-05,
      "loss": 1.1764,
      "step": 23000
    },
    {
      "epoch": 1.56,
      "learning_rate": 1.3742927511149571e-05,
      "loss": 1.1546,
      "step": 23500
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.3609798309259136e-05,
      "loss": 1.1689,
      "step": 24000
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.3476669107368703e-05,
      "loss": 1.1949,
      "step": 24500
    },
    {
      "epoch": 1.66,
      "learning_rate": 1.3343539905478268e-05,
      "loss": 1.1748,
      "step": 25000
    },
    {
      "epoch": 1.7,
      "learning_rate": 1.3210410703587833e-05,
      "loss": 1.1464,
      "step": 25500
    },
    {
      "epoch": 1.73,
      "learning_rate": 1.3077281501697398e-05,
      "loss": 1.1431,
      "step": 26000
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.2944152299806965e-05,
      "loss": 1.1687,
      "step": 26500
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.2811023097916529e-05,
      "loss": 1.1576,
      "step": 27000
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.2677893896026094e-05,
      "loss": 1.1535,
      "step": 27500
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.254476469413566e-05,
      "loss": 1.1313,
      "step": 28000
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.2411635492245224e-05,
      "loss": 1.161,
      "step": 28500
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.227850629035479e-05,
      "loss": 1.1313,
      "step": 29000
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.2145377088464356e-05,
      "loss": 1.1716,
      "step": 29500
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.201224788657392e-05,
      "loss": 1.147,
      "step": 30000
    },
    {
      "epoch": 2.0,
      "eval_loss": 1.2820119857788086,
      "eval_runtime": 194.1116,
      "eval_samples_per_second": 8.603,
      "eval_steps_per_second": 8.603,
      "step": 30046
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.1879118684683486e-05,
      "loss": 0.9294,
      "step": 30500
    },
    {
      "epoch": 2.06,
      "learning_rate": 1.1745989482793053e-05,
      "loss": 0.9088,
      "step": 31000
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.1612860280902616e-05,
      "loss": 0.9354,
      "step": 31500
    },
    {
      "epoch": 2.13,
      "learning_rate": 1.1479731079012181e-05,
      "loss": 0.922,
      "step": 32000
    },
    {
      "epoch": 2.16,
      "learning_rate": 1.1346601877121748e-05,
      "loss": 0.9055,
      "step": 32500
    },
    {
      "epoch": 2.2,
      "learning_rate": 1.1213472675231313e-05,
      "loss": 0.9151,
      "step": 33000
    },
    {
      "epoch": 2.23,
      "learning_rate": 1.1080343473340878e-05,
      "loss": 0.9178,
      "step": 33500
    },
    {
      "epoch": 2.26,
      "learning_rate": 1.0947214271450443e-05,
      "loss": 0.9004,
      "step": 34000
    },
    {
      "epoch": 2.3,
      "learning_rate": 1.081408506956001e-05,
      "loss": 0.9073,
      "step": 34500
    },
    {
      "epoch": 2.33,
      "learning_rate": 1.0680955867669574e-05,
      "loss": 0.9197,
      "step": 35000
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.0547826665779139e-05,
      "loss": 0.9201,
      "step": 35500
    },
    {
      "epoch": 2.4,
      "learning_rate": 1.0414697463888705e-05,
      "loss": 0.9365,
      "step": 36000
    },
    {
      "epoch": 2.43,
      "learning_rate": 1.0281568261998269e-05,
      "loss": 0.9298,
      "step": 36500
    },
    {
      "epoch": 2.46,
      "learning_rate": 1.0148439060107836e-05,
      "loss": 0.9154,
      "step": 37000
    },
    {
      "epoch": 2.5,
      "learning_rate": 1.00153098582174e-05,
      "loss": 0.9471,
      "step": 37500
    },
    {
      "epoch": 2.53,
      "learning_rate": 9.882180656326966e-06,
      "loss": 0.9379,
      "step": 38000
    },
    {
      "epoch": 2.56,
      "learning_rate": 9.749051454436531e-06,
      "loss": 0.9176,
      "step": 38500
    },
    {
      "epoch": 2.6,
      "learning_rate": 9.615922252546098e-06,
      "loss": 0.9211,
      "step": 39000
    },
    {
      "epoch": 2.63,
      "learning_rate": 9.482793050655663e-06,
      "loss": 0.9013,
      "step": 39500
    },
    {
      "epoch": 2.66,
      "learning_rate": 9.349663848765226e-06,
      "loss": 0.9283,
      "step": 40000
    },
    {
      "epoch": 2.7,
      "learning_rate": 9.216534646874793e-06,
      "loss": 0.943,
      "step": 40500
    },
    {
      "epoch": 2.73,
      "learning_rate": 9.083405444984358e-06,
      "loss": 0.9222,
      "step": 41000
    },
    {
      "epoch": 2.76,
      "learning_rate": 8.950276243093923e-06,
      "loss": 0.9055,
      "step": 41500
    },
    {
      "epoch": 2.8,
      "learning_rate": 8.817147041203488e-06,
      "loss": 0.9233,
      "step": 42000
    },
    {
      "epoch": 2.83,
      "learning_rate": 8.684017839313053e-06,
      "loss": 0.9208,
      "step": 42500
    },
    {
      "epoch": 2.86,
      "learning_rate": 8.55088863742262e-06,
      "loss": 0.9159,
      "step": 43000
    },
    {
      "epoch": 2.9,
      "learning_rate": 8.417759435532185e-06,
      "loss": 0.9155,
      "step": 43500
    },
    {
      "epoch": 2.93,
      "learning_rate": 8.284630233641749e-06,
      "loss": 0.9309,
      "step": 44000
    },
    {
      "epoch": 2.96,
      "learning_rate": 8.151501031751316e-06,
      "loss": 0.9202,
      "step": 44500
    },
    {
      "epoch": 3.0,
      "learning_rate": 8.01837182986088e-06,
      "loss": 0.9113,
      "step": 45000
    },
    {
      "epoch": 3.0,
      "eval_loss": 1.2868494987487793,
      "eval_runtime": 194.1014,
      "eval_samples_per_second": 8.604,
      "eval_steps_per_second": 8.604,
      "step": 45069
    },
    {
      "epoch": 3.03,
      "learning_rate": 7.885242627970446e-06,
      "loss": 0.7201,
      "step": 45500
    },
    {
      "epoch": 3.06,
      "learning_rate": 7.75211342608001e-06,
      "loss": 0.703,
      "step": 46000
    },
    {
      "epoch": 3.1,
      "learning_rate": 7.618984224189576e-06,
      "loss": 0.7127,
      "step": 46500
    },
    {
      "epoch": 3.13,
      "learning_rate": 7.485855022299142e-06,
      "loss": 0.7036,
      "step": 47000
    },
    {
      "epoch": 3.16,
      "learning_rate": 7.352725820408707e-06,
      "loss": 0.7042,
      "step": 47500
    },
    {
      "epoch": 3.2,
      "learning_rate": 7.219596618518273e-06,
      "loss": 0.7037,
      "step": 48000
    },
    {
      "epoch": 3.23,
      "learning_rate": 7.086467416627838e-06,
      "loss": 0.7129,
      "step": 48500
    },
    {
      "epoch": 3.26,
      "learning_rate": 6.953338214737403e-06,
      "loss": 0.6955,
      "step": 49000
    },
    {
      "epoch": 3.29,
      "learning_rate": 6.820209012846969e-06,
      "loss": 0.7161,
      "step": 49500
    },
    {
      "epoch": 3.33,
      "learning_rate": 6.687079810956534e-06,
      "loss": 0.7202,
      "step": 50000
    },
    {
      "epoch": 3.36,
      "learning_rate": 6.553950609066098e-06,
      "loss": 0.7123,
      "step": 50500
    },
    {
      "epoch": 3.39,
      "learning_rate": 6.420821407175664e-06,
      "loss": 0.7299,
      "step": 51000
    },
    {
      "epoch": 3.43,
      "learning_rate": 6.287692205285229e-06,
      "loss": 0.7074,
      "step": 51500
    },
    {
      "epoch": 3.46,
      "learning_rate": 6.154563003394795e-06,
      "loss": 0.7159,
      "step": 52000
    },
    {
      "epoch": 3.49,
      "learning_rate": 6.0214338015043604e-06,
      "loss": 0.7009,
      "step": 52500
    },
    {
      "epoch": 3.53,
      "learning_rate": 5.8883045996139255e-06,
      "loss": 0.7261,
      "step": 53000
    },
    {
      "epoch": 3.56,
      "learning_rate": 5.7551753977234915e-06,
      "loss": 0.7171,
      "step": 53500
    },
    {
      "epoch": 3.59,
      "learning_rate": 5.6220461958330566e-06,
      "loss": 0.6988,
      "step": 54000
    },
    {
      "epoch": 3.63,
      "learning_rate": 5.488916993942621e-06,
      "loss": 0.7362,
      "step": 54500
    },
    {
      "epoch": 3.66,
      "learning_rate": 5.355787792052188e-06,
      "loss": 0.699,
      "step": 55000
    },
    {
      "epoch": 3.69,
      "learning_rate": 5.222658590161752e-06,
      "loss": 0.7174,
      "step": 55500
    },
    {
      "epoch": 3.73,
      "learning_rate": 5.089529388271318e-06,
      "loss": 0.6933,
      "step": 56000
    },
    {
      "epoch": 3.76,
      "learning_rate": 4.956400186380883e-06,
      "loss": 0.6889,
      "step": 56500
    },
    {
      "epoch": 3.79,
      "learning_rate": 4.823270984490449e-06,
      "loss": 0.7,
      "step": 57000
    },
    {
      "epoch": 3.83,
      "learning_rate": 4.690141782600014e-06,
      "loss": 0.7188,
      "step": 57500
    },
    {
      "epoch": 3.86,
      "learning_rate": 4.557012580709579e-06,
      "loss": 0.7152,
      "step": 58000
    },
    {
      "epoch": 3.89,
      "learning_rate": 4.423883378819144e-06,
      "loss": 0.7102,
      "step": 58500
    },
    {
      "epoch": 3.93,
      "learning_rate": 4.29075417692871e-06,
      "loss": 0.7014,
      "step": 59000
    },
    {
      "epoch": 3.96,
      "learning_rate": 4.157624975038275e-06,
      "loss": 0.6923,
      "step": 59500
    },
    {
      "epoch": 3.99,
      "learning_rate": 4.02449577314784e-06,
      "loss": 0.7021,
      "step": 60000
    },
    {
      "epoch": 4.0,
      "eval_loss": 1.394597053527832,
      "eval_runtime": 194.1121,
      "eval_samples_per_second": 8.603,
      "eval_steps_per_second": 8.603,
      "step": 60092
    },
    {
      "epoch": 4.03,
      "learning_rate": 3.891366571257405e-06,
      "loss": 0.5703,
      "step": 60500
    },
    {
      "epoch": 4.06,
      "learning_rate": 3.758237369366971e-06,
      "loss": 0.5317,
      "step": 61000
    },
    {
      "epoch": 4.09,
      "learning_rate": 3.6251081674765364e-06,
      "loss": 0.5341,
      "step": 61500
    },
    {
      "epoch": 4.13,
      "learning_rate": 3.4919789655861015e-06,
      "loss": 0.5351,
      "step": 62000
    },
    {
      "epoch": 4.16,
      "learning_rate": 3.358849763695667e-06,
      "loss": 0.5277,
      "step": 62500
    },
    {
      "epoch": 4.19,
      "learning_rate": 3.225720561805232e-06,
      "loss": 0.5347,
      "step": 63000
    },
    {
      "epoch": 4.23,
      "learning_rate": 3.0925913599147976e-06,
      "loss": 0.5219,
      "step": 63500
    },
    {
      "epoch": 4.26,
      "learning_rate": 2.9594621580243627e-06,
      "loss": 0.5344,
      "step": 64000
    },
    {
      "epoch": 4.29,
      "learning_rate": 2.8263329561339283e-06,
      "loss": 0.5426,
      "step": 64500
    },
    {
      "epoch": 4.33,
      "learning_rate": 2.6932037542434938e-06,
      "loss": 0.5367,
      "step": 65000
    },
    {
      "epoch": 4.36,
      "learning_rate": 2.560074552353059e-06,
      "loss": 0.5447,
      "step": 65500
    },
    {
      "epoch": 4.39,
      "learning_rate": 2.4269453504626244e-06,
      "loss": 0.5361,
      "step": 66000
    },
    {
      "epoch": 4.43,
      "learning_rate": 2.2938161485721895e-06,
      "loss": 0.5324,
      "step": 66500
    },
    {
      "epoch": 4.46,
      "learning_rate": 2.160686946681755e-06,
      "loss": 0.5199,
      "step": 67000
    },
    {
      "epoch": 4.49,
      "learning_rate": 2.02755774479132e-06,
      "loss": 0.5278,
      "step": 67500
    },
    {
      "epoch": 4.53,
      "learning_rate": 1.8944285429008854e-06,
      "loss": 0.5472,
      "step": 68000
    },
    {
      "epoch": 4.56,
      "learning_rate": 1.761299341010451e-06,
      "loss": 0.5417,
      "step": 68500
    },
    {
      "epoch": 4.59,
      "learning_rate": 1.628170139120016e-06,
      "loss": 0.5268,
      "step": 69000
    },
    {
      "epoch": 4.63,
      "learning_rate": 1.4950409372295815e-06,
      "loss": 0.5361,
      "step": 69500
    },
    {
      "epoch": 4.66,
      "learning_rate": 1.3619117353391466e-06,
      "loss": 0.5345,
      "step": 70000
    },
    {
      "epoch": 4.69,
      "learning_rate": 1.228782533448712e-06,
      "loss": 0.5455,
      "step": 70500
    },
    {
      "epoch": 4.73,
      "learning_rate": 1.0956533315582773e-06,
      "loss": 0.5279,
      "step": 71000
    },
    {
      "epoch": 4.76,
      "learning_rate": 9.625241296678428e-07,
      "loss": 0.529,
      "step": 71500
    },
    {
      "epoch": 4.79,
      "learning_rate": 8.29394927777408e-07,
      "loss": 0.5297,
      "step": 72000
    },
    {
      "epoch": 4.83,
      "learning_rate": 6.962657258869733e-07,
      "loss": 0.5343,
      "step": 72500
    },
    {
      "epoch": 4.86,
      "learning_rate": 5.631365239965387e-07,
      "loss": 0.5295,
      "step": 73000
    },
    {
      "epoch": 4.89,
      "learning_rate": 4.30007322106104e-07,
      "loss": 0.5167,
      "step": 73500
    },
    {
      "epoch": 4.93,
      "learning_rate": 2.968781202156693e-07,
      "loss": 0.5465,
      "step": 74000
    },
    {
      "epoch": 4.96,
      "learning_rate": 1.6374891832523464e-07,
      "loss": 0.5374,
      "step": 74500
    },
    {
      "epoch": 4.99,
      "learning_rate": 3.061971643479997e-08,
      "loss": 0.5422,
      "step": 75000
    },
    {
      "epoch": 5.0,
      "eval_loss": 1.581581473350525,
      "eval_runtime": 194.1135,
      "eval_samples_per_second": 8.603,
      "eval_steps_per_second": 8.603,
      "step": 75115
    }
  ],
  "max_steps": 75115,
  "num_train_epochs": 5,
  "total_flos": 1.4915577452212224e+17,
  "trial_name": null,
  "trial_params": null
}