{
  "best_metric": 0.6414651274681091,
  "best_model_checkpoint": "output/checkpoint-170000",
  "epoch": 15.0,
  "global_step": 171660,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "learning_rate": 4.98e-06,
      "loss": 2.0401,
      "step": 500
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.980000000000001e-06,
      "loss": 1.397,
      "step": 1000
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.4979999999999999e-05,
      "loss": 1.302,
      "step": 1500
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.9980000000000002e-05,
      "loss": 1.242,
      "step": 2000
    },
    {
      "epoch": 0.22,
      "learning_rate": 2.4970000000000003e-05,
      "loss": 1.1985,
      "step": 2500
    },
    {
      "epoch": 0.22,
      "eval_loss": 1.0940483808517456,
      "eval_runtime": 3092.4183,
      "eval_samples_per_second": 24.931,
      "eval_steps_per_second": 1.558,
      "step": 2500
    },
    {
      "epoch": 0.26,
      "learning_rate": 2.9970000000000003e-05,
      "loss": 1.1712,
      "step": 3000
    },
    {
      "epoch": 0.31,
      "learning_rate": 3.4970000000000006e-05,
      "loss": 1.1419,
      "step": 3500
    },
    {
      "epoch": 0.35,
      "learning_rate": 3.9970000000000005e-05,
      "loss": 1.1259,
      "step": 4000
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.497e-05,
      "loss": 1.1088,
      "step": 4500
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.997e-05,
      "loss": 1.0937,
      "step": 5000
    },
    {
      "epoch": 0.44,
      "eval_loss": 1.0032908916473389,
      "eval_runtime": 3130.3678,
      "eval_samples_per_second": 24.628,
      "eval_steps_per_second": 1.539,
      "step": 5000
    },
    {
      "epoch": 0.48,
      "learning_rate": 5.497e-05,
      "loss": 1.0865,
      "step": 5500
    },
    {
      "epoch": 0.52,
      "learning_rate": 5.9970000000000004e-05,
      "loss": 1.0763,
      "step": 6000
    },
    {
      "epoch": 0.57,
      "learning_rate": 6.497000000000001e-05,
      "loss": 1.0666,
      "step": 6500
    },
    {
      "epoch": 0.61,
      "learning_rate": 6.997e-05,
      "loss": 1.0704,
      "step": 7000
    },
    {
      "epoch": 0.66,
      "learning_rate": 7.497000000000001e-05,
      "loss": 1.0675,
      "step": 7500
    },
    {
      "epoch": 0.66,
      "eval_loss": 0.9753177165985107,
      "eval_runtime": 3110.5948,
      "eval_samples_per_second": 24.785,
      "eval_steps_per_second": 1.549,
      "step": 7500
    },
    {
      "epoch": 0.7,
      "learning_rate": 7.997e-05,
      "loss": 1.0647,
      "step": 8000
    },
    {
      "epoch": 0.74,
      "learning_rate": 8.495e-05,
      "loss": 1.0493,
      "step": 8500
    },
    {
      "epoch": 0.79,
      "learning_rate": 8.995e-05,
      "loss": 1.0583,
      "step": 9000
    },
    {
      "epoch": 0.83,
      "learning_rate": 9.495e-05,
      "loss": 1.0582,
      "step": 9500
    },
    {
      "epoch": 0.87,
      "learning_rate": 9.995e-05,
      "loss": 1.0565,
      "step": 10000
    },
    {
      "epoch": 0.87,
      "eval_loss": 0.9801222085952759,
      "eval_runtime": 3136.3647,
      "eval_samples_per_second": 24.581,
      "eval_steps_per_second": 1.536,
      "step": 10000
    },
    {
      "epoch": 0.92,
      "learning_rate": 9.895171537484117e-05,
      "loss": 1.0507,
      "step": 10500
    },
    {
      "epoch": 0.96,
      "learning_rate": 9.789284201609487e-05,
      "loss": 1.0443,
      "step": 11000
    },
    {
      "epoch": 1.0,
      "learning_rate": 9.683396865734859e-05,
      "loss": 1.0318,
      "step": 11500
    },
    {
      "epoch": 1.05,
      "learning_rate": 9.577721304531978e-05,
      "loss": 1.0298,
      "step": 12000
    },
    {
      "epoch": 1.09,
      "learning_rate": 9.471833968657349e-05,
      "loss": 1.0244,
      "step": 12500
    },
    {
      "epoch": 1.09,
      "eval_loss": 0.9525517225265503,
      "eval_runtime": 3137.7232,
      "eval_samples_per_second": 24.571,
      "eval_steps_per_second": 1.536,
      "step": 12500
    },
    {
      "epoch": 1.14,
      "learning_rate": 9.365946632782719e-05,
      "loss": 1.024,
      "step": 13000
    },
    {
      "epoch": 1.18,
      "learning_rate": 9.26005929690809e-05,
      "loss": 1.0054,
      "step": 13500
    },
    {
      "epoch": 1.22,
      "learning_rate": 9.154383735705209e-05,
      "loss": 1.0105,
      "step": 14000
    },
    {
      "epoch": 1.27,
      "learning_rate": 9.04849639983058e-05,
      "loss": 1.0035,
      "step": 14500
    },
    {
      "epoch": 1.31,
      "learning_rate": 8.942609063955952e-05,
      "loss": 0.9943,
      "step": 15000
    },
    {
      "epoch": 1.31,
      "eval_loss": 0.9298429489135742,
      "eval_runtime": 3051.8254,
      "eval_samples_per_second": 25.262,
      "eval_steps_per_second": 1.579,
      "step": 15000
    },
    {
      "epoch": 1.35,
      "learning_rate": 8.836721728081322e-05,
      "loss": 0.9959,
      "step": 15500
    },
    {
      "epoch": 1.4,
      "learning_rate": 8.731046166878442e-05,
      "loss": 0.9874,
      "step": 16000
    },
    {
      "epoch": 1.44,
      "learning_rate": 8.625158831003812e-05,
      "loss": 0.9834,
      "step": 16500
    },
    {
      "epoch": 1.49,
      "learning_rate": 8.519483269800933e-05,
      "loss": 0.9764,
      "step": 17000
    },
    {
      "epoch": 1.53,
      "learning_rate": 8.413595933926302e-05,
      "loss": 0.9799,
      "step": 17500
    },
    {
      "epoch": 1.53,
      "eval_loss": 0.9034547805786133,
      "eval_runtime": 3114.4981,
      "eval_samples_per_second": 24.754,
      "eval_steps_per_second": 1.547,
      "step": 17500
    },
    {
      "epoch": 1.57,
      "learning_rate": 8.307708598051674e-05,
      "loss": 0.9642,
      "step": 18000
    },
    {
      "epoch": 1.62,
      "learning_rate": 8.201821262177044e-05,
      "loss": 0.9671,
      "step": 18500
    },
    {
      "epoch": 1.66,
      "learning_rate": 8.095933926302415e-05,
      "loss": 0.9637,
      "step": 19000
    },
    {
      "epoch": 1.7,
      "learning_rate": 7.990046590427785e-05,
      "loss": 0.9545,
      "step": 19500
    },
    {
      "epoch": 1.75,
      "learning_rate": 7.884159254553156e-05,
      "loss": 0.95,
      "step": 20000
    },
    {
      "epoch": 1.75,
      "eval_loss": 0.8834716081619263,
      "eval_runtime": 3125.2028,
      "eval_samples_per_second": 24.669,
      "eval_steps_per_second": 1.542,
      "step": 20000
    },
    {
      "epoch": 1.79,
      "learning_rate": 7.778483693350275e-05,
      "loss": 0.9556,
      "step": 20500
    },
    {
      "epoch": 1.84,
      "learning_rate": 7.672596357475647e-05,
      "loss": 0.9519,
      "step": 21000
    },
    {
      "epoch": 1.88,
      "learning_rate": 7.566709021601018e-05,
      "loss": 0.9448,
      "step": 21500
    },
    {
      "epoch": 1.92,
      "learning_rate": 7.460821685726388e-05,
      "loss": 0.9382,
      "step": 22000
    },
    {
      "epoch": 1.97,
      "learning_rate": 7.354934349851758e-05,
      "loss": 0.933,
      "step": 22500
    },
    {
      "epoch": 1.97,
      "eval_loss": 0.8636178970336914,
      "eval_runtime": 3087.2735,
      "eval_samples_per_second": 24.972,
      "eval_steps_per_second": 1.561,
      "step": 22500
    },
    {
      "epoch": 2.01,
      "learning_rate": 7.249258788648878e-05,
      "loss": 0.93,
      "step": 23000
    },
    {
      "epoch": 2.05,
      "learning_rate": 7.14337145277425e-05,
      "loss": 0.9193,
      "step": 23500
    },
    {
      "epoch": 2.1,
      "learning_rate": 7.037484116899618e-05,
      "loss": 0.9141,
      "step": 24000
    },
    {
      "epoch": 2.14,
      "learning_rate": 6.931596781024989e-05,
      "loss": 0.9092,
      "step": 24500
    },
    {
      "epoch": 2.18,
      "learning_rate": 6.82570944515036e-05,
      "loss": 0.9079,
      "step": 25000
    },
    {
      "epoch": 2.18,
      "eval_loss": 0.8507079482078552,
      "eval_runtime": 3106.3211,
      "eval_samples_per_second": 24.819,
      "eval_steps_per_second": 1.551,
      "step": 25000
    },
    {
      "epoch": 2.23,
      "learning_rate": 6.72003388394748e-05,
      "loss": 0.9041,
      "step": 25500
    },
    {
      "epoch": 2.27,
      "learning_rate": 6.614146548072851e-05,
      "loss": 0.9004,
      "step": 26000
    },
    {
      "epoch": 2.32,
      "learning_rate": 6.508259212198221e-05,
      "loss": 0.8995,
      "step": 26500
    },
    {
      "epoch": 2.36,
      "learning_rate": 6.402371876323592e-05,
      "loss": 0.8922,
      "step": 27000
    },
    {
      "epoch": 2.4,
      "learning_rate": 6.296484540448962e-05,
      "loss": 0.8938,
      "step": 27500
    },
    {
      "epoch": 2.4,
      "eval_loss": 0.8396568894386292,
      "eval_runtime": 3033.9082,
      "eval_samples_per_second": 25.411,
      "eval_steps_per_second": 1.588,
      "step": 27500
    },
    {
      "epoch": 2.45,
      "learning_rate": 6.190597204574333e-05,
      "loss": 0.8897,
      "step": 28000
    },
    {
      "epoch": 2.49,
      "learning_rate": 6.084709868699704e-05,
      "loss": 0.8855,
      "step": 28500
    },
    {
      "epoch": 2.53,
      "learning_rate": 5.979034307496824e-05,
      "loss": 0.8795,
      "step": 29000
    },
    {
      "epoch": 2.58,
      "learning_rate": 5.873146971622194e-05,
      "loss": 0.8841,
      "step": 29500
    },
    {
      "epoch": 2.62,
      "learning_rate": 5.767259635747565e-05,
      "loss": 0.8781,
      "step": 30000
    },
    {
      "epoch": 2.62,
      "eval_loss": 0.8194745182991028,
      "eval_runtime": 3075.2096,
      "eval_samples_per_second": 25.07,
      "eval_steps_per_second": 1.567,
      "step": 30000
    },
    {
      "epoch": 2.67,
      "learning_rate": 5.6613722998729355e-05,
      "loss": 0.8776,
      "step": 30500
    },
    {
      "epoch": 2.71,
      "learning_rate": 5.555484963998306e-05,
      "loss": 0.8777,
      "step": 31000
    },
    {
      "epoch": 2.75,
      "learning_rate": 5.4495976281236774e-05,
      "loss": 0.865,
      "step": 31500
    },
    {
      "epoch": 2.8,
      "learning_rate": 5.3437102922490467e-05,
      "loss": 0.8752,
      "step": 32000
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.237822956374417e-05,
      "loss": 0.8647,
      "step": 32500
    },
    {
      "epoch": 2.84,
      "eval_loss": 0.8087666630744934,
      "eval_runtime": 3169.435,
      "eval_samples_per_second": 24.325,
      "eval_steps_per_second": 1.52,
      "step": 32500
    },
    {
      "epoch": 2.88,
      "learning_rate": 5.1321473951715384e-05,
      "loss": 0.8599,
      "step": 33000
    },
    {
      "epoch": 2.93,
      "learning_rate": 5.0262600592969076e-05,
      "loss": 0.8568,
      "step": 33500
    },
    {
      "epoch": 2.97,
      "learning_rate": 4.920372723422279e-05,
      "loss": 0.8528,
      "step": 34000
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.8144853875476495e-05,
      "loss": 0.8463,
      "step": 34500
    },
    {
      "epoch": 3.06,
      "learning_rate": 4.70859805167302e-05,
      "loss": 0.8422,
      "step": 35000
    },
    {
      "epoch": 3.06,
      "eval_loss": 0.7954000234603882,
      "eval_runtime": 3127.0452,
      "eval_samples_per_second": 24.655,
      "eval_steps_per_second": 1.541,
      "step": 35000
    },
    {
      "epoch": 3.1,
      "learning_rate": 4.602710715798391e-05,
      "loss": 0.8373,
      "step": 35500
    },
    {
      "epoch": 3.15,
      "learning_rate": 4.4968233799237613e-05,
      "loss": 0.8268,
      "step": 36000
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.390936044049132e-05,
      "loss": 0.8426,
      "step": 36500
    },
    {
      "epoch": 3.23,
      "learning_rate": 4.2850487081745026e-05,
      "loss": 0.837,
      "step": 37000
    },
    {
      "epoch": 3.28,
      "learning_rate": 4.179373146971622e-05,
      "loss": 0.831,
      "step": 37500
    },
    {
      "epoch": 3.28,
      "eval_loss": 0.7871229648590088,
      "eval_runtime": 3095.9065,
      "eval_samples_per_second": 24.903,
      "eval_steps_per_second": 1.557,
      "step": 37500
    },
    {
      "epoch": 3.32,
      "learning_rate": 4.073485811096993e-05,
      "loss": 0.8343,
      "step": 38000
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.9675984752223635e-05,
      "loss": 0.825,
      "step": 38500
    },
    {
      "epoch": 3.41,
      "learning_rate": 3.861711139347734e-05,
      "loss": 0.8195,
      "step": 39000
    },
    {
      "epoch": 3.45,
      "learning_rate": 3.755823803473105e-05,
      "loss": 0.8219,
      "step": 39500
    },
    {
      "epoch": 3.5,
      "learning_rate": 3.650148242270225e-05,
      "loss": 0.8173,
      "step": 40000
    },
    {
      "epoch": 3.5,
      "eval_loss": 0.7720773220062256,
      "eval_runtime": 3173.0037,
      "eval_samples_per_second": 24.297,
      "eval_steps_per_second": 1.519,
      "step": 40000
    },
    {
      "epoch": 3.54,
      "learning_rate": 3.544260906395595e-05,
      "loss": 0.8148,
      "step": 40500
    },
    {
      "epoch": 3.58,
      "learning_rate": 3.4385853451927155e-05,
      "loss": 0.8102,
      "step": 41000
    },
    {
      "epoch": 3.63,
      "learning_rate": 3.3329097839898346e-05,
      "loss": 0.8103,
      "step": 41500
    },
    {
      "epoch": 3.67,
      "learning_rate": 3.227022448115206e-05,
      "loss": 0.8116,
      "step": 42000
    },
    {
      "epoch": 3.71,
      "learning_rate": 3.121346886912325e-05,
      "loss": 0.8072,
      "step": 42500
    },
    {
      "epoch": 3.71,
      "eval_loss": 0.7610893845558167,
      "eval_runtime": 3130.5399,
      "eval_samples_per_second": 24.627,
      "eval_steps_per_second": 1.539,
      "step": 42500
    },
    {
      "epoch": 3.76,
      "learning_rate": 3.015459551037696e-05,
      "loss": 0.802,
      "step": 43000
    },
    {
      "epoch": 3.8,
      "learning_rate": 2.9095722151630668e-05,
      "loss": 0.7994,
      "step": 43500
    },
    {
      "epoch": 3.84,
      "learning_rate": 2.8036848792884374e-05,
      "loss": 0.8029,
      "step": 44000
    },
    {
      "epoch": 3.89,
      "learning_rate": 2.697797543413808e-05,
      "loss": 0.7975,
      "step": 44500
    },
    {
      "epoch": 3.93,
      "learning_rate": 2.5919102075391783e-05,
      "loss": 0.8011,
      "step": 45000
    },
    {
      "epoch": 3.93,
      "eval_loss": 0.7531821131706238,
      "eval_runtime": 3062.0684,
      "eval_samples_per_second": 25.178,
      "eval_steps_per_second": 1.574,
      "step": 45000
    },
    {
      "epoch": 3.98,
      "learning_rate": 2.486022871664549e-05,
      "loss": 0.7935,
      "step": 45500
    },
    {
      "epoch": 4.02,
      "learning_rate": 2.3801355357899195e-05,
      "loss": 0.793,
      "step": 46000
    },
    {
      "epoch": 4.06,
      "learning_rate": 2.2742481999152905e-05,
      "loss": 0.788,
      "step": 46500
    },
    {
      "epoch": 4.11,
      "learning_rate": 2.1683608640406607e-05,
      "loss": 0.7853,
      "step": 47000
    },
    {
      "epoch": 4.15,
      "learning_rate": 2.0624735281660313e-05,
      "loss": 0.7828,
      "step": 47500
    },
    {
      "epoch": 4.15,
      "eval_loss": 0.7431035041809082,
      "eval_runtime": 3053.0232,
      "eval_samples_per_second": 25.252,
      "eval_steps_per_second": 1.578,
      "step": 47500
    },
    {
      "epoch": 4.19,
      "learning_rate": 1.956586192291402e-05,
      "loss": 0.7715,
      "step": 48000
    },
    {
      "epoch": 4.24,
      "learning_rate": 1.8506988564167726e-05,
      "loss": 0.7752,
      "step": 48500
    },
    {
      "epoch": 4.28,
      "learning_rate": 1.7450232952138926e-05,
      "loss": 0.7714,
      "step": 49000
    },
    {
      "epoch": 4.33,
      "learning_rate": 1.639135959339263e-05,
      "loss": 0.7736,
      "step": 49500
    },
    {
      "epoch": 4.37,
      "learning_rate": 1.533248623464634e-05,
      "loss": 0.7691,
      "step": 50000
    },
    {
      "epoch": 4.37,
      "eval_loss": 0.7367000579833984,
      "eval_runtime": 3035.0963,
      "eval_samples_per_second": 25.402,
      "eval_steps_per_second": 1.588,
      "step": 50000
    },
    {
      "epoch": 4.41,
      "learning_rate": 1.4273612875900045e-05,
      "loss": 0.7685,
      "step": 50500
    },
    {
      "epoch": 4.46,
      "learning_rate": 1.3214739517153749e-05,
      "loss": 0.7639,
      "step": 51000
    },
    {
      "epoch": 4.5,
      "learning_rate": 1.2157983905124948e-05,
      "loss": 0.7654,
      "step": 51500
    },
    {
      "epoch": 4.54,
      "learning_rate": 1.1099110546378654e-05,
      "loss": 0.7677,
      "step": 52000
    },
    {
      "epoch": 4.59,
      "learning_rate": 1.0040237187632359e-05,
      "loss": 0.7659,
      "step": 52500
    },
    {
      "epoch": 4.59,
      "eval_loss": 0.7291901111602783,
      "eval_runtime": 3110.3355,
      "eval_samples_per_second": 24.787,
      "eval_steps_per_second": 1.549,
      "step": 52500
    },
    {
      "epoch": 4.63,
      "learning_rate": 8.981363828886067e-06,
      "loss": 0.763,
      "step": 53000
    },
    {
      "epoch": 4.67,
      "learning_rate": 7.922490470139773e-06,
      "loss": 0.759,
      "step": 53500
    },
    {
      "epoch": 4.72,
      "learning_rate": 6.863617111393478e-06,
      "loss": 0.7559,
      "step": 54000
    },
    {
      "epoch": 4.76,
      "learning_rate": 5.804743752647183e-06,
      "loss": 0.7547,
      "step": 54500
    },
    {
      "epoch": 4.81,
      "learning_rate": 4.74587039390089e-06,
      "loss": 0.7606,
      "step": 55000
    },
    {
      "epoch": 4.81,
      "eval_loss": 0.7244983911514282,
      "eval_runtime": 3076.3307,
      "eval_samples_per_second": 25.061,
      "eval_steps_per_second": 1.566,
      "step": 55000
    },
    {
      "epoch": 4.85,
      "learning_rate": 5.6455381080045965e-05,
      "loss": 0.7842,
      "step": 55500
    },
    {
      "epoch": 4.89,
      "learning_rate": 5.597663730371505e-05,
      "loss": 0.7983,
      "step": 56000
    },
    {
      "epoch": 4.94,
      "learning_rate": 5.549789352738415e-05,
      "loss": 0.8032,
      "step": 56500
    },
    {
      "epoch": 4.98,
      "learning_rate": 5.5019149751053236e-05,
      "loss": 0.8037,
      "step": 57000
    },
    {
      "epoch": 5.02,
      "learning_rate": 5.454040597472233e-05,
      "loss": 0.8082,
      "step": 57500
    },
    {
      "epoch": 5.02,
      "eval_loss": 0.7695716023445129,
      "eval_runtime": 3095.7063,
      "eval_samples_per_second": 24.904,
      "eval_steps_per_second": 1.557,
      "step": 57500
    },
    {
      "epoch": 5.07,
      "learning_rate": 5.406166219839143e-05,
      "loss": 0.8102,
      "step": 58000
    },
    {
      "epoch": 5.11,
      "learning_rate": 5.358291842206051e-05,
      "loss": 0.8067,
      "step": 58500
    },
    {
      "epoch": 5.16,
      "learning_rate": 5.310417464572961e-05,
      "loss": 0.8103,
      "step": 59000
    },
    {
      "epoch": 5.2,
      "learning_rate": 5.2627345844504025e-05,
      "loss": 0.8088,
      "step": 59500
    },
    {
      "epoch": 5.24,
      "learning_rate": 5.214955955572578e-05,
      "loss": 0.8114,
      "step": 60000
    },
    {
      "epoch": 5.24,
      "eval_loss": 0.7694710493087769,
      "eval_runtime": 3084.9083,
      "eval_samples_per_second": 24.991,
      "eval_steps_per_second": 1.562,
      "step": 60000
    },
    {
      "epoch": 5.29,
      "learning_rate": 5.167081577939487e-05,
      "loss": 0.8073,
      "step": 60500
    },
    {
      "epoch": 5.33,
      "learning_rate": 5.119207200306396e-05,
      "loss": 0.8019,
      "step": 61000
    },
    {
      "epoch": 5.37,
      "learning_rate": 5.071332822673306e-05,
      "loss": 0.7971,
      "step": 61500
    },
    {
      "epoch": 5.42,
      "learning_rate": 5.0234584450402144e-05,
      "loss": 0.8024,
      "step": 62000
    },
    {
      "epoch": 5.46,
      "learning_rate": 4.97567981616239e-05,
      "loss": 0.8022,
      "step": 62500
    },
    {
      "epoch": 5.46,
      "eval_loss": 0.7612630128860474,
      "eval_runtime": 3135.2609,
      "eval_samples_per_second": 24.59,
      "eval_steps_per_second": 1.537,
      "step": 62500
    },
    {
      "epoch": 5.51,
      "learning_rate": 4.927805438529299e-05,
      "loss": 0.8071,
      "step": 63000
    },
    {
      "epoch": 5.55,
      "learning_rate": 4.8799310608962085e-05,
      "loss": 0.793,
      "step": 63500
    },
    {
      "epoch": 5.59,
      "learning_rate": 4.832152432018384e-05,
      "loss": 0.798,
      "step": 64000
    },
    {
      "epoch": 5.64,
      "learning_rate": 4.7842780543852934e-05,
      "loss": 0.7976,
      "step": 64500
    },
    {
      "epoch": 5.68,
      "learning_rate": 4.7364036767522026e-05,
      "loss": 0.7986,
      "step": 65000
    },
    {
      "epoch": 5.68,
      "eval_loss": 0.7558260560035706,
      "eval_runtime": 3154.5823,
      "eval_samples_per_second": 24.439,
      "eval_steps_per_second": 1.528,
      "step": 65000
    },
    {
      "epoch": 5.72,
      "learning_rate": 4.688529299119111e-05,
      "loss": 0.7918,
      "step": 65500
    },
    {
      "epoch": 5.77,
      "learning_rate": 4.6406549214860205e-05,
      "loss": 0.7941,
      "step": 66000
    },
    {
      "epoch": 5.81,
      "learning_rate": 4.5927805438529304e-05,
      "loss": 0.7976,
      "step": 66500
    },
    {
      "epoch": 5.85,
      "learning_rate": 4.5449061662198396e-05,
      "loss": 0.7953,
      "step": 67000
    },
    {
      "epoch": 5.9,
      "learning_rate": 4.497031788586749e-05,
      "loss": 0.8018,
      "step": 67500
    },
    {
      "epoch": 5.9,
      "eval_loss": 0.7478091716766357,
      "eval_runtime": 3109.9687,
      "eval_samples_per_second": 24.79,
      "eval_steps_per_second": 1.55,
      "step": 67500
    },
    {
      "epoch": 5.94,
      "learning_rate": 4.4491574109536574e-05,
      "loss": 0.7939,
      "step": 68000
    },
    {
      "epoch": 5.99,
      "learning_rate": 4.401283033320567e-05,
      "loss": 0.7844,
      "step": 68500
    },
    {
      "epoch": 6.03,
      "learning_rate": 4.3534086556874766e-05,
      "loss": 0.7853,
      "step": 69000
    },
    {
      "epoch": 6.07,
      "learning_rate": 4.305534278054386e-05,
      "loss": 0.7776,
      "step": 69500
    },
    {
      "epoch": 6.12,
      "learning_rate": 4.2576599004212944e-05,
      "loss": 0.782,
      "step": 70000
    },
    {
      "epoch": 6.12,
      "eval_loss": 0.7434529662132263,
      "eval_runtime": 3111.4708,
      "eval_samples_per_second": 24.778,
      "eval_steps_per_second": 1.549,
      "step": 70000
    },
    {
      "epoch": 6.16,
      "learning_rate": 4.209785522788204e-05,
      "loss": 0.7812,
      "step": 70500
    },
    {
      "epoch": 6.2,
      "learning_rate": 4.161911145155113e-05,
      "loss": 0.7816,
      "step": 71000
    },
    {
      "epoch": 6.25,
      "learning_rate": 4.1141325162772885e-05,
      "loss": 0.7745,
      "step": 71500
    },
    {
      "epoch": 6.29,
      "learning_rate": 4.066258138644198e-05,
      "loss": 0.7776,
      "step": 72000
    },
    {
      "epoch": 6.34,
      "learning_rate": 4.018383761011107e-05,
      "loss": 0.7743,
      "step": 72500
    },
    {
      "epoch": 6.34,
      "eval_loss": 0.7367435097694397,
      "eval_runtime": 3109.6155,
      "eval_samples_per_second": 24.793,
      "eval_steps_per_second": 1.55,
      "step": 72500
    },
    {
      "epoch": 6.38,
      "learning_rate": 3.970509383378016e-05,
      "loss": 0.7762,
      "step": 73000
    },
    {
      "epoch": 6.42,
      "learning_rate": 3.9226350057449255e-05,
      "loss": 0.7728,
      "step": 73500
    },
    {
      "epoch": 6.47,
      "learning_rate": 3.874760628111835e-05,
      "loss": 0.7743,
      "step": 74000
    },
    {
      "epoch": 6.51,
      "learning_rate": 3.826886250478744e-05,
      "loss": 0.7715,
      "step": 74500
    },
    {
      "epoch": 6.55,
      "learning_rate": 3.779011872845653e-05,
      "loss": 0.774,
      "step": 75000
    },
    {
      "epoch": 6.55,
      "eval_loss": 0.7312998175621033,
      "eval_runtime": 3143.7734,
      "eval_samples_per_second": 24.523,
      "eval_steps_per_second": 1.533,
      "step": 75000
    },
    {
      "epoch": 6.6,
      "learning_rate": 3.7311374952125625e-05,
      "loss": 0.7656,
      "step": 75500
    },
    {
      "epoch": 6.64,
      "learning_rate": 3.683263117579472e-05,
      "loss": 0.7653,
      "step": 76000
    },
    {
      "epoch": 6.68,
      "learning_rate": 3.635388739946381e-05,
      "loss": 0.7619,
      "step": 76500
    },
    {
      "epoch": 6.73,
      "learning_rate": 3.587610111068556e-05,
      "loss": 0.765,
      "step": 77000
    },
    {
      "epoch": 6.77,
      "learning_rate": 3.539735733435465e-05,
      "loss": 0.7692,
      "step": 77500
    },
    {
      "epoch": 6.77,
      "eval_loss": 0.7270153164863586,
      "eval_runtime": 3083.4104,
      "eval_samples_per_second": 25.003,
      "eval_steps_per_second": 1.563,
      "step": 77500
    },
    {
      "epoch": 6.82,
      "learning_rate": 3.491957104557641e-05,
      "loss": 0.7676,
      "step": 78000
    },
    {
      "epoch": 6.86,
      "learning_rate": 3.44408272692455e-05,
      "loss": 0.7617,
      "step": 78500
    },
    {
      "epoch": 6.9,
      "learning_rate": 3.396208349291459e-05,
      "loss": 0.765,
      "step": 79000
    },
    {
      "epoch": 6.95,
      "learning_rate": 3.3483339716583685e-05,
      "loss": 0.7609,
      "step": 79500
    },
    {
      "epoch": 6.99,
      "learning_rate": 3.300459594025278e-05,
      "loss": 0.7604,
      "step": 80000
    },
    {
      "epoch": 6.99,
      "eval_loss": 0.7200314998626709,
      "eval_runtime": 3077.861,
      "eval_samples_per_second": 25.049,
      "eval_steps_per_second": 1.566,
      "step": 80000
    },
    {
      "epoch": 7.03,
      "learning_rate": 3.252585216392187e-05,
      "loss": 0.7513,
      "step": 80500
    },
    {
      "epoch": 7.08,
      "learning_rate": 3.204710838759096e-05,
      "loss": 0.7549,
      "step": 81000
    },
    {
      "epoch": 7.12,
      "learning_rate": 3.156932209881271e-05,
      "loss": 0.7485,
      "step": 81500
    },
    {
      "epoch": 7.17,
      "learning_rate": 3.109057832248181e-05,
      "loss": 0.7486,
      "step": 82000
    },
    {
      "epoch": 7.21,
      "learning_rate": 3.0611834546150903e-05,
      "loss": 0.7468,
      "step": 82500
    },
    {
      "epoch": 7.21,
      "eval_loss": 0.7163689136505127,
      "eval_runtime": 3165.721,
      "eval_samples_per_second": 24.353,
      "eval_steps_per_second": 1.522,
      "step": 82500
    },
    {
      "epoch": 7.25,
      "learning_rate": 3.0133090769819993e-05,
      "loss": 0.7471,
      "step": 83000
    },
    {
      "epoch": 7.3,
      "learning_rate": 2.9654346993489085e-05,
      "loss": 0.7473,
      "step": 83500
    },
    {
      "epoch": 7.34,
      "learning_rate": 2.9175603217158177e-05,
      "loss": 0.747,
      "step": 84000
    },
    {
      "epoch": 7.38,
      "learning_rate": 2.8696859440827273e-05,
      "loss": 0.743,
      "step": 84500
    },
    {
      "epoch": 7.43,
      "learning_rate": 2.8219073152049026e-05,
      "loss": 0.7486,
      "step": 85000
    },
    {
      "epoch": 7.43,
      "eval_loss": 0.7117038369178772,
      "eval_runtime": 3102.6798,
      "eval_samples_per_second": 24.848,
      "eval_steps_per_second": 1.553,
      "step": 85000
    },
    {
      "epoch": 7.47,
      "learning_rate": 2.7740329375718115e-05,
      "loss": 0.7411,
      "step": 85500
    },
    {
      "epoch": 7.51,
      "learning_rate": 2.7261585599387208e-05,
      "loss": 0.7418,
      "step": 86000
    },
    {
      "epoch": 7.56,
      "learning_rate": 2.6782841823056303e-05,
      "loss": 0.7443,
      "step": 86500
    },
    {
      "epoch": 7.6,
      "learning_rate": 2.6305055534278056e-05,
      "loss": 0.7403,
      "step": 87000
    },
    {
      "epoch": 7.65,
      "learning_rate": 2.582631175794715e-05,
      "loss": 0.7399,
      "step": 87500
    },
    {
      "epoch": 7.65,
      "eval_loss": 0.7042549252510071,
      "eval_runtime": 3110.6014,
      "eval_samples_per_second": 24.785,
      "eval_steps_per_second": 1.549,
      "step": 87500
    },
    {
      "epoch": 7.69,
      "learning_rate": 2.5347567981616238e-05,
      "loss": 0.7393,
      "step": 88000
    },
    {
      "epoch": 7.73,
      "learning_rate": 2.4868824205285333e-05,
      "loss": 0.7404,
      "step": 88500
    },
    {
      "epoch": 7.78,
      "learning_rate": 2.4390080428954426e-05,
      "loss": 0.7356,
      "step": 89000
    },
    {
      "epoch": 7.82,
      "learning_rate": 2.391133665262352e-05,
      "loss": 0.7395,
      "step": 89500
    },
    {
      "epoch": 7.86,
      "learning_rate": 2.3432592876292607e-05,
      "loss": 0.7306,
      "step": 90000
    },
    {
      "epoch": 7.86,
      "eval_loss": 0.6955912709236145,
      "eval_runtime": 3139.2914,
      "eval_samples_per_second": 24.558,
      "eval_steps_per_second": 1.535,
      "step": 90000
    },
    {
      "epoch": 7.91,
      "learning_rate": 2.2953849099961703e-05,
      "loss": 0.735,
      "step": 90500
    },
    {
      "epoch": 7.95,
      "learning_rate": 2.2475105323630792e-05,
      "loss": 0.7281,
      "step": 91000
    },
    {
      "epoch": 8.0,
      "learning_rate": 2.1996361547299888e-05,
      "loss": 0.732,
      "step": 91500
    },
    {
      "epoch": 8.04,
      "learning_rate": 2.1517617770968977e-05,
      "loss": 0.7219,
      "step": 92000
    },
    {
      "epoch": 8.08,
      "learning_rate": 2.1039831482190733e-05,
      "loss": 0.7243,
      "step": 92500
    },
    {
      "epoch": 8.08,
      "eval_loss": 0.695923924446106,
      "eval_runtime": 3092.5373,
      "eval_samples_per_second": 24.93,
      "eval_steps_per_second": 1.558,
      "step": 92500
    },
    {
      "epoch": 8.13,
      "learning_rate": 2.0561087705859826e-05,
      "loss": 0.7243,
      "step": 93000
    },
    {
      "epoch": 8.17,
      "learning_rate": 2.0082343929528918e-05,
      "loss": 0.7258,
      "step": 93500
    },
    {
      "epoch": 8.21,
      "learning_rate": 1.960360015319801e-05,
      "loss": 0.7211,
      "step": 94000
    },
    {
      "epoch": 8.26,
      "learning_rate": 1.91248563768671e-05,
      "loss": 0.7214,
      "step": 94500
    },
    {
      "epoch": 8.3,
      "learning_rate": 1.8646112600536196e-05,
      "loss": 0.7132,
      "step": 95000
    },
    {
      "epoch": 8.3,
      "eval_loss": 0.6916212439537048,
      "eval_runtime": 3163.2312,
      "eval_samples_per_second": 24.373,
      "eval_steps_per_second": 1.523,
      "step": 95000
    },
    {
      "epoch": 8.34,
      "learning_rate": 1.8167368824205285e-05,
      "loss": 0.7188,
      "step": 95500
    },
    {
      "epoch": 8.39,
      "learning_rate": 1.768862504787438e-05,
      "loss": 0.717,
      "step": 96000
    },
    {
      "epoch": 8.43,
      "learning_rate": 1.7210838759096133e-05,
      "loss": 0.7189,
      "step": 96500
    },
    {
      "epoch": 8.48,
      "learning_rate": 1.6732094982765226e-05,
      "loss": 0.7206,
      "step": 97000
    },
    {
      "epoch": 8.52,
      "learning_rate": 1.625430869398698e-05,
      "loss": 0.71,
      "step": 97500
    },
    {
      "epoch": 8.52,
      "eval_loss": 0.6853311061859131,
      "eval_runtime": 3146.2403,
      "eval_samples_per_second": 24.504,
      "eval_steps_per_second": 1.532,
      "step": 97500
    },
    {
      "epoch": 8.56,
      "learning_rate": 1.577556491765607e-05,
      "loss": 0.7148,
      "step": 98000
    },
    {
      "epoch": 8.61,
      "learning_rate": 1.5296821141325163e-05,
      "loss": 0.7169,
      "step": 98500
    },
    {
      "epoch": 8.65,
      "learning_rate": 1.4818077364994256e-05,
      "loss": 0.7155,
      "step": 99000
    },
    {
      "epoch": 8.69,
      "learning_rate": 1.4339333588663348e-05,
      "loss": 0.713,
      "step": 99500
    },
    {
      "epoch": 8.74,
      "learning_rate": 1.3860589812332439e-05,
      "loss": 0.7128,
      "step": 100000
    },
    {
      "epoch": 8.74,
      "eval_loss": 0.6855071783065796,
      "eval_runtime": 3110.846,
      "eval_samples_per_second": 24.783,
      "eval_steps_per_second": 1.549,
      "step": 100000
    },
    {
      "epoch": 8.78,
      "learning_rate": 1.3381846036001533e-05,
      "loss": 0.7107,
      "step": 100500
    },
    {
      "epoch": 8.83,
      "learning_rate": 1.2903102259670624e-05,
      "loss": 0.7142,
      "step": 101000
    },
    {
      "epoch": 8.87,
      "learning_rate": 1.2424358483339716e-05,
      "loss": 0.7035,
      "step": 101500
    },
    {
      "epoch": 8.91,
      "learning_rate": 1.1945614707008809e-05,
      "loss": 0.7061,
      "step": 102000
    },
    {
      "epoch": 8.96,
      "learning_rate": 1.1467828418230563e-05,
      "loss": 0.7088,
      "step": 102500
    },
    {
      "epoch": 8.96,
      "eval_loss": 0.680884063243866,
      "eval_runtime": 3113.4696,
      "eval_samples_per_second": 24.762,
      "eval_steps_per_second": 1.548,
      "step": 102500
    },
    {
      "epoch": 9.0,
      "learning_rate": 1.0989084641899656e-05,
      "loss": 0.7039,
      "step": 103000
    },
    {
      "epoch": 9.04,
      "learning_rate": 1.0510340865568748e-05,
      "loss": 0.6983,
      "step": 103500
    },
    {
      "epoch": 9.09,
      "learning_rate": 1.0032554576790503e-05,
      "loss": 0.7028,
      "step": 104000
    },
    {
      "epoch": 9.13,
      "learning_rate": 9.553810800459595e-06,
      "loss": 0.7002,
      "step": 104500
    },
    {
      "epoch": 9.18,
      "learning_rate": 9.075067024128686e-06,
      "loss": 0.7002,
      "step": 105000
    },
    {
      "epoch": 9.18,
      "eval_loss": 0.6783619523048401,
      "eval_runtime": 3133.9738,
      "eval_samples_per_second": 24.6,
      "eval_steps_per_second": 1.538,
      "step": 105000
    },
    {
      "epoch": 9.22,
      "learning_rate": 8.596323247797778e-06,
      "loss": 0.699,
      "step": 105500
    },
    {
      "epoch": 9.26,
      "learning_rate": 8.11757947146687e-06,
      "loss": 0.6927,
      "step": 106000
    },
    {
      "epoch": 9.31,
      "learning_rate": 7.638835695135963e-06,
      "loss": 0.6976,
      "step": 106500
    },
    {
      "epoch": 9.35,
      "learning_rate": 7.1600919188050565e-06,
      "loss": 0.6978,
      "step": 107000
    },
    {
      "epoch": 9.39,
      "learning_rate": 6.681348142474147e-06,
      "loss": 0.6953,
      "step": 107500
    },
    {
      "epoch": 9.39,
      "eval_loss": 0.6736627221107483,
      "eval_runtime": 3133.4547,
      "eval_samples_per_second": 24.604,
      "eval_steps_per_second": 1.538,
      "step": 107500
    },
    {
      "epoch": 9.44,
      "learning_rate": 6.2026043661432406e-06,
      "loss": 0.6892,
      "step": 108000
    },
    {
      "epoch": 9.48,
      "learning_rate": 5.723860589812333e-06,
      "loss": 0.6989,
      "step": 108500
    },
    {
      "epoch": 9.52,
      "learning_rate": 5.2451168134814254e-06,
      "loss": 0.6975,
      "step": 109000
    },
    {
      "epoch": 9.57,
      "learning_rate": 4.767330524703179e-06,
      "loss": 0.6977,
      "step": 109500
    },
    {
      "epoch": 9.61,
      "learning_rate": 4.288586748372271e-06,
      "loss": 0.695,
      "step": 110000
    },
    {
      "epoch": 9.61,
      "eval_loss": 0.6714410185813904,
      "eval_runtime": 3139.9991,
      "eval_samples_per_second": 24.553,
      "eval_steps_per_second": 1.535,
      "step": 110000
    },
    {
      "epoch": 9.66,
      "learning_rate": 3.810800459594025e-06,
      "loss": 0.6947,
      "step": 110500
    },
    {
      "epoch": 9.7,
      "learning_rate": 3.3320566832631176e-06,
      "loss": 0.6967,
      "step": 111000
    },
    {
      "epoch": 9.74,
      "learning_rate": 2.85331290693221e-06,
      "loss": 0.6904,
      "step": 111500
    },
    {
      "epoch": 9.79,
      "learning_rate": 2.3745691306013025e-06,
      "loss": 0.6949,
      "step": 112000
    },
    {
      "epoch": 9.83,
      "learning_rate": 1.8958253542703947e-06,
      "loss": 0.6871,
      "step": 112500
    },
    {
      "epoch": 9.83,
      "eval_loss": 0.6687204837799072,
      "eval_runtime": 3111.3683,
      "eval_samples_per_second": 24.779,
      "eval_steps_per_second": 1.549,
      "step": 112500
    },
    {
      "epoch": 9.87,
      "learning_rate": 3.6311394284300384e-05,
      "loss": 0.7034,
      "step": 113000
    },
    {
      "epoch": 9.92,
      "learning_rate": 3.600210317951256e-05,
      "loss": 0.7149,
      "step": 113500
    },
    {
      "epoch": 9.96,
      "learning_rate": 3.5692812074724736e-05,
      "loss": 0.716,
      "step": 114000
    },
    {
      "epoch": 10.01,
      "learning_rate": 3.5383520969936905e-05,
      "loss": 0.7236,
      "step": 114500
    },
    {
      "epoch": 10.05,
      "learning_rate": 3.5074229865149074e-05,
      "loss": 0.7161,
      "step": 115000
    },
    {
      "epoch": 10.05,
      "eval_loss": 0.6960982084274292,
      "eval_runtime": 3127.0335,
      "eval_samples_per_second": 24.655,
      "eval_steps_per_second": 1.541,
      "step": 115000
    },
    {
      "epoch": 10.09,
      "learning_rate": 3.476493876036126e-05,
      "loss": 0.7149,
      "step": 115500
    },
    {
      "epoch": 10.14,
      "learning_rate": 3.4455647655573426e-05,
      "loss": 0.7203,
      "step": 116000
    },
    {
      "epoch": 10.18,
      "learning_rate": 3.41463565507856e-05,
      "loss": 0.7259,
      "step": 116500
    },
    {
      "epoch": 10.22,
      "learning_rate": 3.383706544599777e-05,
      "loss": 0.7213,
      "step": 117000
    },
    {
      "epoch": 10.27,
      "learning_rate": 3.35290115056291e-05,
      "loss": 0.7265,
      "step": 117500
    },
    {
      "epoch": 10.27,
      "eval_loss": 0.7006255984306335,
      "eval_runtime": 3139.1903,
      "eval_samples_per_second": 24.559,
      "eval_steps_per_second": 1.535,
      "step": 117500
    },
    {
      "epoch": 10.31,
      "learning_rate": 3.321972040084128e-05,
      "loss": 0.7228,
      "step": 118000
    },
    {
      "epoch": 10.35,
      "learning_rate": 3.2910429296053447e-05,
      "loss": 0.7214,
      "step": 118500
    },
    {
      "epoch": 10.4,
      "learning_rate": 3.260113819126562e-05,
      "loss": 0.7309,
      "step": 119000
    },
    {
      "epoch": 10.44,
      "learning_rate": 3.229184708647779e-05,
      "loss": 0.7208,
      "step": 119500
    },
    {
      "epoch": 10.49,
      "learning_rate": 3.198255598168997e-05,
      "loss": 0.7284,
      "step": 120000
    },
    {
      "epoch": 10.49,
      "eval_loss": 0.6941153407096863,
      "eval_runtime": 3132.0931,
      "eval_samples_per_second": 24.615,
      "eval_steps_per_second": 1.539,
      "step": 120000
    },
    {
      "epoch": 10.53,
      "learning_rate": 3.167326487690214e-05,
      "loss": 0.7233,
      "step": 120500
    },
    {
      "epoch": 10.57,
      "learning_rate": 3.136397377211431e-05,
      "loss": 0.7219,
      "step": 121000
    },
    {
      "epoch": 10.62,
      "learning_rate": 3.105468266732649e-05,
      "loss": 0.7163,
      "step": 121500
    },
    {
      "epoch": 10.66,
      "learning_rate": 3.0745391562538664e-05,
      "loss": 0.7223,
      "step": 122000
    },
    {
      "epoch": 10.7,
      "learning_rate": 3.0436100457750836e-05,
      "loss": 0.724,
      "step": 122500
    },
    {
      "epoch": 10.7,
      "eval_loss": 0.6886956095695496,
      "eval_runtime": 3117.6812,
      "eval_samples_per_second": 24.729,
      "eval_steps_per_second": 1.546,
      "step": 122500
    },
    {
      "epoch": 10.75,
      "learning_rate": 3.012680935296301e-05,
      "loss": 0.7181,
      "step": 123000
    },
    {
      "epoch": 10.79,
      "learning_rate": 2.9817518248175185e-05,
      "loss": 0.7215,
      "step": 123500
    },
    {
      "epoch": 10.84,
      "learning_rate": 2.9508845725596935e-05,
      "loss": 0.726,
      "step": 124000
    },
    {
      "epoch": 10.88,
      "learning_rate": 2.9199554620809104e-05,
      "loss": 0.7205,
      "step": 124500
    },
    {
      "epoch": 10.92,
      "learning_rate": 2.8890882098230854e-05,
      "loss": 0.7266,
      "step": 125000
    },
    {
      "epoch": 10.92,
      "eval_loss": 0.6931244134902954,
      "eval_runtime": 3180.6149,
      "eval_samples_per_second": 24.239,
      "eval_steps_per_second": 1.515,
      "step": 125000
    },
    {
      "epoch": 10.97,
      "learning_rate": 2.858159099344303e-05,
      "loss": 0.7214,
      "step": 125500
    },
    {
      "epoch": 11.01,
      "learning_rate": 2.8272299888655202e-05,
      "loss": 0.7165,
      "step": 126000
    },
    {
      "epoch": 11.05,
      "learning_rate": 2.7963627366076955e-05,
      "loss": 0.7044,
      "step": 126500
    },
    {
      "epoch": 11.1,
      "learning_rate": 2.7654954843498705e-05,
      "loss": 0.7071,
      "step": 127000
    },
    {
      "epoch": 11.14,
      "learning_rate": 2.7345663738710874e-05,
      "loss": 0.7051,
      "step": 127500
    },
    {
      "epoch": 11.14,
      "eval_loss": 0.6846074461936951,
      "eval_runtime": 3127.4042,
      "eval_samples_per_second": 24.652,
      "eval_steps_per_second": 1.541,
      "step": 127500
    },
    {
      "epoch": 11.18,
      "learning_rate": 2.703637263392305e-05,
      "loss": 0.7139,
      "step": 128000
    },
    {
      "epoch": 11.23,
      "learning_rate": 2.6727081529135223e-05,
      "loss": 0.7057,
      "step": 128500
    },
    {
      "epoch": 11.27,
      "learning_rate": 2.64177904243474e-05,
      "loss": 0.7104,
      "step": 129000
    },
    {
      "epoch": 11.32,
      "learning_rate": 2.610849931955957e-05,
      "loss": 0.7036,
      "step": 129500
    },
    {
      "epoch": 11.36,
      "learning_rate": 2.5799208214771743e-05,
      "loss": 0.7106,
      "step": 130000
    },
    {
      "epoch": 11.36,
      "eval_loss": 0.68162602186203,
      "eval_runtime": 3071.6142,
      "eval_samples_per_second": 25.1,
      "eval_steps_per_second": 1.569,
      "step": 130000
    },
    {
      "epoch": 11.4,
      "learning_rate": 2.548991710998392e-05,
      "loss": 0.7072,
      "step": 130500
    },
    {
      "epoch": 11.45,
      "learning_rate": 2.5180626005196088e-05,
      "loss": 0.7073,
      "step": 131000
    },
    {
      "epoch": 11.49,
      "learning_rate": 2.4871334900408268e-05,
      "loss": 0.7109,
      "step": 131500
    },
    {
      "epoch": 11.53,
      "learning_rate": 2.456204379562044e-05,
      "loss": 0.7041,
      "step": 132000
    },
    {
      "epoch": 11.58,
      "learning_rate": 2.4252752690832612e-05,
      "loss": 0.7011,
      "step": 132500
    },
    {
      "epoch": 11.58,
      "eval_loss": 0.6830089688301086,
      "eval_runtime": 3073.7046,
      "eval_samples_per_second": 25.082,
      "eval_steps_per_second": 1.568,
      "step": 132500
    },
    {
      "epoch": 11.62,
      "learning_rate": 2.3943461586044785e-05,
      "loss": 0.7041,
      "step": 133000
    },
    {
      "epoch": 11.67,
      "learning_rate": 2.3634789063466535e-05,
      "loss": 0.7057,
      "step": 133500
    },
    {
      "epoch": 11.71,
      "learning_rate": 2.332549795867871e-05,
      "loss": 0.7068,
      "step": 134000
    },
    {
      "epoch": 11.75,
      "learning_rate": 2.3016825436100457e-05,
      "loss": 0.7073,
      "step": 134500
    },
    {
      "epoch": 11.8,
      "learning_rate": 2.2707534331312633e-05,
      "loss": 0.6997,
      "step": 135000
    },
    {
      "epoch": 11.8,
      "eval_loss": 0.6784160137176514,
      "eval_runtime": 3076.1607,
      "eval_samples_per_second": 25.062,
      "eval_steps_per_second": 1.567,
      "step": 135000
    },
    {
      "epoch": 11.84,
      "learning_rate": 2.2398243226524806e-05,
      "loss": 0.7062,
      "step": 135500
    },
    {
      "epoch": 11.88,
      "learning_rate": 2.208895212173698e-05,
      "loss": 0.6982,
      "step": 136000
    },
    {
      "epoch": 11.93,
      "learning_rate": 2.1779661016949154e-05,
      "loss": 0.7,
      "step": 136500
    },
    {
      "epoch": 11.97,
      "learning_rate": 2.147036991216133e-05,
      "loss": 0.6981,
      "step": 137000
    },
    {
      "epoch": 12.02,
      "learning_rate": 2.1161697389583076e-05,
      "loss": 0.6969,
      "step": 137500
    },
    {
      "epoch": 12.02,
      "eval_loss": 0.673393189907074,
      "eval_runtime": 3053.9536,
      "eval_samples_per_second": 25.245,
      "eval_steps_per_second": 1.578,
      "step": 137500
    },
    {
      "epoch": 12.06,
      "learning_rate": 2.0852406284795252e-05,
      "loss": 0.6961,
      "step": 138000
    },
    {
      "epoch": 12.1,
      "learning_rate": 2.0543115180007425e-05,
      "loss": 0.6941,
      "step": 138500
    },
    {
      "epoch": 12.15,
      "learning_rate": 2.0233824075219597e-05,
      "loss": 0.6938,
      "step": 139000
    },
    {
      "epoch": 12.19,
      "learning_rate": 1.9924532970431773e-05,
      "loss": 0.6968,
      "step": 139500
    },
    {
      "epoch": 12.23,
      "learning_rate": 1.9615241865643945e-05,
      "loss": 0.6968,
      "step": 140000
    },
    {
      "epoch": 12.23,
      "eval_loss": 0.6708864569664001,
      "eval_runtime": 3020.4771,
      "eval_samples_per_second": 25.524,
      "eval_steps_per_second": 1.595,
      "step": 140000
    },
    {
      "epoch": 12.28,
      "learning_rate": 1.9305950760856118e-05,
      "loss": 0.6928,
      "step": 140500
    },
    {
      "epoch": 12.32,
      "learning_rate": 1.899665965606829e-05,
      "loss": 0.6975,
      "step": 141000
    },
    {
      "epoch": 12.36,
      "learning_rate": 1.8687368551280466e-05,
      "loss": 0.6912,
      "step": 141500
    },
    {
      "epoch": 12.41,
      "learning_rate": 1.837807744649264e-05,
      "loss": 0.6923,
      "step": 142000
    },
    {
      "epoch": 12.45,
      "learning_rate": 1.8068786341704814e-05,
      "loss": 0.6867,
      "step": 142500
    },
    {
      "epoch": 12.45,
      "eval_loss": 0.6655944585800171,
      "eval_runtime": 3033.3682,
      "eval_samples_per_second": 25.416,
      "eval_steps_per_second": 1.589,
      "step": 142500
    },
    {
      "epoch": 12.5,
      "learning_rate": 1.7759495236916987e-05,
      "loss": 0.6894,
      "step": 143000
    },
    {
      "epoch": 12.54,
      "learning_rate": 1.745020413212916e-05,
      "loss": 0.6861,
      "step": 143500
    },
    {
      "epoch": 12.58,
      "learning_rate": 1.714153160955091e-05,
      "loss": 0.6917,
      "step": 144000
    },
    {
      "epoch": 12.63,
      "learning_rate": 1.6832240504763082e-05,
      "loss": 0.688,
      "step": 144500
    },
    {
      "epoch": 12.67,
      "learning_rate": 1.6522949399975258e-05,
      "loss": 0.6925,
      "step": 145000
    },
    {
      "epoch": 12.67,
      "eval_loss": 0.6661481261253357,
      "eval_runtime": 3083.2268,
      "eval_samples_per_second": 25.005,
      "eval_steps_per_second": 1.563,
      "step": 145000
    },
    {
      "epoch": 12.71,
      "learning_rate": 1.621365829518743e-05,
      "loss": 0.6943,
      "step": 145500
    },
    {
      "epoch": 12.76,
      "learning_rate": 1.5904367190399606e-05,
      "loss": 0.6891,
      "step": 146000
    },
    {
      "epoch": 12.8,
      "learning_rate": 1.559507608561178e-05,
      "loss": 0.6848,
      "step": 146500
    },
    {
      "epoch": 12.85,
      "learning_rate": 1.5285784980823954e-05,
      "loss": 0.6816,
      "step": 147000
    },
    {
      "epoch": 12.89,
      "learning_rate": 1.49771124582457e-05,
      "loss": 0.6795,
      "step": 147500
    },
    {
      "epoch": 12.89,
      "eval_loss": 0.6605859398841858,
      "eval_runtime": 3089.4045,
      "eval_samples_per_second": 24.955,
      "eval_steps_per_second": 1.56,
      "step": 147500
    },
    {
      "epoch": 12.93,
      "learning_rate": 1.4667821353457875e-05,
      "loss": 0.6779,
      "step": 148000
    },
    {
      "epoch": 12.98,
      "learning_rate": 1.4358530248670049e-05,
      "loss": 0.6846,
      "step": 148500
    },
    {
      "epoch": 13.02,
      "learning_rate": 1.4049239143882223e-05,
      "loss": 0.6739,
      "step": 149000
    },
    {
      "epoch": 13.06,
      "learning_rate": 1.3740566621303971e-05,
      "loss": 0.6772,
      "step": 149500
    },
    {
      "epoch": 13.11,
      "learning_rate": 1.3431275516516146e-05,
      "loss": 0.6774,
      "step": 150000
    },
    {
      "epoch": 13.11,
      "eval_loss": 0.6616868376731873,
      "eval_runtime": 3137.3755,
      "eval_samples_per_second": 24.573,
      "eval_steps_per_second": 1.536,
      "step": 150000
    },
    {
      "epoch": 13.15,
      "learning_rate": 1.3122602993937894e-05,
      "loss": 0.6768,
      "step": 150500
    },
    {
      "epoch": 13.19,
      "learning_rate": 1.2813311889150068e-05,
      "loss": 0.6794,
      "step": 151000
    },
    {
      "epoch": 13.24,
      "learning_rate": 1.2504020784362242e-05,
      "loss": 0.6816,
      "step": 151500
    },
    {
      "epoch": 13.28,
      "learning_rate": 1.2194729679574415e-05,
      "loss": 0.6748,
      "step": 152000
    },
    {
      "epoch": 13.33,
      "learning_rate": 1.1886057156996165e-05,
      "loss": 0.6756,
      "step": 152500
    },
    {
      "epoch": 13.33,
      "eval_loss": 0.6562890410423279,
      "eval_runtime": 3110.9217,
      "eval_samples_per_second": 24.782,
      "eval_steps_per_second": 1.549,
      "step": 152500
    },
    {
      "epoch": 13.37,
      "learning_rate": 1.1576766052208339e-05,
      "loss": 0.672,
      "step": 153000
    },
    {
      "epoch": 13.41,
      "learning_rate": 1.1267474947420513e-05,
      "loss": 0.6711,
      "step": 153500
    },
    {
      "epoch": 13.46,
      "learning_rate": 1.0958183842632687e-05,
      "loss": 0.6731,
      "step": 154000
    },
    {
      "epoch": 13.5,
      "learning_rate": 1.064889273784486e-05,
      "loss": 0.6707,
      "step": 154500
    },
    {
      "epoch": 13.54,
      "learning_rate": 1.0339601633057034e-05,
      "loss": 0.6728,
      "step": 155000
    },
    {
      "epoch": 13.54,
      "eval_loss": 0.6547101140022278,
      "eval_runtime": 3126.1621,
      "eval_samples_per_second": 24.662,
      "eval_steps_per_second": 1.542,
      "step": 155000
    },
    {
      "epoch": 13.59,
      "learning_rate": 1.0030929110478784e-05,
      "loss": 0.6727,
      "step": 155500
    },
    {
      "epoch": 13.63,
      "learning_rate": 9.721638005690958e-06,
      "loss": 0.6679,
      "step": 156000
    },
    {
      "epoch": 13.68,
      "learning_rate": 9.41234690090313e-06,
      "loss": 0.6626,
      "step": 156500
    },
    {
      "epoch": 13.72,
      "learning_rate": 9.103055796115304e-06,
      "loss": 0.6703,
      "step": 157000
    },
    {
      "epoch": 13.76,
      "learning_rate": 8.793764691327477e-06,
      "loss": 0.6732,
      "step": 157500
    },
    {
      "epoch": 13.76,
      "eval_loss": 0.6520426273345947,
      "eval_runtime": 3118.7072,
      "eval_samples_per_second": 24.72,
      "eval_steps_per_second": 1.545,
      "step": 157500
    },
    {
      "epoch": 13.81,
      "learning_rate": 8.484473586539651e-06,
      "loss": 0.6734,
      "step": 158000
    },
    {
      "epoch": 13.85,
      "learning_rate": 8.175182481751825e-06,
      "loss": 0.6707,
      "step": 158500
    },
    {
      "epoch": 13.89,
      "learning_rate": 7.865891376964e-06,
      "loss": 0.667,
      "step": 159000
    },
    {
      "epoch": 13.94,
      "learning_rate": 7.557218854385749e-06,
      "loss": 0.6667,
      "step": 159500
    },
    {
      "epoch": 13.98,
      "learning_rate": 7.2485463318074974e-06,
      "loss": 0.6704,
      "step": 160000
    },
    {
      "epoch": 13.98,
      "eval_loss": 0.6492029428482056,
      "eval_runtime": 3114.0251,
      "eval_samples_per_second": 24.758,
      "eval_steps_per_second": 1.548,
      "step": 160000
    },
    {
      "epoch": 14.02,
      "learning_rate": 6.939255227019672e-06,
      "loss": 0.6631,
      "step": 160500
    },
    {
      "epoch": 14.07,
      "learning_rate": 6.629964122231845e-06,
      "loss": 0.663,
      "step": 161000
    },
    {
      "epoch": 14.11,
      "learning_rate": 6.320673017444019e-06,
      "loss": 0.6661,
      "step": 161500
    },
    {
      "epoch": 14.16,
      "learning_rate": 6.011381912656192e-06,
      "loss": 0.6641,
      "step": 162000
    },
    {
      "epoch": 14.2,
      "learning_rate": 5.7020908078683665e-06,
      "loss": 0.6666,
      "step": 162500
    },
    {
      "epoch": 14.2,
      "eval_loss": 0.6445870995521545,
      "eval_runtime": 3137.0802,
      "eval_samples_per_second": 24.576,
      "eval_steps_per_second": 1.536,
      "step": 162500
    },
    {
      "epoch": 14.24,
      "learning_rate": 5.392799703080539e-06,
      "loss": 0.6612,
      "step": 163000
    },
    {
      "epoch": 14.29,
      "learning_rate": 5.083508598292713e-06,
      "loss": 0.6645,
      "step": 163500
    },
    {
      "epoch": 14.33,
      "learning_rate": 4.774217493504887e-06,
      "loss": 0.6613,
      "step": 164000
    },
    {
      "epoch": 14.37,
      "learning_rate": 4.464926388717061e-06,
      "loss": 0.6608,
      "step": 164500
    },
    {
      "epoch": 14.42,
      "learning_rate": 4.155635283929235e-06,
      "loss": 0.6615,
      "step": 165000
    },
    {
      "epoch": 14.42,
      "eval_loss": 0.6487849950790405,
      "eval_runtime": 3095.1776,
      "eval_samples_per_second": 24.908,
      "eval_steps_per_second": 1.557,
      "step": 165000
    },
    {
      "epoch": 14.46,
      "learning_rate": 3.846344179141408e-06,
      "loss": 0.6598,
      "step": 165500
    },
    {
      "epoch": 14.51,
      "learning_rate": 3.537671656563157e-06,
      "loss": 0.6569,
      "step": 166000
    },
    {
      "epoch": 14.55,
      "learning_rate": 3.2289991339849062e-06,
      "loss": 0.6587,
      "step": 166500
    },
    {
      "epoch": 14.59,
      "learning_rate": 2.920326611406656e-06,
      "loss": 0.6575,
      "step": 167000
    },
    {
      "epoch": 14.64,
      "learning_rate": 2.61103550661883e-06,
      "loss": 0.6638,
      "step": 167500
    },
    {
      "epoch": 14.64,
      "eval_loss": 0.6522655487060547,
      "eval_runtime": 3117.0073,
      "eval_samples_per_second": 24.734,
      "eval_steps_per_second": 1.546,
      "step": 167500
    },
    {
      "epoch": 14.68,
      "learning_rate": 2.3017444018310032e-06,
      "loss": 0.6553,
      "step": 168000
    },
    {
      "epoch": 14.72,
      "learning_rate": 1.9930718792527527e-06,
      "loss": 0.6617,
      "step": 168500
    },
    {
      "epoch": 14.77,
      "learning_rate": 1.6837807744649262e-06,
      "loss": 0.6583,
      "step": 169000
    },
    {
      "epoch": 14.81,
      "learning_rate": 1.3744896696771002e-06,
      "loss": 0.6579,
      "step": 169500
    },
    {
      "epoch": 14.85,
      "learning_rate": 1.0651985648892737e-06,
      "loss": 0.6588,
      "step": 170000
    },
    {
      "epoch": 14.85,
      "eval_loss": 0.6414651274681091,
      "eval_runtime": 3135.6411,
      "eval_samples_per_second": 24.587,
      "eval_steps_per_second": 1.537,
      "step": 170000
    },
    {
      "epoch": 14.9,
      "learning_rate": 7.559074601014475e-07,
      "loss": 0.6589,
      "step": 170500
    },
    {
      "epoch": 14.94,
      "learning_rate": 4.4661635531362117e-07,
      "loss": 0.6582,
      "step": 171000
    },
    {
      "epoch": 14.99,
      "learning_rate": 1.3732525052579488e-07,
      "loss": 0.6569,
      "step": 171500
    },
    {
      "epoch": 15.0,
      "step": 171660,
      "total_flos": 2.5739338132512957e+18,
      "train_loss": 0.23807151880656507,
      "train_runtime": 725915.7505,
      "train_samples_per_second": 30.268,
      "train_steps_per_second": 0.236
    }
  ],
  "max_steps": 171660,
  "num_train_epochs": 15,
  "total_flos": 2.5739338132512957e+18,
  "trial_name": null,
  "trial_params": null
}