{
  "best_metric": 0.3365662097930908,
  "best_model_checkpoint": "/output/xlm-roberta-base-finetuned-full/checkpoint-205000",
  "epoch": 2.9999680346646302,
  "global_step": 211164,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 4.9881608607527804e-05,
      "loss": 1.5253,
      "step": 500
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.97632172150556e-05,
      "loss": 1.1464,
      "step": 1000
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.964482582258339e-05,
      "loss": 1.0217,
      "step": 1500
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.9526434430111195e-05,
      "loss": 0.9542,
      "step": 2000
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.9408043037638996e-05,
      "loss": 0.9065,
      "step": 2500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.928965164516679e-05,
      "loss": 0.8708,
      "step": 3000
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.917126025269459e-05,
      "loss": 0.8342,
      "step": 3500
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.905286886022239e-05,
      "loss": 0.806,
      "step": 4000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.893447746775019e-05,
      "loss": 0.79,
      "step": 4500
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.881608607527798e-05,
      "loss": 0.7697,
      "step": 5000
    },
    {
      "epoch": 0.07,
      "eval_accuracy": 0.8496130075487093,
      "eval_loss": 0.6976155042648315,
      "eval_runtime": 4472.5845,
      "eval_samples_per_second": 31.522,
      "eval_steps_per_second": 7.881,
      "step": 5000
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.8697694682805785e-05,
      "loss": 0.7583,
      "step": 5500
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.8579303290333586e-05,
      "loss": 0.7434,
      "step": 6000
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.846091189786138e-05,
      "loss": 0.7366,
      "step": 6500
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.8342520505389175e-05,
      "loss": 0.7189,
      "step": 7000
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.822412911291698e-05,
      "loss": 0.7056,
      "step": 7500
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.810573772044478e-05,
      "loss": 0.6993,
      "step": 8000
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.798734632797257e-05,
      "loss": 0.682,
      "step": 8500
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.786895493550037e-05,
      "loss": 0.6792,
      "step": 9000
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.775056354302817e-05,
      "loss": 0.6652,
      "step": 9500
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.763217215055597e-05,
      "loss": 0.6638,
      "step": 10000
    },
    {
      "epoch": 0.14,
      "eval_accuracy": 0.866376552224583,
      "eval_loss": 0.6067455410957336,
      "eval_runtime": 4473.7903,
      "eval_samples_per_second": 31.514,
      "eval_steps_per_second": 7.879,
      "step": 10000
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.7513780758083765e-05,
      "loss": 0.6513,
      "step": 10500
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.739538936561157e-05,
      "loss": 0.6492,
      "step": 11000
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.727699797313936e-05,
      "loss": 0.6443,
      "step": 11500
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.715860658066716e-05,
      "loss": 0.6369,
      "step": 12000
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.704021518819496e-05,
      "loss": 0.6319,
      "step": 12500
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.692182379572276e-05,
      "loss": 0.6247,
      "step": 13000
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.680343240325056e-05,
      "loss": 0.6208,
      "step": 13500
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.6685041010778355e-05,
      "loss": 0.619,
      "step": 14000
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.656664961830615e-05,
      "loss": 0.6193,
      "step": 14500
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.644825822583395e-05,
      "loss": 0.6101,
      "step": 15000
    },
    {
      "epoch": 0.21,
      "eval_accuracy": 0.8746108749481479,
      "eval_loss": 0.5597154498100281,
      "eval_runtime": 4477.6442,
      "eval_samples_per_second": 31.487,
      "eval_steps_per_second": 7.872,
      "step": 15000
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.632986683336175e-05,
      "loss": 0.6035,
      "step": 15500
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.621147544088955e-05,
      "loss": 0.6028,
      "step": 16000
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.609308404841734e-05,
      "loss": 0.5945,
      "step": 16500
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.5974692655945144e-05,
      "loss": 0.5905,
      "step": 17000
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.5856301263472945e-05,
      "loss": 0.5869,
      "step": 17500
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.573790987100074e-05,
      "loss": 0.5849,
      "step": 18000
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.561951847852854e-05,
      "loss": 0.579,
      "step": 18500
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.5501127086056336e-05,
      "loss": 0.5755,
      "step": 19000
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.538273569358414e-05,
      "loss": 0.5809,
      "step": 19500
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.526434430111193e-05,
      "loss": 0.5702,
      "step": 20000
    },
    {
      "epoch": 0.28,
      "eval_accuracy": 0.8805759285133897,
      "eval_loss": 0.5277929306030273,
      "eval_runtime": 4476.5281,
      "eval_samples_per_second": 31.494,
      "eval_steps_per_second": 7.874,
      "step": 20000
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.5145952908639734e-05,
      "loss": 0.5763,
      "step": 20500
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.5027561516167535e-05,
      "loss": 0.564,
      "step": 21000
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.490917012369533e-05,
      "loss": 0.566,
      "step": 21500
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.4790778731223124e-05,
      "loss": 0.5614,
      "step": 22000
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.4672387338750926e-05,
      "loss": 0.5602,
      "step": 22500
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.455399594627873e-05,
      "loss": 0.5566,
      "step": 23000
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.443560455380652e-05,
      "loss": 0.5521,
      "step": 23500
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.431721316133432e-05,
      "loss": 0.5505,
      "step": 24000
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.419882176886212e-05,
      "loss": 0.5505,
      "step": 24500
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.408043037638992e-05,
      "loss": 0.5433,
      "step": 25000
    },
    {
      "epoch": 0.36,
      "eval_accuracy": 0.8846786107773023,
      "eval_loss": 0.5012275576591492,
      "eval_runtime": 4482.4883,
      "eval_samples_per_second": 31.453,
      "eval_steps_per_second": 7.863,
      "step": 25000
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.3962038983917714e-05,
      "loss": 0.5462,
      "step": 25500
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.3843647591445516e-05,
      "loss": 0.5402,
      "step": 26000
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.372525619897331e-05,
      "loss": 0.543,
      "step": 26500
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.3606864806501105e-05,
      "loss": 0.5411,
      "step": 27000
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.348847341402891e-05,
      "loss": 0.5359,
      "step": 27500
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.337008202155671e-05,
      "loss": 0.5375,
      "step": 28000
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.325169062908451e-05,
      "loss": 0.5293,
      "step": 28500
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.3133299236612304e-05,
      "loss": 0.5255,
      "step": 29000
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.30149078441401e-05,
      "loss": 0.5281,
      "step": 29500
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.28965164516679e-05,
      "loss": 0.5308,
      "step": 30000
    },
    {
      "epoch": 0.43,
      "eval_accuracy": 0.8875591084578017,
      "eval_loss": 0.48577040433883667,
      "eval_runtime": 4481.7631,
      "eval_samples_per_second": 31.458,
      "eval_steps_per_second": 7.865,
      "step": 30000
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.27781250591957e-05,
      "loss": 0.5304,
      "step": 30500
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.2659733666723497e-05,
      "loss": 0.5272,
      "step": 31000
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.254134227425129e-05,
      "loss": 0.523,
      "step": 31500
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.242295088177909e-05,
      "loss": 0.5197,
      "step": 32000
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.2304559489306894e-05,
      "loss": 0.5212,
      "step": 32500
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.218616809683469e-05,
      "loss": 0.5203,
      "step": 33000
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.206777670436249e-05,
      "loss": 0.5132,
      "step": 33500
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.1949385311890285e-05,
      "loss": 0.5154,
      "step": 34000
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.183099391941808e-05,
      "loss": 0.5193,
      "step": 34500
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.171260252694588e-05,
      "loss": 0.5118,
      "step": 35000
    },
    {
      "epoch": 0.5,
      "eval_accuracy": 0.890265619775884,
      "eval_loss": 0.4702432155609131,
      "eval_runtime": 4481.2911,
      "eval_samples_per_second": 31.461,
      "eval_steps_per_second": 7.865,
      "step": 35000
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.159421113447368e-05,
      "loss": 0.5079,
      "step": 35500
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.1475819742001484e-05,
      "loss": 0.5104,
      "step": 36000
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.135742834952928e-05,
      "loss": 0.5047,
      "step": 36500
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.1239036957057073e-05,
      "loss": 0.5102,
      "step": 37000
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.1120645564584875e-05,
      "loss": 0.5054,
      "step": 37500
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.1002254172112676e-05,
      "loss": 0.5052,
      "step": 38000
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.088386277964047e-05,
      "loss": 0.502,
      "step": 38500
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.076547138716827e-05,
      "loss": 0.5037,
      "step": 39000
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.064707999469607e-05,
      "loss": 0.4998,
      "step": 39500
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.052868860222386e-05,
      "loss": 0.4971,
      "step": 40000
    },
    {
      "epoch": 0.57,
      "eval_accuracy": 0.8922914063763604,
      "eval_loss": 0.4587230086326599,
      "eval_runtime": 4481.3983,
      "eval_samples_per_second": 31.46,
      "eval_steps_per_second": 7.865,
      "step": 40000
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.041029720975166e-05,
      "loss": 0.5026,
      "step": 40500
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.0291905817279465e-05,
      "loss": 0.496,
      "step": 41000
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.017351442480726e-05,
      "loss": 0.4965,
      "step": 41500
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.0055123032335054e-05,
      "loss": 0.496,
      "step": 42000
    },
    {
      "epoch": 0.6,
      "learning_rate": 3.9936731639862856e-05,
      "loss": 0.4873,
      "step": 42500
    },
    {
      "epoch": 0.61,
      "learning_rate": 3.981834024739066e-05,
      "loss": 0.4977,
      "step": 43000
    },
    {
      "epoch": 0.62,
      "learning_rate": 3.969994885491846e-05,
      "loss": 0.4896,
      "step": 43500
    },
    {
      "epoch": 0.63,
      "learning_rate": 3.958155746244625e-05,
      "loss": 0.4858,
      "step": 44000
    },
    {
      "epoch": 0.63,
      "learning_rate": 3.946316606997405e-05,
      "loss": 0.4877,
      "step": 44500
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.934477467750185e-05,
      "loss": 0.4857,
      "step": 45000
    },
    {
      "epoch": 0.64,
      "eval_accuracy": 0.894234712380511,
      "eval_loss": 0.4484991431236267,
      "eval_runtime": 4480.0256,
      "eval_samples_per_second": 31.47,
      "eval_steps_per_second": 7.868,
      "step": 45000
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.922638328502965e-05,
      "loss": 0.4888,
      "step": 45500
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.9107991892557446e-05,
      "loss": 0.4827,
      "step": 46000
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.898960050008525e-05,
      "loss": 0.4799,
      "step": 46500
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.887120910761304e-05,
      "loss": 0.4815,
      "step": 47000
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.8752817715140836e-05,
      "loss": 0.4847,
      "step": 47500
    },
    {
      "epoch": 0.68,
      "learning_rate": 3.863442632266864e-05,
      "loss": 0.4845,
      "step": 48000
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.851603493019644e-05,
      "loss": 0.4832,
      "step": 48500
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.8397643537724234e-05,
      "loss": 0.4783,
      "step": 49000
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.827925214525203e-05,
      "loss": 0.4785,
      "step": 49500
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.816086075277983e-05,
      "loss": 0.4812,
      "step": 50000
    },
    {
      "epoch": 0.71,
      "eval_accuracy": 0.8959396424676428,
      "eval_loss": 0.43902069330215454,
      "eval_runtime": 4484.5246,
      "eval_samples_per_second": 31.438,
      "eval_steps_per_second": 7.86,
      "step": 50000
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.804246936030763e-05,
      "loss": 0.4753,
      "step": 50500
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.792407796783543e-05,
      "loss": 0.48,
      "step": 51000
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.780568657536323e-05,
      "loss": 0.4777,
      "step": 51500
    },
    {
      "epoch": 0.74,
      "learning_rate": 3.768729518289102e-05,
      "loss": 0.4748,
      "step": 52000
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.7568903790418824e-05,
      "loss": 0.4728,
      "step": 52500
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.745051239794662e-05,
      "loss": 0.4677,
      "step": 53000
    },
    {
      "epoch": 0.76,
      "learning_rate": 3.733212100547442e-05,
      "loss": 0.4731,
      "step": 53500
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.721372961300222e-05,
      "loss": 0.464,
      "step": 54000
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.7095338220530016e-05,
      "loss": 0.4708,
      "step": 54500
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.697694682805781e-05,
      "loss": 0.4716,
      "step": 55000
    },
    {
      "epoch": 0.78,
      "eval_accuracy": 0.8970586319380052,
      "eval_loss": 0.43407291173934937,
      "eval_runtime": 4489.7395,
      "eval_samples_per_second": 31.402,
      "eval_steps_per_second": 7.851,
      "step": 55000
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.685855543558561e-05,
      "loss": 0.469,
      "step": 55500
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.6740164043113414e-05,
      "loss": 0.4684,
      "step": 56000
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.662177265064121e-05,
      "loss": 0.4682,
      "step": 56500
    },
    {
      "epoch": 0.81,
      "learning_rate": 3.6503381258169e-05,
      "loss": 0.466,
      "step": 57000
    },
    {
      "epoch": 0.82,
      "learning_rate": 3.6384989865696805e-05,
      "loss": 0.4629,
      "step": 57500
    },
    {
      "epoch": 0.82,
      "learning_rate": 3.6266598473224606e-05,
      "loss": 0.4641,
      "step": 58000
    },
    {
      "epoch": 0.83,
      "learning_rate": 3.614820708075241e-05,
      "loss": 0.4602,
      "step": 58500
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.60298156882802e-05,
      "loss": 0.463,
      "step": 59000
    },
    {
      "epoch": 0.85,
      "learning_rate": 3.5911424295808e-05,
      "loss": 0.4586,
      "step": 59500
    },
    {
      "epoch": 0.85,
      "learning_rate": 3.57930329033358e-05,
      "loss": 0.4601,
      "step": 60000
    },
    {
      "epoch": 0.85,
      "eval_accuracy": 0.8986926934131002,
      "eval_loss": 0.42562565207481384,
      "eval_runtime": 4489.1177,
      "eval_samples_per_second": 31.406,
      "eval_steps_per_second": 7.852,
      "step": 60000
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.567464151086359e-05,
      "loss": 0.4612,
      "step": 60500
    },
    {
      "epoch": 0.87,
      "learning_rate": 3.5556250118391395e-05,
      "loss": 0.4609,
      "step": 61000
    },
    {
      "epoch": 0.87,
      "learning_rate": 3.5437858725919196e-05,
      "loss": 0.4552,
      "step": 61500
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.531946733344699e-05,
      "loss": 0.4624,
      "step": 62000
    },
    {
      "epoch": 0.89,
      "learning_rate": 3.5201075940974785e-05,
      "loss": 0.4572,
      "step": 62500
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.508268454850259e-05,
      "loss": 0.4577,
      "step": 63000
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.496429315603039e-05,
      "loss": 0.4527,
      "step": 63500
    },
    {
      "epoch": 0.91,
      "learning_rate": 3.484590176355819e-05,
      "loss": 0.4542,
      "step": 64000
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.472751037108598e-05,
      "loss": 0.4549,
      "step": 64500
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.460911897861378e-05,
      "loss": 0.4548,
      "step": 65000
    },
    {
      "epoch": 0.92,
      "eval_accuracy": 0.9000358274024964,
      "eval_loss": 0.418160617351532,
      "eval_runtime": 4478.7532,
      "eval_samples_per_second": 31.479,
      "eval_steps_per_second": 7.87,
      "step": 65000
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.449072758614158e-05,
      "loss": 0.4543,
      "step": 65500
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.4372336193669375e-05,
      "loss": 0.4563,
      "step": 66000
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.425394480119718e-05,
      "loss": 0.4534,
      "step": 66500
    },
    {
      "epoch": 0.95,
      "learning_rate": 3.413555340872497e-05,
      "loss": 0.4532,
      "step": 67000
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.401716201625277e-05,
      "loss": 0.4512,
      "step": 67500
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.389877062378057e-05,
      "loss": 0.4477,
      "step": 68000
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.378037923130837e-05,
      "loss": 0.4487,
      "step": 68500
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.366198783883617e-05,
      "loss": 0.4503,
      "step": 69000
    },
    {
      "epoch": 0.99,
      "learning_rate": 3.3543596446363965e-05,
      "loss": 0.445,
      "step": 69500
    },
    {
      "epoch": 0.99,
      "learning_rate": 3.342520505389176e-05,
      "loss": 0.4497,
      "step": 70000
    },
    {
      "epoch": 0.99,
      "eval_accuracy": 0.9010131119019583,
      "eval_loss": 0.4137929081916809,
      "eval_runtime": 4478.1902,
      "eval_samples_per_second": 31.483,
      "eval_steps_per_second": 7.871,
      "step": 70000
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.330681366141956e-05,
      "loss": 0.448,
      "step": 70500
    },
    {
      "epoch": 1.01,
      "learning_rate": 3.318842226894736e-05,
      "loss": 0.4412,
      "step": 71000
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.3070030876475164e-05,
      "loss": 0.4441,
      "step": 71500
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.295163948400295e-05,
      "loss": 0.4442,
      "step": 72000
    },
    {
      "epoch": 1.03,
      "learning_rate": 3.2833248091530754e-05,
      "loss": 0.4415,
      "step": 72500
    },
    {
      "epoch": 1.04,
      "learning_rate": 3.2714856699058555e-05,
      "loss": 0.4457,
      "step": 73000
    },
    {
      "epoch": 1.04,
      "learning_rate": 3.259646530658635e-05,
      "loss": 0.4438,
      "step": 73500
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.247807391411415e-05,
      "loss": 0.4453,
      "step": 74000
    },
    {
      "epoch": 1.06,
      "learning_rate": 3.2359682521641946e-05,
      "loss": 0.4407,
      "step": 74500
    },
    {
      "epoch": 1.07,
      "learning_rate": 3.224129112916975e-05,
      "loss": 0.4376,
      "step": 75000
    },
    {
      "epoch": 1.07,
      "eval_accuracy": 0.9019829116607336,
      "eval_loss": 0.408815860748291,
      "eval_runtime": 4480.6277,
      "eval_samples_per_second": 31.466,
      "eval_steps_per_second": 7.867,
      "step": 75000
    },
    {
      "epoch": 1.07,
      "learning_rate": 3.212289973669754e-05,
      "loss": 0.4371,
      "step": 75500
    },
    {
      "epoch": 1.08,
      "learning_rate": 3.2004508344225344e-05,
      "loss": 0.4397,
      "step": 76000
    },
    {
      "epoch": 1.09,
      "learning_rate": 3.1886116951753145e-05,
      "loss": 0.4312,
      "step": 76500
    },
    {
      "epoch": 1.09,
      "learning_rate": 3.176772555928094e-05,
      "loss": 0.435,
      "step": 77000
    },
    {
      "epoch": 1.1,
      "learning_rate": 3.1649334166808734e-05,
      "loss": 0.4398,
      "step": 77500
    },
    {
      "epoch": 1.11,
      "learning_rate": 3.1530942774336536e-05,
      "loss": 0.4393,
      "step": 78000
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.141255138186434e-05,
      "loss": 0.4371,
      "step": 78500
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.129415998939213e-05,
      "loss": 0.4355,
      "step": 79000
    },
    {
      "epoch": 1.13,
      "learning_rate": 3.117576859691993e-05,
      "loss": 0.4328,
      "step": 79500
    },
    {
      "epoch": 1.14,
      "learning_rate": 3.105737720444773e-05,
      "loss": 0.4349,
      "step": 80000
    },
    {
      "epoch": 1.14,
      "eval_accuracy": 0.9031310444806294,
      "eval_loss": 0.40232041478157043,
      "eval_runtime": 4482.6913,
      "eval_samples_per_second": 31.451,
      "eval_steps_per_second": 7.863,
      "step": 80000
    },
    {
      "epoch": 1.14,
      "learning_rate": 3.093898581197553e-05,
      "loss": 0.4324,
      "step": 80500
    },
    {
      "epoch": 1.15,
      "learning_rate": 3.0820594419503324e-05,
      "loss": 0.4326,
      "step": 81000
    },
    {
      "epoch": 1.16,
      "learning_rate": 3.0702203027031126e-05,
      "loss": 0.4308,
      "step": 81500
    },
    {
      "epoch": 1.16,
      "learning_rate": 3.058381163455892e-05,
      "loss": 0.4317,
      "step": 82000
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.0465420242086722e-05,
      "loss": 0.4322,
      "step": 82500
    },
    {
      "epoch": 1.18,
      "learning_rate": 3.0347028849614517e-05,
      "loss": 0.43,
      "step": 83000
    },
    {
      "epoch": 1.19,
      "learning_rate": 3.0228637457142318e-05,
      "loss": 0.4321,
      "step": 83500
    },
    {
      "epoch": 1.19,
      "learning_rate": 3.0110246064670116e-05,
      "loss": 0.4309,
      "step": 84000
    },
    {
      "epoch": 1.2,
      "learning_rate": 2.9991854672197918e-05,
      "loss": 0.4274,
      "step": 84500
    },
    {
      "epoch": 1.21,
      "learning_rate": 2.9873463279725712e-05,
      "loss": 0.4262,
      "step": 85000
    },
    {
      "epoch": 1.21,
      "eval_accuracy": 0.9039536284020224,
      "eval_loss": 0.39793965220451355,
      "eval_runtime": 4483.3415,
      "eval_samples_per_second": 31.447,
      "eval_steps_per_second": 7.862,
      "step": 85000
    },
    {
      "epoch": 1.21,
      "learning_rate": 2.975507188725351e-05,
      "loss": 0.4282,
      "step": 85500
    },
    {
      "epoch": 1.22,
      "learning_rate": 2.9636680494781312e-05,
      "loss": 0.4283,
      "step": 86000
    },
    {
      "epoch": 1.23,
      "learning_rate": 2.9518289102309103e-05,
      "loss": 0.4275,
      "step": 86500
    },
    {
      "epoch": 1.24,
      "learning_rate": 2.9399897709836905e-05,
      "loss": 0.4251,
      "step": 87000
    },
    {
      "epoch": 1.24,
      "learning_rate": 2.9281506317364703e-05,
      "loss": 0.4272,
      "step": 87500
    },
    {
      "epoch": 1.25,
      "learning_rate": 2.9163114924892504e-05,
      "loss": 0.428,
      "step": 88000
    },
    {
      "epoch": 1.26,
      "learning_rate": 2.90447235324203e-05,
      "loss": 0.4281,
      "step": 88500
    },
    {
      "epoch": 1.26,
      "learning_rate": 2.8926332139948097e-05,
      "loss": 0.4242,
      "step": 89000
    },
    {
      "epoch": 1.27,
      "learning_rate": 2.8807940747475898e-05,
      "loss": 0.4233,
      "step": 89500
    },
    {
      "epoch": 1.28,
      "learning_rate": 2.8689549355003696e-05,
      "loss": 0.4187,
      "step": 90000
    },
    {
      "epoch": 1.28,
      "eval_accuracy": 0.905044954941931,
      "eval_loss": 0.3930475413799286,
      "eval_runtime": 4483.2316,
      "eval_samples_per_second": 31.447,
      "eval_steps_per_second": 7.862,
      "step": 90000
    },
    {
      "epoch": 1.29,
      "learning_rate": 2.857115796253149e-05,
      "loss": 0.4248,
      "step": 90500
    },
    {
      "epoch": 1.29,
      "learning_rate": 2.8452766570059293e-05,
      "loss": 0.4256,
      "step": 91000
    },
    {
      "epoch": 1.3,
      "learning_rate": 2.833437517758709e-05,
      "loss": 0.4254,
      "step": 91500
    },
    {
      "epoch": 1.31,
      "learning_rate": 2.8215983785114885e-05,
      "loss": 0.4213,
      "step": 92000
    },
    {
      "epoch": 1.31,
      "learning_rate": 2.8097592392642687e-05,
      "loss": 0.4219,
      "step": 92500
    },
    {
      "epoch": 1.32,
      "learning_rate": 2.7979201000170485e-05,
      "loss": 0.4231,
      "step": 93000
    },
    {
      "epoch": 1.33,
      "learning_rate": 2.7860809607698286e-05,
      "loss": 0.4206,
      "step": 93500
    },
    {
      "epoch": 1.34,
      "learning_rate": 2.7742418215226078e-05,
      "loss": 0.4203,
      "step": 94000
    },
    {
      "epoch": 1.34,
      "learning_rate": 2.762402682275388e-05,
      "loss": 0.4174,
      "step": 94500
    },
    {
      "epoch": 1.35,
      "learning_rate": 2.7505635430281677e-05,
      "loss": 0.4183,
      "step": 95000
    },
    {
      "epoch": 1.35,
      "eval_accuracy": 0.9058014412128497,
      "eval_loss": 0.3888256549835205,
      "eval_runtime": 4483.1222,
      "eval_samples_per_second": 31.448,
      "eval_steps_per_second": 7.862,
      "step": 95000
    },
    {
      "epoch": 1.36,
      "learning_rate": 2.738724403780948e-05,
      "loss": 0.4226,
      "step": 95500
    },
    {
      "epoch": 1.36,
      "learning_rate": 2.7268852645337273e-05,
      "loss": 0.4213,
      "step": 96000
    },
    {
      "epoch": 1.37,
      "learning_rate": 2.715046125286507e-05,
      "loss": 0.4191,
      "step": 96500
    },
    {
      "epoch": 1.38,
      "learning_rate": 2.7032069860392873e-05,
      "loss": 0.4156,
      "step": 97000
    },
    {
      "epoch": 1.39,
      "learning_rate": 2.691367846792067e-05,
      "loss": 0.4179,
      "step": 97500
    },
    {
      "epoch": 1.39,
      "learning_rate": 2.6795287075448466e-05,
      "loss": 0.4158,
      "step": 98000
    },
    {
      "epoch": 1.4,
      "learning_rate": 2.6676895682976267e-05,
      "loss": 0.421,
      "step": 98500
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.6558504290504065e-05,
      "loss": 0.417,
      "step": 99000
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.644011289803186e-05,
      "loss": 0.415,
      "step": 99500
    },
    {
      "epoch": 1.42,
      "learning_rate": 2.632172150555966e-05,
      "loss": 0.4126,
      "step": 100000
    },
    {
      "epoch": 1.42,
      "eval_accuracy": 0.9066390531479637,
      "eval_loss": 0.3853781521320343,
      "eval_runtime": 4483.6399,
      "eval_samples_per_second": 31.445,
      "eval_steps_per_second": 7.861,
      "step": 100000
    },
    {
      "epoch": 1.43,
      "learning_rate": 2.620333011308746e-05,
      "loss": 0.4131,
      "step": 100500
    },
    {
      "epoch": 1.43,
      "learning_rate": 2.608493872061526e-05,
      "loss": 0.4168,
      "step": 101000
    },
    {
      "epoch": 1.44,
      "learning_rate": 2.5966547328143055e-05,
      "loss": 0.4114,
      "step": 101500
    },
    {
      "epoch": 1.45,
      "learning_rate": 2.5848155935670854e-05,
      "loss": 0.4164,
      "step": 102000
    },
    {
      "epoch": 1.46,
      "learning_rate": 2.5729764543198655e-05,
      "loss": 0.4157,
      "step": 102500
    },
    {
      "epoch": 1.46,
      "learning_rate": 2.5611373150726453e-05,
      "loss": 0.4155,
      "step": 103000
    },
    {
      "epoch": 1.47,
      "learning_rate": 2.5492981758254248e-05,
      "loss": 0.4123,
      "step": 103500
    },
    {
      "epoch": 1.48,
      "learning_rate": 2.5374590365782046e-05,
      "loss": 0.4124,
      "step": 104000
    },
    {
      "epoch": 1.48,
      "learning_rate": 2.5256198973309847e-05,
      "loss": 0.4148,
      "step": 104500
    },
    {
      "epoch": 1.49,
      "learning_rate": 2.5137807580837645e-05,
      "loss": 0.411,
      "step": 105000
    },
    {
      "epoch": 1.49,
      "eval_accuracy": 0.907043654550547,
      "eval_loss": 0.38217270374298096,
      "eval_runtime": 4482.6373,
      "eval_samples_per_second": 31.452,
      "eval_steps_per_second": 7.863,
      "step": 105000
    },
    {
      "epoch": 1.5,
      "learning_rate": 2.501941618836544e-05,
      "loss": 0.408,
      "step": 105500
    },
    {
      "epoch": 1.51,
      "learning_rate": 2.490102479589324e-05,
      "loss": 0.4123,
      "step": 106000
    },
    {
      "epoch": 1.51,
      "learning_rate": 2.4782633403421036e-05,
      "loss": 0.4125,
      "step": 106500
    },
    {
      "epoch": 1.52,
      "learning_rate": 2.4664242010948838e-05,
      "loss": 0.4117,
      "step": 107000
    },
    {
      "epoch": 1.53,
      "learning_rate": 2.4545850618476636e-05,
      "loss": 0.4083,
      "step": 107500
    },
    {
      "epoch": 1.53,
      "learning_rate": 2.4427459226004434e-05,
      "loss": 0.4117,
      "step": 108000
    },
    {
      "epoch": 1.54,
      "learning_rate": 2.4309067833532232e-05,
      "loss": 0.4079,
      "step": 108500
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.419067644106003e-05,
      "loss": 0.4092,
      "step": 109000
    },
    {
      "epoch": 1.56,
      "learning_rate": 2.4072285048587828e-05,
      "loss": 0.4097,
      "step": 109500
    },
    {
      "epoch": 1.56,
      "learning_rate": 2.395389365611563e-05,
      "loss": 0.409,
      "step": 110000
    },
    {
      "epoch": 1.56,
      "eval_accuracy": 0.907909121265083,
      "eval_loss": 0.3783527910709381,
      "eval_runtime": 4481.8573,
      "eval_samples_per_second": 31.457,
      "eval_steps_per_second": 7.864,
      "step": 110000
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.3835502263643424e-05,
      "loss": 0.403,
      "step": 110500
    },
    {
      "epoch": 1.58,
      "learning_rate": 2.3717110871171226e-05,
      "loss": 0.4109,
      "step": 111000
    },
    {
      "epoch": 1.58,
      "learning_rate": 2.359871947869902e-05,
      "loss": 0.4072,
      "step": 111500
    },
    {
      "epoch": 1.59,
      "learning_rate": 2.348032808622682e-05,
      "loss": 0.4063,
      "step": 112000
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.336193669375462e-05,
      "loss": 0.4087,
      "step": 112500
    },
    {
      "epoch": 1.61,
      "learning_rate": 2.3243545301282415e-05,
      "loss": 0.4034,
      "step": 113000
    },
    {
      "epoch": 1.61,
      "learning_rate": 2.3125153908810216e-05,
      "loss": 0.4041,
      "step": 113500
    },
    {
      "epoch": 1.62,
      "learning_rate": 2.300676251633801e-05,
      "loss": 0.4044,
      "step": 114000
    },
    {
      "epoch": 1.63,
      "learning_rate": 2.2888371123865812e-05,
      "loss": 0.4072,
      "step": 114500
    },
    {
      "epoch": 1.63,
      "learning_rate": 2.276997973139361e-05,
      "loss": 0.4041,
      "step": 115000
    },
    {
      "epoch": 1.63,
      "eval_accuracy": 0.9086945617090499,
      "eval_loss": 0.37441667914390564,
      "eval_runtime": 4484.0758,
      "eval_samples_per_second": 31.441,
      "eval_steps_per_second": 7.86,
      "step": 115000
    },
    {
      "epoch": 1.64,
      "learning_rate": 2.265158833892141e-05,
      "loss": 0.4044,
      "step": 115500
    },
    {
      "epoch": 1.65,
      "learning_rate": 2.2533196946449206e-05,
      "loss": 0.4038,
      "step": 116000
    },
    {
      "epoch": 1.66,
      "learning_rate": 2.2414805553977004e-05,
      "loss": 0.4003,
      "step": 116500
    },
    {
      "epoch": 1.66,
      "learning_rate": 2.2296414161504803e-05,
      "loss": 0.4024,
      "step": 117000
    },
    {
      "epoch": 1.67,
      "learning_rate": 2.2178022769032604e-05,
      "loss": 0.4032,
      "step": 117500
    },
    {
      "epoch": 1.68,
      "learning_rate": 2.20596313765604e-05,
      "loss": 0.3995,
      "step": 118000
    },
    {
      "epoch": 1.68,
      "learning_rate": 2.1941239984088197e-05,
      "loss": 0.4052,
      "step": 118500
    },
    {
      "epoch": 1.69,
      "learning_rate": 2.1822848591615995e-05,
      "loss": 0.4007,
      "step": 119000
    },
    {
      "epoch": 1.7,
      "learning_rate": 2.1704457199143793e-05,
      "loss": 0.4003,
      "step": 119500
    },
    {
      "epoch": 1.7,
      "learning_rate": 2.1586065806671594e-05,
      "loss": 0.4025,
      "step": 120000
    },
    {
      "epoch": 1.7,
      "eval_accuracy": 0.9094200337257503,
      "eval_loss": 0.37063175439834595,
      "eval_runtime": 4483.4685,
      "eval_samples_per_second": 31.446,
      "eval_steps_per_second": 7.862,
      "step": 120000
    },
    {
      "epoch": 1.71,
      "learning_rate": 2.146767441419939e-05,
      "loss": 0.4016,
      "step": 120500
    },
    {
      "epoch": 1.72,
      "learning_rate": 2.134928302172719e-05,
      "loss": 0.4001,
      "step": 121000
    },
    {
      "epoch": 1.73,
      "learning_rate": 2.123089162925499e-05,
      "loss": 0.4006,
      "step": 121500
    },
    {
      "epoch": 1.73,
      "learning_rate": 2.1112500236782787e-05,
      "loss": 0.3957,
      "step": 122000
    },
    {
      "epoch": 1.74,
      "learning_rate": 2.0994108844310585e-05,
      "loss": 0.4004,
      "step": 122500
    },
    {
      "epoch": 1.75,
      "learning_rate": 2.0875717451838383e-05,
      "loss": 0.3994,
      "step": 123000
    },
    {
      "epoch": 1.75,
      "learning_rate": 2.075732605936618e-05,
      "loss": 0.3951,
      "step": 123500
    },
    {
      "epoch": 1.76,
      "learning_rate": 2.063893466689398e-05,
      "loss": 0.3963,
      "step": 124000
    },
    {
      "epoch": 1.77,
      "learning_rate": 2.0520543274421777e-05,
      "loss": 0.3955,
      "step": 124500
    },
    {
      "epoch": 1.78,
      "learning_rate": 2.0402151881949575e-05,
      "loss": 0.398,
      "step": 125000
    },
    {
      "epoch": 1.78,
      "eval_accuracy": 0.9097696447890194,
      "eval_loss": 0.3691096901893616,
      "eval_runtime": 4482.3141,
      "eval_samples_per_second": 31.454,
      "eval_steps_per_second": 7.864,
      "step": 125000
    },
    {
      "epoch": 1.78,
      "learning_rate": 2.0283760489477373e-05,
      "loss": 0.402,
      "step": 125500
    },
    {
      "epoch": 1.79,
      "learning_rate": 2.016536909700517e-05,
      "loss": 0.393,
      "step": 126000
    },
    {
      "epoch": 1.8,
      "learning_rate": 2.004697770453297e-05,
      "loss": 0.3927,
      "step": 126500
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.9928586312060767e-05,
      "loss": 0.3931,
      "step": 127000
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.981019491958857e-05,
      "loss": 0.3922,
      "step": 127500
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.9691803527116364e-05,
      "loss": 0.3984,
      "step": 128000
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.9573412134644165e-05,
      "loss": 0.3958,
      "step": 128500
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.9455020742171963e-05,
      "loss": 0.3958,
      "step": 129000
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.933662934969976e-05,
      "loss": 0.3971,
      "step": 129500
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.921823795722756e-05,
      "loss": 0.3894,
      "step": 130000
    },
    {
      "epoch": 1.85,
      "eval_accuracy": 0.9105847302510419,
      "eval_loss": 0.3647039532661438,
      "eval_runtime": 4483.1278,
      "eval_samples_per_second": 31.448,
      "eval_steps_per_second": 7.862,
      "step": 130000
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.9099846564755357e-05,
      "loss": 0.3912,
      "step": 130500
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.8981455172283155e-05,
      "loss": 0.3895,
      "step": 131000
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.8863063779810953e-05,
      "loss": 0.3906,
      "step": 131500
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.874467238733875e-05,
      "loss": 0.3893,
      "step": 132000
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.862628099486655e-05,
      "loss": 0.3941,
      "step": 132500
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.8507889602394348e-05,
      "loss": 0.394,
      "step": 133000
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.8389498209922146e-05,
      "loss": 0.3893,
      "step": 133500
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.8271106817449947e-05,
      "loss": 0.3911,
      "step": 134000
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.8152715424977742e-05,
      "loss": 0.3899,
      "step": 134500
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.8034324032505543e-05,
      "loss": 0.3898,
      "step": 135000
    },
    {
      "epoch": 1.92,
      "eval_accuracy": 0.9108666750444402,
      "eval_loss": 0.36286085844039917,
      "eval_runtime": 4484.3794,
      "eval_samples_per_second": 31.439,
      "eval_steps_per_second": 7.86,
      "step": 135000
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.7915932640033338e-05,
      "loss": 0.389,
      "step": 135500
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.779754124756114e-05,
      "loss": 0.3878,
      "step": 136000
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.7679149855088938e-05,
      "loss": 0.3926,
      "step": 136500
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.7560758462616736e-05,
      "loss": 0.3916,
      "step": 137000
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.7442367070144534e-05,
      "loss": 0.3903,
      "step": 137500
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.732397567767233e-05,
      "loss": 0.3932,
      "step": 138000
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.720558428520013e-05,
      "loss": 0.3894,
      "step": 138500
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.7087192892727928e-05,
      "loss": 0.3882,
      "step": 139000
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.6968801500255726e-05,
      "loss": 0.3874,
      "step": 139500
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.6850410107783524e-05,
      "loss": 0.3857,
      "step": 140000
    },
    {
      "epoch": 1.99,
      "eval_accuracy": 0.9115911874403287,
      "eval_loss": 0.36069589853286743,
      "eval_runtime": 4496.781,
      "eval_samples_per_second": 31.353,
      "eval_steps_per_second": 7.838,
      "step": 140000
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.6732018715311322e-05,
      "loss": 0.3895,
      "step": 140500
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.661362732283912e-05,
      "loss": 0.3903,
      "step": 141000
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.6495235930366922e-05,
      "loss": 0.3858,
      "step": 141500
    },
    {
      "epoch": 2.02,
      "learning_rate": 1.6376844537894716e-05,
      "loss": 0.3838,
      "step": 142000
    },
    {
      "epoch": 2.02,
      "learning_rate": 1.6258453145422518e-05,
      "loss": 0.3811,
      "step": 142500
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.6140061752950313e-05,
      "loss": 0.3892,
      "step": 143000
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.6021670360478114e-05,
      "loss": 0.3857,
      "step": 143500
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.5903278968005912e-05,
      "loss": 0.3812,
      "step": 144000
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.5784887575533707e-05,
      "loss": 0.3862,
      "step": 144500
    },
    {
      "epoch": 2.06,
      "learning_rate": 1.5666496183061508e-05,
      "loss": 0.3809,
      "step": 145000
    },
    {
      "epoch": 2.06,
      "eval_accuracy": 0.9121499748273744,
      "eval_loss": 0.3579576909542084,
      "eval_runtime": 4496.0613,
      "eval_samples_per_second": 31.358,
      "eval_steps_per_second": 7.84,
      "step": 145000
    },
    {
      "epoch": 2.07,
      "learning_rate": 1.5548104790589303e-05,
      "loss": 0.3831,
      "step": 145500
    },
    {
      "epoch": 2.07,
      "learning_rate": 1.5429713398117104e-05,
      "loss": 0.3831,
      "step": 146000
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.5311322005644902e-05,
      "loss": 0.3822,
      "step": 146500
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.51929306131727e-05,
      "loss": 0.3817,
      "step": 147000
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.5074539220700499e-05,
      "loss": 0.3854,
      "step": 147500
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.4956147828228298e-05,
      "loss": 0.3851,
      "step": 148000
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.4837756435756095e-05,
      "loss": 0.3838,
      "step": 148500
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.4719365043283895e-05,
      "loss": 0.3812,
      "step": 149000
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.4600973650811693e-05,
      "loss": 0.3825,
      "step": 149500
    },
    {
      "epoch": 2.13,
      "learning_rate": 1.4482582258339492e-05,
      "loss": 0.3819,
      "step": 150000
    },
    {
      "epoch": 2.13,
      "eval_accuracy": 0.91259126809248,
      "eval_loss": 0.3549927771091461,
      "eval_runtime": 4493.1359,
      "eval_samples_per_second": 31.378,
      "eval_steps_per_second": 7.845,
      "step": 150000
    },
    {
      "epoch": 2.14,
      "learning_rate": 1.4364190865867289e-05,
      "loss": 0.3793,
      "step": 150500
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.4245799473395085e-05,
      "loss": 0.3832,
      "step": 151000
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.4127408080922885e-05,
      "loss": 0.3806,
      "step": 151500
    },
    {
      "epoch": 2.16,
      "learning_rate": 1.4009016688450683e-05,
      "loss": 0.3784,
      "step": 152000
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3890625295978483e-05,
      "loss": 0.382,
      "step": 152500
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3772233903506279e-05,
      "loss": 0.3788,
      "step": 153000
    },
    {
      "epoch": 2.18,
      "learning_rate": 1.3653842511034079e-05,
      "loss": 0.3837,
      "step": 153500
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.3535451118561875e-05,
      "loss": 0.3843,
      "step": 154000
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.3417059726089675e-05,
      "loss": 0.3813,
      "step": 154500
    },
    {
      "epoch": 2.2,
      "learning_rate": 1.3298668333617473e-05,
      "loss": 0.3795,
      "step": 155000
    },
    {
      "epoch": 2.2,
      "eval_accuracy": 0.913045165928687,
      "eval_loss": 0.3533931076526642,
      "eval_runtime": 4479.8171,
      "eval_samples_per_second": 31.471,
      "eval_steps_per_second": 7.868,
      "step": 155000
    },
    {
      "epoch": 2.21,
      "learning_rate": 1.3180276941145273e-05,
      "loss": 0.3806,
      "step": 155500
    },
    {
      "epoch": 2.22,
      "learning_rate": 1.306188554867307e-05,
      "loss": 0.3763,
      "step": 156000
    },
    {
      "epoch": 2.22,
      "learning_rate": 1.2943494156200869e-05,
      "loss": 0.371,
      "step": 156500
    },
    {
      "epoch": 2.23,
      "learning_rate": 1.2825102763728667e-05,
      "loss": 0.3809,
      "step": 157000
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.2706711371256467e-05,
      "loss": 0.3775,
      "step": 157500
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.2588319978784263e-05,
      "loss": 0.3813,
      "step": 158000
    },
    {
      "epoch": 2.25,
      "learning_rate": 1.2469928586312061e-05,
      "loss": 0.3808,
      "step": 158500
    },
    {
      "epoch": 2.26,
      "learning_rate": 1.235153719383986e-05,
      "loss": 0.3771,
      "step": 159000
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.2233145801367659e-05,
      "loss": 0.3734,
      "step": 159500
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.2114754408895457e-05,
      "loss": 0.3772,
      "step": 160000
    },
    {
      "epoch": 2.27,
      "eval_accuracy": 0.9134737242143287,
      "eval_loss": 0.3512367606163025,
      "eval_runtime": 4479.3521,
      "eval_samples_per_second": 31.475,
      "eval_steps_per_second": 7.869,
      "step": 160000
    },
    {
      "epoch": 2.28,
      "learning_rate": 1.1996363016423255e-05,
      "loss": 0.3765,
      "step": 160500
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.1877971623951052e-05,
      "loss": 0.3771,
      "step": 161000
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.175958023147885e-05,
      "loss": 0.3757,
      "step": 161500
    },
    {
      "epoch": 2.3,
      "learning_rate": 1.164118883900665e-05,
      "loss": 0.3742,
      "step": 162000
    },
    {
      "epoch": 2.31,
      "learning_rate": 1.1522797446534448e-05,
      "loss": 0.3738,
      "step": 162500
    },
    {
      "epoch": 2.32,
      "learning_rate": 1.1404406054062246e-05,
      "loss": 0.3743,
      "step": 163000
    },
    {
      "epoch": 2.32,
      "learning_rate": 1.1286014661590044e-05,
      "loss": 0.3738,
      "step": 163500
    },
    {
      "epoch": 2.33,
      "learning_rate": 1.1167623269117842e-05,
      "loss": 0.3712,
      "step": 164000
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.1049231876645642e-05,
      "loss": 0.373,
      "step": 164500
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.093084048417344e-05,
      "loss": 0.3723,
      "step": 165000
    },
    {
      "epoch": 2.34,
      "eval_accuracy": 0.9138787356297591,
      "eval_loss": 0.3488907516002655,
      "eval_runtime": 4481.5458,
      "eval_samples_per_second": 31.459,
      "eval_steps_per_second": 7.865,
      "step": 165000
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.0812449091701238e-05,
      "loss": 0.373,
      "step": 165500
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.0694057699229036e-05,
      "loss": 0.3715,
      "step": 166000
    },
    {
      "epoch": 2.37,
      "learning_rate": 1.0575666306756834e-05,
      "loss": 0.3753,
      "step": 166500
    },
    {
      "epoch": 2.37,
      "learning_rate": 1.0457274914284634e-05,
      "loss": 0.3755,
      "step": 167000
    },
    {
      "epoch": 2.38,
      "learning_rate": 1.033888352181243e-05,
      "loss": 0.3736,
      "step": 167500
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.0220492129340228e-05,
      "loss": 0.3738,
      "step": 168000
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.0102100736868026e-05,
      "loss": 0.3718,
      "step": 168500
    },
    {
      "epoch": 2.4,
      "learning_rate": 9.983709344395826e-06,
      "loss": 0.3701,
      "step": 169000
    },
    {
      "epoch": 2.41,
      "learning_rate": 9.865317951923624e-06,
      "loss": 0.3747,
      "step": 169500
    },
    {
      "epoch": 2.42,
      "learning_rate": 9.746926559451422e-06,
      "loss": 0.3741,
      "step": 170000
    },
    {
      "epoch": 2.42,
      "eval_accuracy": 0.9141958812185558,
      "eval_loss": 0.34678730368614197,
      "eval_runtime": 4483.3675,
      "eval_samples_per_second": 31.446,
      "eval_steps_per_second": 7.862,
      "step": 170000
    },
    {
      "epoch": 2.42,
      "learning_rate": 9.62853516697922e-06,
      "loss": 0.3731,
      "step": 170500
    },
    {
      "epoch": 2.43,
      "learning_rate": 9.510143774507018e-06,
      "loss": 0.3709,
      "step": 171000
    },
    {
      "epoch": 2.44,
      "learning_rate": 9.391752382034818e-06,
      "loss": 0.3674,
      "step": 171500
    },
    {
      "epoch": 2.44,
      "learning_rate": 9.273360989562616e-06,
      "loss": 0.3688,
      "step": 172000
    },
    {
      "epoch": 2.45,
      "learning_rate": 9.154969597090414e-06,
      "loss": 0.3708,
      "step": 172500
    },
    {
      "epoch": 2.46,
      "learning_rate": 9.036578204618212e-06,
      "loss": 0.3705,
      "step": 173000
    },
    {
      "epoch": 2.46,
      "learning_rate": 8.91818681214601e-06,
      "loss": 0.3688,
      "step": 173500
    },
    {
      "epoch": 2.47,
      "learning_rate": 8.799795419673808e-06,
      "loss": 0.3688,
      "step": 174000
    },
    {
      "epoch": 2.48,
      "learning_rate": 8.681404027201606e-06,
      "loss": 0.3731,
      "step": 174500
    },
    {
      "epoch": 2.49,
      "learning_rate": 8.563012634729405e-06,
      "loss": 0.3678,
      "step": 175000
    },
    {
      "epoch": 2.49,
      "eval_accuracy": 0.9147443492541378,
      "eval_loss": 0.3443525731563568,
      "eval_runtime": 4487.0096,
      "eval_samples_per_second": 31.421,
      "eval_steps_per_second": 7.855,
      "step": 175000
    },
    {
      "epoch": 2.49,
      "learning_rate": 8.444621242257203e-06,
      "loss": 0.3712,
      "step": 175500
    },
    {
      "epoch": 2.5,
      "learning_rate": 8.326229849785e-06,
      "loss": 0.3707,
      "step": 176000
    },
    {
      "epoch": 2.51,
      "learning_rate": 8.2078384573128e-06,
      "loss": 0.3693,
      "step": 176500
    },
    {
      "epoch": 2.51,
      "learning_rate": 8.089447064840599e-06,
      "loss": 0.371,
      "step": 177000
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.971055672368397e-06,
      "loss": 0.371,
      "step": 177500
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.852664279896195e-06,
      "loss": 0.3724,
      "step": 178000
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.734272887423993e-06,
      "loss": 0.3707,
      "step": 178500
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.615881494951792e-06,
      "loss": 0.3673,
      "step": 179000
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.4974901024795906e-06,
      "loss": 0.3684,
      "step": 179500
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.379098710007389e-06,
      "loss": 0.3698,
      "step": 180000
    },
    {
      "epoch": 2.56,
      "eval_accuracy": 0.9149897136925411,
      "eval_loss": 0.3437068462371826,
      "eval_runtime": 4486.3551,
      "eval_samples_per_second": 31.426,
      "eval_steps_per_second": 7.856,
      "step": 180000
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.260707317535187e-06,
      "loss": 0.3644,
      "step": 180500
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.142315925062984e-06,
      "loss": 0.3696,
      "step": 181000
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.023924532590782e-06,
      "loss": 0.3682,
      "step": 181500
    },
    {
      "epoch": 2.59,
      "learning_rate": 6.905533140118581e-06,
      "loss": 0.3703,
      "step": 182000
    },
    {
      "epoch": 2.59,
      "learning_rate": 6.787141747646379e-06,
      "loss": 0.3686,
      "step": 182500
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.668750355174178e-06,
      "loss": 0.3638,
      "step": 183000
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.550358962701976e-06,
      "loss": 0.3652,
      "step": 183500
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.431967570229774e-06,
      "loss": 0.367,
      "step": 184000
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.313576177757573e-06,
      "loss": 0.3665,
      "step": 184500
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.195184785285371e-06,
      "loss": 0.3685,
      "step": 185000
    },
    {
      "epoch": 2.63,
      "eval_accuracy": 0.9152310790161654,
      "eval_loss": 0.34259435534477234,
      "eval_runtime": 4485.6065,
      "eval_samples_per_second": 31.431,
      "eval_steps_per_second": 7.858,
      "step": 185000
    },
    {
      "epoch": 2.64,
      "learning_rate": 6.076793392813169e-06,
      "loss": 0.3635,
      "step": 185500
    },
    {
      "epoch": 2.64,
      "learning_rate": 5.958402000340967e-06,
      "loss": 0.3645,
      "step": 186000
    },
    {
      "epoch": 2.65,
      "learning_rate": 5.840010607868765e-06,
      "loss": 0.3697,
      "step": 186500
    },
    {
      "epoch": 2.66,
      "learning_rate": 5.721619215396564e-06,
      "loss": 0.3693,
      "step": 187000
    },
    {
      "epoch": 2.66,
      "learning_rate": 5.603227822924362e-06,
      "loss": 0.3638,
      "step": 187500
    },
    {
      "epoch": 2.67,
      "learning_rate": 5.484836430452161e-06,
      "loss": 0.3621,
      "step": 188000
    },
    {
      "epoch": 2.68,
      "learning_rate": 5.366445037979959e-06,
      "loss": 0.3602,
      "step": 188500
    },
    {
      "epoch": 2.69,
      "learning_rate": 5.248053645507757e-06,
      "loss": 0.369,
      "step": 189000
    },
    {
      "epoch": 2.69,
      "learning_rate": 5.1296622530355554e-06,
      "loss": 0.3678,
      "step": 189500
    },
    {
      "epoch": 2.7,
      "learning_rate": 5.0112708605633535e-06,
      "loss": 0.3663,
      "step": 190000
    },
    {
      "epoch": 2.7,
      "eval_accuracy": 0.9157103027094395,
      "eval_loss": 0.3402460515499115,
      "eval_runtime": 4483.4518,
      "eval_samples_per_second": 31.446,
      "eval_steps_per_second": 7.862,
      "step": 190000
    },
    {
      "epoch": 2.71,
      "learning_rate": 4.8928794680911524e-06,
      "loss": 0.366,
      "step": 190500
    },
    {
      "epoch": 2.71,
      "learning_rate": 4.7744880756189505e-06,
      "loss": 0.3676,
      "step": 191000
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.656096683146749e-06,
      "loss": 0.3653,
      "step": 191500
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.5377052906745475e-06,
      "loss": 0.3649,
      "step": 192000
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.419313898202345e-06,
      "loss": 0.3626,
      "step": 192500
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.300922505730144e-06,
      "loss": 0.3603,
      "step": 193000
    },
    {
      "epoch": 2.75,
      "learning_rate": 4.182531113257942e-06,
      "loss": 0.3597,
      "step": 193500
    },
    {
      "epoch": 2.76,
      "learning_rate": 4.064139720785741e-06,
      "loss": 0.3666,
      "step": 194000
    },
    {
      "epoch": 2.76,
      "learning_rate": 3.945748328313539e-06,
      "loss": 0.3626,
      "step": 194500
    },
    {
      "epoch": 2.77,
      "learning_rate": 3.827356935841337e-06,
      "loss": 0.3636,
      "step": 195000
    },
    {
      "epoch": 2.77,
      "eval_accuracy": 0.9157984246825175,
      "eval_loss": 0.33979707956314087,
      "eval_runtime": 4482.6546,
      "eval_samples_per_second": 31.451,
      "eval_steps_per_second": 7.863,
      "step": 195000
    },
    {
      "epoch": 2.78,
      "learning_rate": 3.708965543369135e-06,
      "loss": 0.3641,
      "step": 195500
    },
    {
      "epoch": 2.78,
      "learning_rate": 3.5905741508969334e-06,
      "loss": 0.3637,
      "step": 196000
    },
    {
      "epoch": 2.79,
      "learning_rate": 3.4721827584247314e-06,
      "loss": 0.3646,
      "step": 196500
    },
    {
      "epoch": 2.8,
      "learning_rate": 3.35379136595253e-06,
      "loss": 0.3603,
      "step": 197000
    },
    {
      "epoch": 2.81,
      "learning_rate": 3.2353999734803284e-06,
      "loss": 0.3663,
      "step": 197500
    },
    {
      "epoch": 2.81,
      "learning_rate": 3.1170085810081265e-06,
      "loss": 0.3667,
      "step": 198000
    },
    {
      "epoch": 2.82,
      "learning_rate": 2.998617188535925e-06,
      "loss": 0.3638,
      "step": 198500
    },
    {
      "epoch": 2.83,
      "learning_rate": 2.880225796063723e-06,
      "loss": 0.3597,
      "step": 199000
    },
    {
      "epoch": 2.83,
      "learning_rate": 2.761834403591521e-06,
      "loss": 0.3639,
      "step": 199500
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.6434430111193197e-06,
      "loss": 0.3618,
      "step": 200000
    },
    {
      "epoch": 2.84,
      "eval_accuracy": 0.9161026342896044,
      "eval_loss": 0.33811721205711365,
      "eval_runtime": 4483.081,
      "eval_samples_per_second": 31.448,
      "eval_steps_per_second": 7.862,
      "step": 200000
    },
    {
      "epoch": 2.85,
      "learning_rate": 2.525051618647118e-06,
      "loss": 0.3648,
      "step": 200500
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.4066602261749166e-06,
      "loss": 0.3614,
      "step": 201000
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.2882688337027147e-06,
      "loss": 0.3607,
      "step": 201500
    },
    {
      "epoch": 2.87,
      "learning_rate": 2.169877441230513e-06,
      "loss": 0.3638,
      "step": 202000
    },
    {
      "epoch": 2.88,
      "learning_rate": 2.0514860487583113e-06,
      "loss": 0.3634,
      "step": 202500
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.9330946562861094e-06,
      "loss": 0.3628,
      "step": 203000
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.8147032638139079e-06,
      "loss": 0.358,
      "step": 203500
    },
    {
      "epoch": 2.9,
      "learning_rate": 1.6963118713417062e-06,
      "loss": 0.3586,
      "step": 204000
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.5779204788695042e-06,
      "loss": 0.364,
      "step": 204500
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.4595290863973025e-06,
      "loss": 0.3613,
      "step": 205000
    },
    {
      "epoch": 2.91,
      "eval_accuracy": 0.9164505308425883,
      "eval_loss": 0.3365662097930908,
      "eval_runtime": 4483.3011,
      "eval_samples_per_second": 31.447,
      "eval_steps_per_second": 7.862,
      "step": 205000
    },
    {
      "epoch": 2.92,
      "learning_rate": 1.341137693925101e-06,
      "loss": 0.3641,
      "step": 205500
    },
    {
      "epoch": 2.93,
      "learning_rate": 1.2227463014528993e-06,
      "loss": 0.363,
      "step": 206000
    },
    {
      "epoch": 2.93,
      "learning_rate": 1.1043549089806974e-06,
      "loss": 0.3576,
      "step": 206500
    },
    {
      "epoch": 2.94,
      "learning_rate": 9.859635165084959e-07,
      "loss": 0.3627,
      "step": 207000
    },
    {
      "epoch": 2.95,
      "learning_rate": 8.67572124036294e-07,
      "loss": 0.3604,
      "step": 207500
    },
    {
      "epoch": 2.96,
      "learning_rate": 7.491807315640924e-07,
      "loss": 0.3614,
      "step": 208000
    },
    {
      "epoch": 2.96,
      "learning_rate": 6.307893390918907e-07,
      "loss": 0.3584,
      "step": 208500
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.12397946619689e-07,
      "loss": 0.3594,
      "step": 209000
    },
    {
      "epoch": 2.98,
      "learning_rate": 3.9400655414748724e-07,
      "loss": 0.3589,
      "step": 209500
    },
    {
      "epoch": 2.98,
      "learning_rate": 2.756151616752856e-07,
      "loss": 0.3618,
      "step": 210000
    },
    {
      "epoch": 2.98,
      "eval_accuracy": 0.9162037471165208,
      "eval_loss": 0.3379279673099518,
      "eval_runtime": 4487.2562,
      "eval_samples_per_second": 31.419,
      "eval_steps_per_second": 7.855,
      "step": 210000
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.5722376920308387e-07,
      "loss": 0.3621,
      "step": 210500
    },
    {
      "epoch": 3.0,
      "learning_rate": 3.883237673088216e-08,
      "loss": 0.3606,
      "step": 211000
    },
    {
      "epoch": 3.0,
      "step": 211164,
      "total_flos": 8.915475817785139e+17,
      "train_loss": 0.4472081179925129,
      "train_runtime": 493604.9481,
      "train_samples_per_second": 6.845,
      "train_steps_per_second": 0.428
    }
  ],
  "max_steps": 211164,
  "num_train_epochs": 3,
  "total_flos": 8.915475817785139e+17,
  "trial_name": null,
  "trial_params": null
}