{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.261958573126121,
  "global_step": 140000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 5.999999999999999e-06,
      "loss": 0.9072,
      "step": 500
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.1999999999999999e-05,
      "loss": 0.6779,
      "step": 1000
    },
    {
      "epoch": 0.02,
      "eval_loss": 0.6512336730957031,
      "eval_runtime": 1.3899,
      "eval_samples_per_second": 1581.436,
      "eval_steps_per_second": 25.182,
      "step": 1000
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.7999999999999997e-05,
      "loss": 0.6516,
      "step": 1500
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.3999999999999997e-05,
      "loss": 0.6505,
      "step": 2000
    },
    {
      "epoch": 0.05,
      "eval_loss": 0.6479398608207703,
      "eval_runtime": 1.2819,
      "eval_samples_per_second": 1714.632,
      "eval_steps_per_second": 27.303,
      "step": 2000
    },
    {
      "epoch": 0.06,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 0.6487,
      "step": 2500
    },
    {
      "epoch": 0.07,
      "learning_rate": 3.5999999999999994e-05,
      "loss": 0.6475,
      "step": 3000
    },
    {
      "epoch": 0.07,
      "eval_loss": 0.6461147665977478,
      "eval_runtime": 1.2932,
      "eval_samples_per_second": 1699.633,
      "eval_steps_per_second": 27.064,
      "step": 3000
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.2e-05,
      "loss": 0.6469,
      "step": 3500
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.7999999999999994e-05,
      "loss": 0.6464,
      "step": 4000
    },
    {
      "epoch": 0.09,
      "eval_loss": 0.6451290249824524,
      "eval_runtime": 1.2355,
      "eval_samples_per_second": 1779.085,
      "eval_steps_per_second": 28.329,
      "step": 4000
    },
    {
      "epoch": 0.1,
      "learning_rate": 5.399999999999999e-05,
      "loss": 0.6459,
      "step": 4500
    },
    {
      "epoch": 0.12,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 0.6455,
      "step": 5000
    },
    {
      "epoch": 0.12,
      "eval_loss": 0.6444417238235474,
      "eval_runtime": 1.2972,
      "eval_samples_per_second": 1694.435,
      "eval_steps_per_second": 26.981,
      "step": 5000
    },
    {
      "epoch": 0.13,
      "learning_rate": 6.599999999999999e-05,
      "loss": 0.6452,
      "step": 5500
    },
    {
      "epoch": 0.14,
      "learning_rate": 7.199999999999999e-05,
      "loss": 0.6448,
      "step": 6000
    },
    {
      "epoch": 0.14,
      "eval_loss": 0.6438854336738586,
      "eval_runtime": 1.2871,
      "eval_samples_per_second": 1707.767,
      "eval_steps_per_second": 27.194,
      "step": 6000
    },
    {
      "epoch": 0.15,
      "learning_rate": 7.8e-05,
      "loss": 0.64,
      "step": 6500
    },
    {
      "epoch": 0.16,
      "learning_rate": 8.4e-05,
      "loss": 0.6222,
      "step": 7000
    },
    {
      "epoch": 0.16,
      "eval_loss": 0.6123253703117371,
      "eval_runtime": 1.2501,
      "eval_samples_per_second": 1758.21,
      "eval_steps_per_second": 27.997,
      "step": 7000
    },
    {
      "epoch": 0.17,
      "learning_rate": 8.999999999999999e-05,
      "loss": 0.6115,
      "step": 7500
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.599999999999999e-05,
      "loss": 0.6019,
      "step": 8000
    },
    {
      "epoch": 0.19,
      "eval_loss": 0.5872684717178345,
      "eval_runtime": 1.2778,
      "eval_samples_per_second": 1720.181,
      "eval_steps_per_second": 27.391,
      "step": 8000
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.000102,
      "loss": 0.5924,
      "step": 8500
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00010799999999999998,
      "loss": 0.5838,
      "step": 9000
    },
    {
      "epoch": 0.21,
      "eval_loss": 0.5689075589179993,
      "eval_runtime": 1.2743,
      "eval_samples_per_second": 1724.838,
      "eval_steps_per_second": 27.466,
      "step": 9000
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00011399999999999999,
      "loss": 0.5737,
      "step": 9500
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00011999999999999999,
      "loss": 0.5612,
      "step": 10000
    },
    {
      "epoch": 0.23,
      "eval_loss": 0.5473007559776306,
      "eval_runtime": 1.2648,
      "eval_samples_per_second": 1737.882,
      "eval_steps_per_second": 27.673,
      "step": 10000
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00012599999999999997,
      "loss": 0.5516,
      "step": 10500
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00013199999999999998,
      "loss": 0.5428,
      "step": 11000
    },
    {
      "epoch": 0.26,
      "eval_loss": 0.5300613045692444,
      "eval_runtime": 1.2658,
      "eval_samples_per_second": 1736.485,
      "eval_steps_per_second": 27.651,
      "step": 11000
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.000138,
      "loss": 0.5346,
      "step": 11500
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00014399999999999998,
      "loss": 0.527,
      "step": 12000
    },
    {
      "epoch": 0.28,
      "eval_loss": 0.5140743255615234,
      "eval_runtime": 1.2609,
      "eval_samples_per_second": 1743.192,
      "eval_steps_per_second": 27.758,
      "step": 12000
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00015,
      "loss": 0.5203,
      "step": 12500
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.000156,
      "loss": 0.5142,
      "step": 13000
    },
    {
      "epoch": 0.3,
      "eval_loss": 0.5022566318511963,
      "eval_runtime": 1.2481,
      "eval_samples_per_second": 1761.066,
      "eval_steps_per_second": 28.042,
      "step": 13000
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.000162,
      "loss": 0.5079,
      "step": 13500
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.000168,
      "loss": 0.5002,
      "step": 14000
    },
    {
      "epoch": 0.33,
      "eval_loss": 0.4841987192630768,
      "eval_runtime": 1.3176,
      "eval_samples_per_second": 1668.178,
      "eval_steps_per_second": 26.563,
      "step": 14000
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00017399999999999997,
      "loss": 0.4916,
      "step": 14500
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00017999999999999998,
      "loss": 0.4852,
      "step": 15000
    },
    {
      "epoch": 0.35,
      "eval_loss": 0.47066375613212585,
      "eval_runtime": 1.3026,
      "eval_samples_per_second": 1687.422,
      "eval_steps_per_second": 26.87,
      "step": 15000
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.000186,
      "loss": 0.4778,
      "step": 15500
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00019199999999999998,
      "loss": 0.4697,
      "step": 16000
    },
    {
      "epoch": 0.37,
      "eval_loss": 0.4562186300754547,
      "eval_runtime": 1.2895,
      "eval_samples_per_second": 1704.503,
      "eval_steps_per_second": 27.142,
      "step": 16000
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.000198,
      "loss": 0.4636,
      "step": 16500
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.000204,
      "loss": 0.459,
      "step": 17000
    },
    {
      "epoch": 0.4,
      "eval_loss": 0.44295796751976013,
      "eval_runtime": 1.2516,
      "eval_samples_per_second": 1756.093,
      "eval_steps_per_second": 27.963,
      "step": 17000
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.00020999999999999998,
      "loss": 0.4544,
      "step": 17500
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00021599999999999996,
      "loss": 0.45,
      "step": 18000
    },
    {
      "epoch": 0.42,
      "eval_loss": 0.4343169629573822,
      "eval_runtime": 1.327,
      "eval_samples_per_second": 1656.351,
      "eval_steps_per_second": 26.375,
      "step": 18000
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.00022199999999999998,
      "loss": 0.4458,
      "step": 18500
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00022799999999999999,
      "loss": 0.4411,
      "step": 19000
    },
    {
      "epoch": 0.44,
      "eval_loss": 0.42496559023857117,
      "eval_runtime": 1.3285,
      "eval_samples_per_second": 1654.492,
      "eval_steps_per_second": 26.345,
      "step": 19000
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.000234,
      "loss": 0.4368,
      "step": 19500
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.00023999999999999998,
      "loss": 0.4322,
      "step": 20000
    },
    {
      "epoch": 0.47,
      "eval_loss": 0.4132905900478363,
      "eval_runtime": 1.3158,
      "eval_samples_per_second": 1670.425,
      "eval_steps_per_second": 26.599,
      "step": 20000
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00024599999999999996,
      "loss": 0.4278,
      "step": 20500
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.00025199999999999995,
      "loss": 0.4233,
      "step": 21000
    },
    {
      "epoch": 0.49,
      "eval_loss": 0.4032529294490814,
      "eval_runtime": 1.2842,
      "eval_samples_per_second": 1711.577,
      "eval_steps_per_second": 27.254,
      "step": 21000
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.000258,
      "loss": 0.4189,
      "step": 21500
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00026399999999999997,
      "loss": 0.4146,
      "step": 22000
    },
    {
      "epoch": 0.51,
      "eval_loss": 0.39436739683151245,
      "eval_runtime": 1.2633,
      "eval_samples_per_second": 1739.901,
      "eval_steps_per_second": 27.705,
      "step": 22000
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.00027,
      "loss": 0.4109,
      "step": 22500
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.000276,
      "loss": 0.4077,
      "step": 23000
    },
    {
      "epoch": 0.54,
      "eval_loss": 0.38647642731666565,
      "eval_runtime": 1.2503,
      "eval_samples_per_second": 1758.021,
      "eval_steps_per_second": 27.994,
      "step": 23000
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.00028199999999999997,
      "loss": 0.4047,
      "step": 23500
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00028799999999999995,
      "loss": 0.4019,
      "step": 24000
    },
    {
      "epoch": 0.56,
      "eval_loss": 0.381785124540329,
      "eval_runtime": 1.2485,
      "eval_samples_per_second": 1760.518,
      "eval_steps_per_second": 28.034,
      "step": 24000
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.000294,
      "loss": 0.3994,
      "step": 24500
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0003,
      "loss": 0.3968,
      "step": 25000
    },
    {
      "epoch": 0.58,
      "eval_loss": 0.3785216510295868,
      "eval_runtime": 1.2913,
      "eval_samples_per_second": 1702.153,
      "eval_steps_per_second": 27.104,
      "step": 25000
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00029999920715161553,
      "loss": 0.3945,
      "step": 25500
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002999968286151326,
      "loss": 0.392,
      "step": 26000
    },
    {
      "epoch": 0.61,
      "eval_loss": 0.37318962812423706,
      "eval_runtime": 1.2758,
      "eval_samples_per_second": 1722.782,
      "eval_steps_per_second": 27.433,
      "step": 26000
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002999928644165624,
      "loss": 0.3897,
      "step": 26500
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002999873145992569,
      "loss": 0.3869,
      "step": 27000
    },
    {
      "epoch": 0.63,
      "eval_loss": 0.3669659197330475,
      "eval_runtime": 1.2869,
      "eval_samples_per_second": 1707.967,
      "eval_steps_per_second": 27.197,
      "step": 27000
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.000299980179223908,
      "loss": 0.385,
      "step": 27500
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002999714583685469,
      "loss": 0.3828,
      "step": 28000
    },
    {
      "epoch": 0.65,
      "eval_loss": 0.3615981340408325,
      "eval_runtime": 1.2961,
      "eval_samples_per_second": 1695.914,
      "eval_steps_per_second": 27.005,
      "step": 28000
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00029996115212854366,
      "loss": 0.3808,
      "step": 28500
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00029994926061660554,
      "loss": 0.3786,
      "step": 29000
    },
    {
      "epoch": 0.68,
      "eval_loss": 0.3613170087337494,
      "eval_runtime": 1.3013,
      "eval_samples_per_second": 1689.114,
      "eval_steps_per_second": 26.897,
      "step": 29000
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002999357839627762,
      "loss": 0.377,
      "step": 29500
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00029992072231443425,
      "loss": 0.3751,
      "step": 30000
    },
    {
      "epoch": 0.7,
      "eval_loss": 0.35551005601882935,
      "eval_runtime": 1.2661,
      "eval_samples_per_second": 1736.053,
      "eval_steps_per_second": 27.644,
      "step": 30000
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002999040758362914,
      "loss": 0.3731,
      "step": 30500
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00029988584471039094,
      "loss": 0.3713,
      "step": 31000
    },
    {
      "epoch": 0.72,
      "eval_loss": 0.35287660360336304,
      "eval_runtime": 1.3423,
      "eval_samples_per_second": 1637.545,
      "eval_steps_per_second": 26.076,
      "step": 31000
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002998660291361054,
      "loss": 0.3699,
      "step": 31500
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002998446293301349,
      "loss": 0.3681,
      "step": 32000
    },
    {
      "epoch": 0.75,
      "eval_loss": 0.35174888372421265,
      "eval_runtime": 1.3006,
      "eval_samples_per_second": 1690.029,
      "eval_steps_per_second": 26.911,
      "step": 32000
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002998216455265042,
      "loss": 0.3667,
      "step": 32500
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00029979707797656046,
      "loss": 0.3656,
      "step": 33000
    },
    {
      "epoch": 0.77,
      "eval_loss": 0.34766125679016113,
      "eval_runtime": 1.3363,
      "eval_samples_per_second": 1644.872,
      "eval_steps_per_second": 26.192,
      "step": 33000
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00029977092694897053,
      "loss": 0.3641,
      "step": 33500
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002997431927297178,
      "loss": 0.3626,
      "step": 34000
    },
    {
      "epoch": 0.79,
      "eval_loss": 0.3423311412334442,
      "eval_runtime": 1.3208,
      "eval_samples_per_second": 1664.116,
      "eval_steps_per_second": 26.499,
      "step": 34000
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.00029971387562209936,
      "loss": 0.3616,
      "step": 34500
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00029968297594672226,
      "loss": 0.3601,
      "step": 35000
    },
    {
      "epoch": 0.82,
      "eval_loss": 0.34427037835121155,
      "eval_runtime": 1.2929,
      "eval_samples_per_second": 1700.099,
      "eval_steps_per_second": 27.072,
      "step": 35000
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002996504940415005,
      "loss": 0.3589,
      "step": 35500
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00029961643026165096,
      "loss": 0.3577,
      "step": 36000
    },
    {
      "epoch": 0.84,
      "eval_loss": 0.34002143144607544,
      "eval_runtime": 1.2974,
      "eval_samples_per_second": 1694.186,
      "eval_steps_per_second": 26.977,
      "step": 36000
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00029958078497968973,
      "loss": 0.3568,
      "step": 36500
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002995435585854278,
      "loss": 0.3555,
      "step": 37000
    },
    {
      "epoch": 0.86,
      "eval_loss": 0.3368704915046692,
      "eval_runtime": 1.3166,
      "eval_samples_per_second": 1669.417,
      "eval_steps_per_second": 26.583,
      "step": 37000
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002995047514859671,
      "loss": 0.3542,
      "step": 37500
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002994643641056959,
      "loss": 0.3532,
      "step": 38000
    },
    {
      "epoch": 0.89,
      "eval_loss": 0.3357352018356323,
      "eval_runtime": 1.2847,
      "eval_samples_per_second": 1710.953,
      "eval_steps_per_second": 27.244,
      "step": 38000
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.000299422396886284,
      "loss": 0.3523,
      "step": 38500
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002993788502866783,
      "loss": 0.3514,
      "step": 39000
    },
    {
      "epoch": 0.91,
      "eval_loss": 0.33294928073883057,
      "eval_runtime": 1.272,
      "eval_samples_per_second": 1728.032,
      "eval_steps_per_second": 27.516,
      "step": 39000
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00029933372478309746,
      "loss": 0.3503,
      "step": 39500
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00029928702086902664,
      "loss": 0.3493,
      "step": 40000
    },
    {
      "epoch": 0.93,
      "eval_loss": 0.332314670085907,
      "eval_runtime": 1.318,
      "eval_samples_per_second": 1667.699,
      "eval_steps_per_second": 26.556,
      "step": 40000
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00029923873905521244,
      "loss": 0.3483,
      "step": 40500
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.000299188879869657,
      "loss": 0.3477,
      "step": 41000
    },
    {
      "epoch": 0.96,
      "eval_loss": 0.33138296008110046,
      "eval_runtime": 1.2875,
      "eval_samples_per_second": 1707.216,
      "eval_steps_per_second": 27.185,
      "step": 41000
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.00029913744385761244,
      "loss": 0.3466,
      "step": 41500
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00029908443158157465,
      "loss": 0.3459,
      "step": 42000
    },
    {
      "epoch": 0.98,
      "eval_loss": 0.32563212513923645,
      "eval_runtime": 1.2983,
      "eval_samples_per_second": 1692.964,
      "eval_steps_per_second": 26.958,
      "step": 42000
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002990298436212775,
      "loss": 0.3448,
      "step": 42500
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002989736805736861,
      "loss": 0.3441,
      "step": 43000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.32700827717781067,
      "eval_runtime": 1.3967,
      "eval_samples_per_second": 1573.716,
      "eval_steps_per_second": 25.059,
      "step": 43000
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00029891594305299065,
      "loss": 0.3432,
      "step": 43500
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00029885663169059926,
      "loss": 0.3424,
      "step": 44000
    },
    {
      "epoch": 1.03,
      "eval_loss": 0.3216361701488495,
      "eval_runtime": 1.4298,
      "eval_samples_per_second": 1537.25,
      "eval_steps_per_second": 24.479,
      "step": 44000
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002987957471351316,
      "loss": 0.3415,
      "step": 44500
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.00029873329005241137,
      "loss": 0.3408,
      "step": 45000
    },
    {
      "epoch": 1.05,
      "eval_loss": 0.324531614780426,
      "eval_runtime": 1.3974,
      "eval_samples_per_second": 1572.95,
      "eval_steps_per_second": 25.047,
      "step": 45000
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00029866926112545925,
      "loss": 0.3399,
      "step": 45500
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.00029860366105448534,
      "loss": 0.3394,
      "step": 46000
    },
    {
      "epoch": 1.07,
      "eval_loss": 0.3231622278690338,
      "eval_runtime": 1.3921,
      "eval_samples_per_second": 1578.889,
      "eval_steps_per_second": 25.142,
      "step": 46000
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00029853649055688143,
      "loss": 0.3387,
      "step": 46500
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00029846775036721337,
      "loss": 0.338,
      "step": 47000
    },
    {
      "epoch": 1.1,
      "eval_loss": 0.32254886627197266,
      "eval_runtime": 1.4182,
      "eval_samples_per_second": 1549.86,
      "eval_steps_per_second": 24.679,
      "step": 47000
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002983974412372129,
      "loss": 0.337,
      "step": 47500
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00029832556393576934,
      "loss": 0.3363,
      "step": 48000
    },
    {
      "epoch": 1.12,
      "eval_loss": 0.3195147216320038,
      "eval_runtime": 1.408,
      "eval_samples_per_second": 1561.07,
      "eval_steps_per_second": 24.858,
      "step": 48000
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002982521192489214,
      "loss": 0.3357,
      "step": 48500
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002981771079798483,
      "loss": 0.3353,
      "step": 49000
    },
    {
      "epoch": 1.14,
      "eval_loss": 0.3167419731616974,
      "eval_runtime": 1.4463,
      "eval_samples_per_second": 1519.733,
      "eval_steps_per_second": 24.2,
      "step": 49000
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00029810053094886136,
      "loss": 0.3345,
      "step": 49500
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00029802238899339473,
      "loss": 0.3338,
      "step": 50000
    },
    {
      "epoch": 1.16,
      "eval_loss": 0.3171909749507904,
      "eval_runtime": 1.4105,
      "eval_samples_per_second": 1558.283,
      "eval_steps_per_second": 24.813,
      "step": 50000
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002979426829679962,
      "loss": 0.3334,
      "step": 50500
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002978614137443183,
      "loss": 0.3325,
      "step": 51000
    },
    {
      "epoch": 1.19,
      "eval_loss": 0.31682637333869934,
      "eval_runtime": 1.4213,
      "eval_samples_per_second": 1546.472,
      "eval_steps_per_second": 24.625,
      "step": 51000
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.000297778582211108,
      "loss": 0.3322,
      "step": 51500
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.00029769418927419786,
      "loss": 0.3316,
      "step": 52000
    },
    {
      "epoch": 1.21,
      "eval_loss": 0.31410133838653564,
      "eval_runtime": 1.4261,
      "eval_samples_per_second": 1541.269,
      "eval_steps_per_second": 24.542,
      "step": 52000
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002976082358564954,
      "loss": 0.331,
      "step": 52500
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00029752072289797353,
      "loss": 0.3305,
      "step": 53000
    },
    {
      "epoch": 1.23,
      "eval_loss": 0.3114263713359833,
      "eval_runtime": 1.4182,
      "eval_samples_per_second": 1549.888,
      "eval_steps_per_second": 24.68,
      "step": 53000
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00029743165135565986,
      "loss": 0.3301,
      "step": 53500
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00029734102220362654,
      "loss": 0.3303,
      "step": 54000
    },
    {
      "epoch": 1.26,
      "eval_loss": 0.3110273778438568,
      "eval_runtime": 1.3899,
      "eval_samples_per_second": 1581.462,
      "eval_steps_per_second": 25.183,
      "step": 54000
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.00029724883643297937,
      "loss": 0.3289,
      "step": 54500
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002971550950518473,
      "loss": 0.3284,
      "step": 55000
    },
    {
      "epoch": 1.28,
      "eval_loss": 0.3112471401691437,
      "eval_runtime": 1.3713,
      "eval_samples_per_second": 1602.906,
      "eval_steps_per_second": 25.524,
      "step": 55000
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.000297059799085371,
      "loss": 0.3284,
      "step": 55500
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.00029696294957569196,
      "loss": 0.3276,
      "step": 56000
    },
    {
      "epoch": 1.3,
      "eval_loss": 0.30989283323287964,
      "eval_runtime": 1.4309,
      "eval_samples_per_second": 1536.096,
      "eval_steps_per_second": 24.46,
      "step": 56000
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.00029686454758194076,
      "loss": 0.3269,
      "step": 56500
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00029676459418022594,
      "loss": 0.3264,
      "step": 57000
    },
    {
      "epoch": 1.33,
      "eval_loss": 0.3089369833469391,
      "eval_runtime": 1.4226,
      "eval_samples_per_second": 1545.068,
      "eval_steps_per_second": 24.603,
      "step": 57000
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002966630904636219,
      "loss": 0.3261,
      "step": 57500
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002965600375421569,
      "loss": 0.3256,
      "step": 58000
    },
    {
      "epoch": 1.35,
      "eval_loss": 0.31043851375579834,
      "eval_runtime": 1.4043,
      "eval_samples_per_second": 1565.245,
      "eval_steps_per_second": 24.924,
      "step": 58000
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002964554365428013,
      "loss": 0.3251,
      "step": 58500
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.00029634928860945486,
      "loss": 0.3247,
      "step": 59000
    },
    {
      "epoch": 1.37,
      "eval_loss": 0.3088381588459015,
      "eval_runtime": 1.4281,
      "eval_samples_per_second": 1539.081,
      "eval_steps_per_second": 24.508,
      "step": 59000
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002962415949029343,
      "loss": 0.3243,
      "step": 59500
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00029613235660096084,
      "loss": 0.3241,
      "step": 60000
    },
    {
      "epoch": 1.4,
      "eval_loss": 0.3055568337440491,
      "eval_runtime": 1.4302,
      "eval_samples_per_second": 1536.88,
      "eval_steps_per_second": 24.473,
      "step": 60000
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00029602157489814693,
      "loss": 0.3234,
      "step": 60500
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00029590925100598365,
      "loss": 0.3229,
      "step": 61000
    },
    {
      "epoch": 1.42,
      "eval_loss": 0.3047891855239868,
      "eval_runtime": 1.4211,
      "eval_samples_per_second": 1546.647,
      "eval_steps_per_second": 24.628,
      "step": 61000
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.000295795386152827,
      "loss": 0.323,
      "step": 61500
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002956799815838848,
      "loss": 0.3223,
      "step": 62000
    },
    {
      "epoch": 1.44,
      "eval_loss": 0.30619609355926514,
      "eval_runtime": 1.3977,
      "eval_samples_per_second": 1572.599,
      "eval_steps_per_second": 25.041,
      "step": 62000
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002955630385612029,
      "loss": 0.3217,
      "step": 62500
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002954445583636515,
      "loss": 0.3215,
      "step": 63000
    },
    {
      "epoch": 1.47,
      "eval_loss": 0.3042140007019043,
      "eval_runtime": 1.3779,
      "eval_samples_per_second": 1595.193,
      "eval_steps_per_second": 25.401,
      "step": 63000
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00029532454228691103,
      "loss": 0.3212,
      "step": 63500
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002952029916434581,
      "loss": 0.3208,
      "step": 64000
    },
    {
      "epoch": 1.49,
      "eval_loss": 0.30290499329566956,
      "eval_runtime": 1.3826,
      "eval_samples_per_second": 1589.79,
      "eval_steps_per_second": 25.315,
      "step": 64000
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.00029507990776255107,
      "loss": 0.3206,
      "step": 64500
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00029495529199021555,
      "loss": 0.3201,
      "step": 65000
    },
    {
      "epoch": 1.51,
      "eval_loss": 0.30093201994895935,
      "eval_runtime": 1.3496,
      "eval_samples_per_second": 1628.653,
      "eval_steps_per_second": 25.934,
      "step": 65000
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002948291456892296,
      "loss": 0.3199,
      "step": 65500
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.00029470147023910907,
      "loss": 0.3198,
      "step": 66000
    },
    {
      "epoch": 1.54,
      "eval_loss": 0.3037390410900116,
      "eval_runtime": 1.4125,
      "eval_samples_per_second": 1556.101,
      "eval_steps_per_second": 24.779,
      "step": 66000
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002945722670360921,
      "loss": 0.3189,
      "step": 66500
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002944415374931243,
      "loss": 0.3187,
      "step": 67000
    },
    {
      "epoch": 1.56,
      "eval_loss": 0.30209919810295105,
      "eval_runtime": 1.3711,
      "eval_samples_per_second": 1603.144,
      "eval_steps_per_second": 25.528,
      "step": 67000
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00029430928303984295,
      "loss": 0.3184,
      "step": 67500
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002941755051225616,
      "loss": 0.3179,
      "step": 68000
    },
    {
      "epoch": 1.58,
      "eval_loss": 0.29896965622901917,
      "eval_runtime": 1.5715,
      "eval_samples_per_second": 1398.699,
      "eval_steps_per_second": 22.272,
      "step": 68000
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00029404020520425417,
      "loss": 0.3177,
      "step": 68500
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002939033847645388,
      "loss": 0.3176,
      "step": 69000
    },
    {
      "epoch": 1.61,
      "eval_loss": 0.30087536573410034,
      "eval_runtime": 1.4497,
      "eval_samples_per_second": 1516.182,
      "eval_steps_per_second": 24.143,
      "step": 69000
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00029376504529966195,
      "loss": 0.3172,
      "step": 69500
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00029362518832248184,
      "loss": 0.3169,
      "step": 70000
    },
    {
      "epoch": 1.63,
      "eval_loss": 0.3002066910266876,
      "eval_runtime": 1.4297,
      "eval_samples_per_second": 1537.403,
      "eval_steps_per_second": 24.481,
      "step": 70000
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002934838153624519,
      "loss": 0.3166,
      "step": 70500
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00029334092796560427,
      "loss": 0.3161,
      "step": 71000
    },
    {
      "epoch": 1.65,
      "eval_loss": 0.2996619939804077,
      "eval_runtime": 1.4139,
      "eval_samples_per_second": 1554.517,
      "eval_steps_per_second": 24.753,
      "step": 71000
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002931965276945326,
      "loss": 0.3159,
      "step": 71500
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002930506161283751,
      "loss": 0.3156,
      "step": 72000
    },
    {
      "epoch": 1.68,
      "eval_loss": 0.29820874333381653,
      "eval_runtime": 1.408,
      "eval_samples_per_second": 1561.093,
      "eval_steps_per_second": 24.858,
      "step": 72000
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00029290319486279724,
      "loss": 0.3152,
      "step": 72500
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002927542655099744,
      "loss": 0.3148,
      "step": 73000
    },
    {
      "epoch": 1.7,
      "eval_loss": 0.2982105612754822,
      "eval_runtime": 1.3918,
      "eval_samples_per_second": 1579.253,
      "eval_steps_per_second": 25.147,
      "step": 73000
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00029260382969857417,
      "loss": 0.3148,
      "step": 73500
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00029245188907373845,
      "loss": 0.3143,
      "step": 74000
    },
    {
      "epoch": 1.72,
      "eval_loss": 0.2960808277130127,
      "eval_runtime": 1.4226,
      "eval_samples_per_second": 1545.079,
      "eval_steps_per_second": 24.603,
      "step": 74000
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002922984452970655,
      "loss": 0.3141,
      "step": 74500
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.000292143500046592,
      "loss": 0.3137,
      "step": 75000
    },
    {
      "epoch": 1.75,
      "eval_loss": 0.2935367524623871,
      "eval_runtime": 1.3992,
      "eval_samples_per_second": 1570.943,
      "eval_steps_per_second": 25.015,
      "step": 75000
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002919870550167743,
      "loss": 0.3133,
      "step": 75500
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002918291119184702,
      "loss": 0.3132,
      "step": 76000
    },
    {
      "epoch": 1.77,
      "eval_loss": 0.2974264919757843,
      "eval_runtime": 1.3828,
      "eval_samples_per_second": 1589.484,
      "eval_steps_per_second": 25.31,
      "step": 76000
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002916696724789201,
      "loss": 0.3134,
      "step": 76500
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00029150873844172823,
      "loss": 0.313,
      "step": 77000
    },
    {
      "epoch": 1.79,
      "eval_loss": 0.29650062322616577,
      "eval_runtime": 1.4037,
      "eval_samples_per_second": 1565.884,
      "eval_steps_per_second": 24.934,
      "step": 77000
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00029134631156684334,
      "loss": 0.3128,
      "step": 77500
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002911823936305398,
      "loss": 0.3147,
      "step": 78000
    },
    {
      "epoch": 1.82,
      "eval_loss": 0.2946864664554596,
      "eval_runtime": 1.3928,
      "eval_samples_per_second": 1578.113,
      "eval_steps_per_second": 25.129,
      "step": 78000
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002910169864253979,
      "loss": 0.3123,
      "step": 78500
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002908500917602842,
      "loss": 0.3118,
      "step": 79000
    },
    {
      "epoch": 1.84,
      "eval_loss": 0.2940049469470978,
      "eval_runtime": 1.405,
      "eval_samples_per_second": 1564.438,
      "eval_steps_per_second": 24.911,
      "step": 79000
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.00029068171146033226,
      "loss": 0.3115,
      "step": 79500
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002905118473669218,
      "loss": 0.3113,
      "step": 80000
    },
    {
      "epoch": 1.86,
      "eval_loss": 0.293503075838089,
      "eval_runtime": 1.4595,
      "eval_samples_per_second": 1505.956,
      "eval_steps_per_second": 23.98,
      "step": 80000
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00029034050133765947,
      "loss": 0.3111,
      "step": 80500
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.00029016767524635804,
      "loss": 0.3109,
      "step": 81000
    },
    {
      "epoch": 1.89,
      "eval_loss": 0.2935539484024048,
      "eval_runtime": 1.3879,
      "eval_samples_per_second": 1583.739,
      "eval_steps_per_second": 25.219,
      "step": 81000
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00028999337098301585,
      "loss": 0.3108,
      "step": 81500
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002898175904537964,
      "loss": 0.3103,
      "step": 82000
    },
    {
      "epoch": 1.91,
      "eval_loss": 0.2912616431713104,
      "eval_runtime": 1.3921,
      "eval_samples_per_second": 1578.853,
      "eval_steps_per_second": 25.141,
      "step": 82000
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002896403355810075,
      "loss": 0.3104,
      "step": 82500
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.00028946160830307997,
      "loss": 0.3102,
      "step": 83000
    },
    {
      "epoch": 1.93,
      "eval_loss": 0.29547759890556335,
      "eval_runtime": 1.4163,
      "eval_samples_per_second": 1551.936,
      "eval_steps_per_second": 24.712,
      "step": 83000
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00028928141057454665,
      "loss": 0.3099,
      "step": 83500
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002890997443660211,
      "loss": 0.3096,
      "step": 84000
    },
    {
      "epoch": 1.96,
      "eval_loss": 0.29173827171325684,
      "eval_runtime": 1.3888,
      "eval_samples_per_second": 1582.667,
      "eval_steps_per_second": 25.202,
      "step": 84000
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00028891661166417586,
      "loss": 0.3089,
      "step": 84500
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.00028873201447172074,
      "loss": 0.3093,
      "step": 85000
    },
    {
      "epoch": 1.98,
      "eval_loss": 0.2930682897567749,
      "eval_runtime": 1.4086,
      "eval_samples_per_second": 1560.366,
      "eval_steps_per_second": 24.847,
      "step": 85000
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002885459548073812,
      "loss": 0.309,
      "step": 85500
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002883584347058758,
      "loss": 0.3084,
      "step": 86000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.2906832695007324,
      "eval_runtime": 1.4896,
      "eval_samples_per_second": 1475.535,
      "eval_steps_per_second": 23.496,
      "step": 86000
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.00028816945621789437,
      "loss": 0.3086,
      "step": 86500
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002879790214100753,
      "loss": 0.3086,
      "step": 87000
    },
    {
      "epoch": 2.03,
      "eval_loss": 0.2908460795879364,
      "eval_runtime": 1.4139,
      "eval_samples_per_second": 1554.553,
      "eval_steps_per_second": 24.754,
      "step": 87000
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002877871323649833,
      "loss": 0.3078,
      "step": 87500
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002875937911810861,
      "loss": 0.3077,
      "step": 88000
    },
    {
      "epoch": 2.05,
      "eval_loss": 0.2881493866443634,
      "eval_runtime": 1.4042,
      "eval_samples_per_second": 1565.313,
      "eval_steps_per_second": 24.925,
      "step": 88000
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.000287398999972732,
      "loss": 0.3075,
      "step": 88500
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.00028720276087012636,
      "loss": 0.3087,
      "step": 89000
    },
    {
      "epoch": 2.07,
      "eval_loss": 0.2877899706363678,
      "eval_runtime": 1.3911,
      "eval_samples_per_second": 1580.034,
      "eval_steps_per_second": 25.16,
      "step": 89000
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002870050760193086,
      "loss": 0.3067,
      "step": 89500
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.00028680594758212854,
      "loss": 0.3065,
      "step": 90000
    },
    {
      "epoch": 2.1,
      "eval_loss": 0.2883204519748688,
      "eval_runtime": 1.3997,
      "eval_samples_per_second": 1570.392,
      "eval_steps_per_second": 25.006,
      "step": 90000
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.00028660537773622294,
      "loss": 0.3073,
      "step": 90500
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.00028640336867499143,
      "loss": 0.3063,
      "step": 91000
    },
    {
      "epoch": 2.12,
      "eval_loss": 0.28914499282836914,
      "eval_runtime": 1.4031,
      "eval_samples_per_second": 1566.538,
      "eval_steps_per_second": 24.945,
      "step": 91000
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002861999226075728,
      "loss": 0.306,
      "step": 91500
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002859950417588206,
      "loss": 0.3057,
      "step": 92000
    },
    {
      "epoch": 2.14,
      "eval_loss": 0.28657251596450806,
      "eval_runtime": 1.3993,
      "eval_samples_per_second": 1570.823,
      "eval_steps_per_second": 25.013,
      "step": 92000
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.00028578872836927904,
      "loss": 0.3057,
      "step": 92500
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002855809846951582,
      "loss": 0.3066,
      "step": 93000
    },
    {
      "epoch": 2.17,
      "eval_loss": 0.2893889248371124,
      "eval_runtime": 1.4241,
      "eval_samples_per_second": 1543.421,
      "eval_steps_per_second": 24.577,
      "step": 93000
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.00028537181300830963,
      "loss": 0.305,
      "step": 93500
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002851612155962014,
      "loss": 0.3048,
      "step": 94000
    },
    {
      "epoch": 2.19,
      "eval_loss": 0.2858821451663971,
      "eval_runtime": 1.4056,
      "eval_samples_per_second": 1563.797,
      "eval_steps_per_second": 24.901,
      "step": 94000
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002849491947618932,
      "loss": 0.3048,
      "step": 94500
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002847357528240107,
      "loss": 0.3051,
      "step": 95000
    },
    {
      "epoch": 2.21,
      "eval_loss": 0.28810954093933105,
      "eval_runtime": 1.4039,
      "eval_samples_per_second": 1565.629,
      "eval_steps_per_second": 24.93,
      "step": 95000
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002845208921167208,
      "loss": 0.3047,
      "step": 95500
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.00028430461498970584,
      "loss": 0.3041,
      "step": 96000
    },
    {
      "epoch": 2.24,
      "eval_loss": 0.28588977456092834,
      "eval_runtime": 1.4454,
      "eval_samples_per_second": 1520.728,
      "eval_steps_per_second": 24.215,
      "step": 96000
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.00028408692380813775,
      "loss": 0.3042,
      "step": 96500
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.00028386782095265247,
      "loss": 0.3039,
      "step": 97000
    },
    {
      "epoch": 2.26,
      "eval_loss": 0.2837139666080475,
      "eval_runtime": 1.414,
      "eval_samples_per_second": 1554.434,
      "eval_steps_per_second": 24.752,
      "step": 97000
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002836473088193237,
      "loss": 0.3036,
      "step": 97500
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.00028342538981963677,
      "loss": 0.3032,
      "step": 98000
    },
    {
      "epoch": 2.28,
      "eval_loss": 0.28603771328926086,
      "eval_runtime": 1.4079,
      "eval_samples_per_second": 1561.179,
      "eval_steps_per_second": 24.86,
      "step": 98000
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002832020663804624,
      "loss": 0.3035,
      "step": 98500
    },
    {
      "epoch": 2.31,
      "learning_rate": 0.00028297734094402986,
      "loss": 0.3039,
      "step": 99000
    },
    {
      "epoch": 2.31,
      "eval_loss": 0.2857549488544464,
      "eval_runtime": 1.4042,
      "eval_samples_per_second": 1565.268,
      "eval_steps_per_second": 24.925,
      "step": 99000
    },
    {
      "epoch": 2.32,
      "learning_rate": 0.0002827512159679005,
      "loss": 0.3027,
      "step": 99500
    },
    {
      "epoch": 2.33,
      "learning_rate": 0.00028252369392494086,
      "loss": 0.3025,
      "step": 100000
    },
    {
      "epoch": 2.33,
      "eval_loss": 0.28487083315849304,
      "eval_runtime": 1.4056,
      "eval_samples_per_second": 1563.737,
      "eval_steps_per_second": 24.9,
      "step": 100000
    },
    {
      "epoch": 2.34,
      "learning_rate": 0.0002822947773032956,
      "loss": 0.3022,
      "step": 100500
    },
    {
      "epoch": 2.35,
      "learning_rate": 0.0002820644686063602,
      "loss": 0.3023,
      "step": 101000
    },
    {
      "epoch": 2.35,
      "eval_loss": 0.2856578528881073,
      "eval_runtime": 1.4144,
      "eval_samples_per_second": 1553.982,
      "eval_steps_per_second": 24.745,
      "step": 101000
    },
    {
      "epoch": 2.36,
      "learning_rate": 0.00028183277035275363,
      "loss": 0.3022,
      "step": 101500
    },
    {
      "epoch": 2.38,
      "learning_rate": 0.0002815996850762909,
      "loss": 0.3024,
      "step": 102000
    },
    {
      "epoch": 2.38,
      "eval_loss": 0.28789734840393066,
      "eval_runtime": 1.8256,
      "eval_samples_per_second": 1203.971,
      "eval_steps_per_second": 19.172,
      "step": 102000
    },
    {
      "epoch": 2.39,
      "learning_rate": 0.00028136521532595515,
      "loss": 0.3022,
      "step": 102500
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.00028112936366587023,
      "loss": 0.3018,
      "step": 103000
    },
    {
      "epoch": 2.4,
      "eval_loss": 0.28655195236206055,
      "eval_runtime": 1.4347,
      "eval_samples_per_second": 1532.011,
      "eval_steps_per_second": 24.395,
      "step": 103000
    },
    {
      "epoch": 2.41,
      "learning_rate": 0.00028089213267527184,
      "loss": 0.3018,
      "step": 103500
    },
    {
      "epoch": 2.42,
      "learning_rate": 0.0002806535249484803,
      "loss": 0.3015,
      "step": 104000
    },
    {
      "epoch": 2.42,
      "eval_loss": 0.2860538959503174,
      "eval_runtime": 1.4232,
      "eval_samples_per_second": 1544.364,
      "eval_steps_per_second": 24.592,
      "step": 104000
    },
    {
      "epoch": 2.43,
      "learning_rate": 0.00028041354309487135,
      "loss": 0.3013,
      "step": 104500
    },
    {
      "epoch": 2.45,
      "learning_rate": 0.0002801721897388482,
      "loss": 0.301,
      "step": 105000
    },
    {
      "epoch": 2.45,
      "eval_loss": 0.28530260920524597,
      "eval_runtime": 1.4494,
      "eval_samples_per_second": 1516.459,
      "eval_steps_per_second": 24.147,
      "step": 105000
    },
    {
      "epoch": 2.46,
      "learning_rate": 0.0002799294675198124,
      "loss": 0.3018,
      "step": 105500
    },
    {
      "epoch": 2.47,
      "learning_rate": 0.00027968537909213524,
      "loss": 0.3008,
      "step": 106000
    },
    {
      "epoch": 2.47,
      "eval_loss": 0.285876989364624,
      "eval_runtime": 1.4506,
      "eval_samples_per_second": 1515.222,
      "eval_steps_per_second": 24.128,
      "step": 106000
    },
    {
      "epoch": 2.48,
      "learning_rate": 0.0002794399271251287,
      "loss": 0.3013,
      "step": 106500
    },
    {
      "epoch": 2.49,
      "learning_rate": 0.0002791931143030162,
      "loss": 0.3006,
      "step": 107000
    },
    {
      "epoch": 2.49,
      "eval_loss": 0.28289294242858887,
      "eval_runtime": 1.4223,
      "eval_samples_per_second": 1545.438,
      "eval_steps_per_second": 24.609,
      "step": 107000
    },
    {
      "epoch": 2.5,
      "learning_rate": 0.00027894494332490315,
      "loss": 0.3001,
      "step": 107500
    },
    {
      "epoch": 2.52,
      "learning_rate": 0.0002786954169047476,
      "loss": 0.3003,
      "step": 108000
    },
    {
      "epoch": 2.52,
      "eval_loss": 0.2835444509983063,
      "eval_runtime": 1.4445,
      "eval_samples_per_second": 1521.603,
      "eval_steps_per_second": 24.229,
      "step": 108000
    },
    {
      "epoch": 2.53,
      "learning_rate": 0.0002784445377713306,
      "loss": 0.2999,
      "step": 108500
    },
    {
      "epoch": 2.54,
      "learning_rate": 0.0002781923086682261,
      "loss": 0.2998,
      "step": 109000
    },
    {
      "epoch": 2.54,
      "eval_loss": 0.27993467450141907,
      "eval_runtime": 1.4177,
      "eval_samples_per_second": 1550.426,
      "eval_steps_per_second": 24.688,
      "step": 109000
    },
    {
      "epoch": 2.55,
      "learning_rate": 0.0002779387323537711,
      "loss": 0.2994,
      "step": 109500
    },
    {
      "epoch": 2.56,
      "learning_rate": 0.0002776838116010356,
      "loss": 0.3001,
      "step": 110000
    },
    {
      "epoch": 2.56,
      "eval_loss": 0.28160494565963745,
      "eval_runtime": 1.4324,
      "eval_samples_per_second": 1534.54,
      "eval_steps_per_second": 24.435,
      "step": 110000
    },
    {
      "epoch": 2.57,
      "learning_rate": 0.0002774275491977922,
      "loss": 0.2993,
      "step": 110500
    },
    {
      "epoch": 2.59,
      "learning_rate": 0.0002771699479464853,
      "loss": 0.2991,
      "step": 111000
    },
    {
      "epoch": 2.59,
      "eval_loss": 0.28305351734161377,
      "eval_runtime": 1.4425,
      "eval_samples_per_second": 1523.692,
      "eval_steps_per_second": 24.263,
      "step": 111000
    },
    {
      "epoch": 2.6,
      "learning_rate": 0.00027691101066420104,
      "loss": 0.2993,
      "step": 111500
    },
    {
      "epoch": 2.61,
      "learning_rate": 0.0002766507401826361,
      "loss": 0.2987,
      "step": 112000
    },
    {
      "epoch": 2.61,
      "eval_loss": 0.2818717956542969,
      "eval_runtime": 1.4177,
      "eval_samples_per_second": 1550.356,
      "eval_steps_per_second": 24.687,
      "step": 112000
    },
    {
      "epoch": 2.62,
      "learning_rate": 0.0002763891393480666,
      "loss": 0.2989,
      "step": 112500
    },
    {
      "epoch": 2.63,
      "learning_rate": 0.0002761262110213175,
      "loss": 0.2985,
      "step": 113000
    },
    {
      "epoch": 2.63,
      "eval_loss": 0.283016562461853,
      "eval_runtime": 1.4225,
      "eval_samples_per_second": 1545.128,
      "eval_steps_per_second": 24.604,
      "step": 113000
    },
    {
      "epoch": 2.64,
      "learning_rate": 0.00027586195807773083,
      "loss": 0.2989,
      "step": 113500
    },
    {
      "epoch": 2.66,
      "learning_rate": 0.00027559638340713435,
      "loss": 0.2981,
      "step": 114000
    },
    {
      "epoch": 2.66,
      "eval_loss": 0.2824258804321289,
      "eval_runtime": 1.412,
      "eval_samples_per_second": 1556.637,
      "eval_steps_per_second": 24.787,
      "step": 114000
    },
    {
      "epoch": 2.67,
      "learning_rate": 0.00027532948991381025,
      "loss": 0.2976,
      "step": 114500
    },
    {
      "epoch": 2.68,
      "learning_rate": 0.00027506128051646287,
      "loss": 0.2978,
      "step": 115000
    },
    {
      "epoch": 2.68,
      "eval_loss": 0.2790142595767975,
      "eval_runtime": 1.436,
      "eval_samples_per_second": 1530.608,
      "eval_steps_per_second": 24.373,
      "step": 115000
    },
    {
      "epoch": 2.69,
      "learning_rate": 0.00027479175814818733,
      "loss": 0.2982,
      "step": 115500
    },
    {
      "epoch": 2.7,
      "learning_rate": 0.000274520925756437,
      "loss": 0.2973,
      "step": 116000
    },
    {
      "epoch": 2.7,
      "eval_loss": 0.27925553917884827,
      "eval_runtime": 1.4557,
      "eval_samples_per_second": 1509.89,
      "eval_steps_per_second": 24.043,
      "step": 116000
    },
    {
      "epoch": 2.71,
      "learning_rate": 0.00027424878630299157,
      "loss": 0.2971,
      "step": 116500
    },
    {
      "epoch": 2.73,
      "learning_rate": 0.0002739753427639244,
      "loss": 0.2977,
      "step": 117000
    },
    {
      "epoch": 2.73,
      "eval_loss": 0.2807333171367645,
      "eval_runtime": 1.4375,
      "eval_samples_per_second": 1529.048,
      "eval_steps_per_second": 24.348,
      "step": 117000
    },
    {
      "epoch": 2.74,
      "learning_rate": 0.0002737005981295704,
      "loss": 0.2973,
      "step": 117500
    },
    {
      "epoch": 2.75,
      "learning_rate": 0.0002734245554044927,
      "loss": 0.2969,
      "step": 118000
    },
    {
      "epoch": 2.75,
      "eval_loss": 0.28166475892066956,
      "eval_runtime": 1.4453,
      "eval_samples_per_second": 1520.838,
      "eval_steps_per_second": 24.217,
      "step": 118000
    },
    {
      "epoch": 2.76,
      "learning_rate": 0.0002731472176074504,
      "loss": 0.2967,
      "step": 118500
    },
    {
      "epoch": 2.77,
      "learning_rate": 0.0002728685877713653,
      "loss": 0.2967,
      "step": 119000
    },
    {
      "epoch": 2.77,
      "eval_loss": 0.2808319330215454,
      "eval_runtime": 1.4174,
      "eval_samples_per_second": 1550.705,
      "eval_steps_per_second": 24.693,
      "step": 119000
    },
    {
      "epoch": 2.78,
      "learning_rate": 0.0002725886689432884,
      "loss": 0.2965,
      "step": 119500
    },
    {
      "epoch": 2.8,
      "learning_rate": 0.0002723074641843674,
      "loss": 0.2965,
      "step": 120000
    },
    {
      "epoch": 2.8,
      "eval_loss": 0.27865585684776306,
      "eval_runtime": 1.4415,
      "eval_samples_per_second": 1524.849,
      "eval_steps_per_second": 24.281,
      "step": 120000
    },
    {
      "epoch": 2.81,
      "learning_rate": 0.0002720249765698123,
      "loss": 0.2963,
      "step": 120500
    },
    {
      "epoch": 2.82,
      "learning_rate": 0.0002717412091888626,
      "loss": 0.2958,
      "step": 121000
    },
    {
      "epoch": 2.82,
      "eval_loss": 0.2785983681678772,
      "eval_runtime": 1.4123,
      "eval_samples_per_second": 1556.34,
      "eval_steps_per_second": 24.782,
      "step": 121000
    },
    {
      "epoch": 2.83,
      "learning_rate": 0.00027145616514475274,
      "loss": 0.2963,
      "step": 121500
    },
    {
      "epoch": 2.84,
      "learning_rate": 0.0002711698475546788,
      "loss": 0.2968,
      "step": 122000
    },
    {
      "epoch": 2.84,
      "eval_loss": 0.27918389439582825,
      "eval_runtime": 1.4238,
      "eval_samples_per_second": 1543.71,
      "eval_steps_per_second": 24.581,
      "step": 122000
    },
    {
      "epoch": 2.85,
      "learning_rate": 0.00027088225954976407,
      "loss": 0.2965,
      "step": 122500
    },
    {
      "epoch": 2.87,
      "learning_rate": 0.0002705934042750249,
      "loss": 0.2956,
      "step": 123000
    },
    {
      "epoch": 2.87,
      "eval_loss": 0.27785566449165344,
      "eval_runtime": 1.3994,
      "eval_samples_per_second": 1570.623,
      "eval_steps_per_second": 25.01,
      "step": 123000
    },
    {
      "epoch": 2.88,
      "learning_rate": 0.00027030328488933625,
      "loss": 0.2962,
      "step": 123500
    },
    {
      "epoch": 2.89,
      "learning_rate": 0.00027001190456539726,
      "loss": 0.2948,
      "step": 124000
    },
    {
      "epoch": 2.89,
      "eval_loss": 0.2769564390182495,
      "eval_runtime": 1.4165,
      "eval_samples_per_second": 1551.741,
      "eval_steps_per_second": 24.709,
      "step": 124000
    },
    {
      "epoch": 2.9,
      "learning_rate": 0.0002697192664896965,
      "loss": 0.295,
      "step": 124500
    },
    {
      "epoch": 2.91,
      "learning_rate": 0.00026942537386247706,
      "loss": 0.2952,
      "step": 125000
    },
    {
      "epoch": 2.91,
      "eval_loss": 0.27958372235298157,
      "eval_runtime": 1.3901,
      "eval_samples_per_second": 1581.168,
      "eval_steps_per_second": 25.178,
      "step": 125000
    },
    {
      "epoch": 2.92,
      "learning_rate": 0.0002691302298977016,
      "loss": 0.2948,
      "step": 125500
    },
    {
      "epoch": 2.94,
      "learning_rate": 0.0002688338378230173,
      "loss": 0.2943,
      "step": 126000
    },
    {
      "epoch": 2.94,
      "eval_loss": 0.2770977318286896,
      "eval_runtime": 1.4225,
      "eval_samples_per_second": 1545.165,
      "eval_steps_per_second": 24.605,
      "step": 126000
    },
    {
      "epoch": 2.95,
      "learning_rate": 0.00026853620087972035,
      "loss": 0.2945,
      "step": 126500
    },
    {
      "epoch": 2.96,
      "learning_rate": 0.00026823732232272065,
      "loss": 0.2945,
      "step": 127000
    },
    {
      "epoch": 2.96,
      "eval_loss": 0.27677854895591736,
      "eval_runtime": 1.4513,
      "eval_samples_per_second": 1514.548,
      "eval_steps_per_second": 24.117,
      "step": 127000
    },
    {
      "epoch": 2.97,
      "learning_rate": 0.0002679372054205063,
      "loss": 0.2949,
      "step": 127500
    },
    {
      "epoch": 2.98,
      "learning_rate": 0.0002676358534551076,
      "loss": 0.2945,
      "step": 128000
    },
    {
      "epoch": 2.98,
      "eval_loss": 0.27655109763145447,
      "eval_runtime": 1.4264,
      "eval_samples_per_second": 1540.932,
      "eval_steps_per_second": 24.537,
      "step": 128000
    },
    {
      "epoch": 2.99,
      "learning_rate": 0.00026733326972206133,
      "loss": 0.2942,
      "step": 128500
    },
    {
      "epoch": 3.01,
      "learning_rate": 0.0002670294575303748,
      "loss": 0.2944,
      "step": 129000
    },
    {
      "epoch": 3.01,
      "eval_loss": 0.2760108709335327,
      "eval_runtime": 1.4342,
      "eval_samples_per_second": 1532.608,
      "eval_steps_per_second": 24.405,
      "step": 129000
    },
    {
      "epoch": 3.02,
      "learning_rate": 0.0002667244202024894,
      "loss": 0.2942,
      "step": 129500
    },
    {
      "epoch": 3.03,
      "learning_rate": 0.00026641816107424453,
      "loss": 0.2944,
      "step": 130000
    },
    {
      "epoch": 3.03,
      "eval_loss": 0.2757641077041626,
      "eval_runtime": 1.4126,
      "eval_samples_per_second": 1555.976,
      "eval_steps_per_second": 24.777,
      "step": 130000
    },
    {
      "epoch": 3.04,
      "learning_rate": 0.0002661106834948409,
      "loss": 0.2939,
      "step": 130500
    },
    {
      "epoch": 3.05,
      "learning_rate": 0.0002658019908268041,
      "loss": 0.2937,
      "step": 131000
    },
    {
      "epoch": 3.05,
      "eval_loss": 0.27508604526519775,
      "eval_runtime": 1.4205,
      "eval_samples_per_second": 1547.302,
      "eval_steps_per_second": 24.639,
      "step": 131000
    },
    {
      "epoch": 3.06,
      "learning_rate": 0.00026549208644594766,
      "loss": 0.294,
      "step": 131500
    },
    {
      "epoch": 3.08,
      "learning_rate": 0.00026518097374133627,
      "loss": 0.2935,
      "step": 132000
    },
    {
      "epoch": 3.08,
      "eval_loss": 0.2736782431602478,
      "eval_runtime": 1.4305,
      "eval_samples_per_second": 1536.563,
      "eval_steps_per_second": 24.468,
      "step": 132000
    },
    {
      "epoch": 3.09,
      "learning_rate": 0.00026486865611524853,
      "loss": 0.2932,
      "step": 132500
    },
    {
      "epoch": 3.1,
      "learning_rate": 0.00026455513698314003,
      "loss": 0.2931,
      "step": 133000
    },
    {
      "epoch": 3.1,
      "eval_loss": 0.27545085549354553,
      "eval_runtime": 1.4136,
      "eval_samples_per_second": 1554.95,
      "eval_steps_per_second": 24.76,
      "step": 133000
    },
    {
      "epoch": 3.11,
      "learning_rate": 0.0002642404197736058,
      "loss": 0.2931,
      "step": 133500
    },
    {
      "epoch": 3.12,
      "learning_rate": 0.0002639245079283428,
      "loss": 0.2936,
      "step": 134000
    },
    {
      "epoch": 3.12,
      "eval_loss": 0.27676478028297424,
      "eval_runtime": 1.4314,
      "eval_samples_per_second": 1535.602,
      "eval_steps_per_second": 24.452,
      "step": 134000
    },
    {
      "epoch": 3.13,
      "learning_rate": 0.00026360740490211234,
      "loss": 0.2932,
      "step": 134500
    },
    {
      "epoch": 3.15,
      "learning_rate": 0.0002632891141627023,
      "loss": 0.2934,
      "step": 135000
    },
    {
      "epoch": 3.15,
      "eval_loss": 0.2775753438472748,
      "eval_runtime": 1.4316,
      "eval_samples_per_second": 1535.332,
      "eval_steps_per_second": 24.448,
      "step": 135000
    },
    {
      "epoch": 3.16,
      "learning_rate": 0.00026296963919088923,
      "loss": 0.2927,
      "step": 135500
    },
    {
      "epoch": 3.17,
      "learning_rate": 0.00026264898348040024,
      "loss": 0.2925,
      "step": 136000
    },
    {
      "epoch": 3.17,
      "eval_loss": 0.27160218358039856,
      "eval_runtime": 1.4163,
      "eval_samples_per_second": 1551.918,
      "eval_steps_per_second": 24.712,
      "step": 136000
    },
    {
      "epoch": 3.18,
      "learning_rate": 0.0002623271505378748,
      "loss": 0.2926,
      "step": 136500
    },
    {
      "epoch": 3.19,
      "learning_rate": 0.00026200414388282637,
      "loss": 0.2924,
      "step": 137000
    },
    {
      "epoch": 3.19,
      "eval_loss": 0.2737308442592621,
      "eval_runtime": 1.4266,
      "eval_samples_per_second": 1540.705,
      "eval_steps_per_second": 24.534,
      "step": 137000
    },
    {
      "epoch": 3.2,
      "learning_rate": 0.00026167996704760406,
      "loss": 0.2929,
      "step": 137500
    },
    {
      "epoch": 3.22,
      "learning_rate": 0.00026135462357735375,
      "loss": 0.2931,
      "step": 138000
    },
    {
      "epoch": 3.22,
      "eval_loss": 0.2752697467803955,
      "eval_runtime": 1.4495,
      "eval_samples_per_second": 1516.388,
      "eval_steps_per_second": 24.146,
      "step": 138000
    },
    {
      "epoch": 3.23,
      "learning_rate": 0.0002610281170299795,
      "loss": 0.292,
      "step": 138500
    },
    {
      "epoch": 3.24,
      "learning_rate": 0.00026070045097610465,
      "loss": 0.2917,
      "step": 139000
    },
    {
      "epoch": 3.24,
      "eval_loss": 0.2744641602039337,
      "eval_runtime": 1.4101,
      "eval_samples_per_second": 1558.726,
      "eval_steps_per_second": 24.82,
      "step": 139000
    },
    {
      "epoch": 3.25,
      "learning_rate": 0.0002603716289990326,
      "loss": 0.2923,
      "step": 139500
    },
    {
      "epoch": 3.26,
      "learning_rate": 0.00026004165469470787,
      "loss": 0.2918,
      "step": 140000
    },
    {
      "epoch": 3.26,
      "eval_loss": 0.2773716449737549,
      "eval_runtime": 1.4122,
      "eval_samples_per_second": 1556.477,
      "eval_steps_per_second": 24.785,
      "step": 140000
    }
  ],
  "max_steps": 500000,
  "num_train_epochs": 12,
  "total_flos": 4.4728124518702797e+21,
  "trial_name": null,
  "trial_params": null
}