{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9851268190633256,
  "global_step": 130000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 2.9999999999999997e-06,
      "loss": 0.8784,
      "step": 1000
    },
    {
      "epoch": 0.03,
      "learning_rate": 5.999999999999999e-06,
      "loss": 0.77,
      "step": 2000
    },
    {
      "epoch": 0.05,
      "learning_rate": 8.999999999999999e-06,
      "loss": 0.7664,
      "step": 3000
    },
    {
      "epoch": 0.06,
      "learning_rate": 1.1999999999999999e-05,
      "loss": 0.7655,
      "step": 4000
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.4999999999999999e-05,
      "loss": 0.765,
      "step": 5000
    },
    {
      "epoch": 0.08,
      "eval_runtime": 1.3797,
      "eval_samples_per_second": 724.791,
      "eval_steps_per_second": 11.597,
      "step": 5000
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.7999999999999997e-05,
      "loss": 0.7647,
      "step": 6000
    },
    {
      "epoch": 0.11,
      "learning_rate": 2.1e-05,
      "loss": 0.7644,
      "step": 7000
    },
    {
      "epoch": 0.12,
      "learning_rate": 2.3999999999999997e-05,
      "loss": 0.7638,
      "step": 8000
    },
    {
      "epoch": 0.14,
      "learning_rate": 2.6999999999999996e-05,
      "loss": 0.7633,
      "step": 9000
    },
    {
      "epoch": 0.15,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 0.76,
      "step": 10000
    },
    {
      "epoch": 0.15,
      "eval_runtime": 1.1376,
      "eval_samples_per_second": 879.066,
      "eval_steps_per_second": 14.065,
      "step": 10000
    },
    {
      "epoch": 0.17,
      "learning_rate": 3.2999999999999996e-05,
      "loss": 0.7148,
      "step": 11000
    },
    {
      "epoch": 0.18,
      "learning_rate": 3.5999999999999994e-05,
      "loss": 0.6963,
      "step": 12000
    },
    {
      "epoch": 0.2,
      "learning_rate": 3.9e-05,
      "loss": 0.6755,
      "step": 13000
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.2e-05,
      "loss": 0.6516,
      "step": 14000
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.4999999999999996e-05,
      "loss": 0.6412,
      "step": 15000
    },
    {
      "epoch": 0.23,
      "eval_runtime": 1.1689,
      "eval_samples_per_second": 855.472,
      "eval_steps_per_second": 13.688,
      "step": 15000
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.7999999999999994e-05,
      "loss": 0.6348,
      "step": 16000
    },
    {
      "epoch": 0.26,
      "learning_rate": 5.1e-05,
      "loss": 0.6295,
      "step": 17000
    },
    {
      "epoch": 0.27,
      "learning_rate": 5.399999999999999e-05,
      "loss": 0.6224,
      "step": 18000
    },
    {
      "epoch": 0.29,
      "learning_rate": 5.6999999999999996e-05,
      "loss": 0.6169,
      "step": 19000
    },
    {
      "epoch": 0.31,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 0.6113,
      "step": 20000
    },
    {
      "epoch": 0.31,
      "eval_runtime": 1.0179,
      "eval_samples_per_second": 982.441,
      "eval_steps_per_second": 15.719,
      "step": 20000
    },
    {
      "epoch": 0.32,
      "learning_rate": 6.299999999999999e-05,
      "loss": 0.6074,
      "step": 21000
    },
    {
      "epoch": 0.34,
      "learning_rate": 6.599999999999999e-05,
      "loss": 0.6039,
      "step": 22000
    },
    {
      "epoch": 0.35,
      "learning_rate": 6.9e-05,
      "loss": 0.6005,
      "step": 23000
    },
    {
      "epoch": 0.37,
      "learning_rate": 7.199999999999999e-05,
      "loss": 0.5968,
      "step": 24000
    },
    {
      "epoch": 0.38,
      "learning_rate": 7.5e-05,
      "loss": 0.5932,
      "step": 25000
    },
    {
      "epoch": 0.38,
      "eval_runtime": 1.1249,
      "eval_samples_per_second": 888.989,
      "eval_steps_per_second": 14.224,
      "step": 25000
    },
    {
      "epoch": 0.4,
      "learning_rate": 7.8e-05,
      "loss": 0.5912,
      "step": 26000
    },
    {
      "epoch": 0.41,
      "learning_rate": 8.1e-05,
      "loss": 0.58,
      "step": 27000
    },
    {
      "epoch": 0.43,
      "learning_rate": 8.4e-05,
      "loss": 0.5698,
      "step": 28000
    },
    {
      "epoch": 0.44,
      "learning_rate": 8.699999999999999e-05,
      "loss": 0.5639,
      "step": 29000
    },
    {
      "epoch": 0.46,
      "learning_rate": 8.999999999999999e-05,
      "loss": 0.5601,
      "step": 30000
    },
    {
      "epoch": 0.46,
      "eval_runtime": 1.0096,
      "eval_samples_per_second": 990.512,
      "eval_steps_per_second": 15.848,
      "step": 30000
    },
    {
      "epoch": 0.47,
      "learning_rate": 9.3e-05,
      "loss": 0.5536,
      "step": 31000
    },
    {
      "epoch": 0.49,
      "learning_rate": 9.599999999999999e-05,
      "loss": 0.5496,
      "step": 32000
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.9e-05,
      "loss": 0.5458,
      "step": 33000
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.000102,
      "loss": 0.5426,
      "step": 34000
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00010499999999999999,
      "loss": 0.5394,
      "step": 35000
    },
    {
      "epoch": 0.53,
      "eval_runtime": 1.3102,
      "eval_samples_per_second": 763.27,
      "eval_steps_per_second": 12.212,
      "step": 35000
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.00010799999999999998,
      "loss": 0.5345,
      "step": 36000
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00011099999999999999,
      "loss": 0.5302,
      "step": 37000
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00011399999999999999,
      "loss": 0.527,
      "step": 38000
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.000117,
      "loss": 0.5232,
      "step": 39000
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00011999999999999999,
      "loss": 0.5202,
      "step": 40000
    },
    {
      "epoch": 0.61,
      "eval_runtime": 1.0146,
      "eval_samples_per_second": 985.598,
      "eval_steps_per_second": 15.77,
      "step": 40000
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.00012299999999999998,
      "loss": 0.5163,
      "step": 41000
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00012599999999999997,
      "loss": 0.5126,
      "step": 42000
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.000129,
      "loss": 0.5094,
      "step": 43000
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00013199999999999998,
      "loss": 0.5061,
      "step": 44000
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.000135,
      "loss": 0.5036,
      "step": 45000
    },
    {
      "epoch": 0.69,
      "eval_runtime": 1.0362,
      "eval_samples_per_second": 965.067,
      "eval_steps_per_second": 15.441,
      "step": 45000
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.000138,
      "loss": 0.4995,
      "step": 46000
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00014099999999999998,
      "loss": 0.4967,
      "step": 47000
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.00014399999999999998,
      "loss": 0.4934,
      "step": 48000
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.000147,
      "loss": 0.4898,
      "step": 49000
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00015,
      "loss": 0.4863,
      "step": 50000
    },
    {
      "epoch": 0.76,
      "eval_runtime": 1.0374,
      "eval_samples_per_second": 963.96,
      "eval_steps_per_second": 15.423,
      "step": 50000
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0001499996172456075,
      "loss": 0.4824,
      "step": 51000
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00014999846898661572,
      "loss": 0.4778,
      "step": 52000
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.00014999655523558183,
      "loss": 0.474,
      "step": 53000
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00014999387601343436,
      "loss": 0.4694,
      "step": 54000
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00014999043134947282,
      "loss": 0.4651,
      "step": 55000
    },
    {
      "epoch": 0.84,
      "eval_runtime": 1.0465,
      "eval_samples_per_second": 955.566,
      "eval_steps_per_second": 15.289,
      "step": 55000
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00014998622128136748,
      "loss": 0.4608,
      "step": 56000
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.000149981245855159,
      "loss": 0.4566,
      "step": 57000
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00014997550512525784,
      "loss": 0.4523,
      "step": 58000
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0001499689991544437,
      "loss": 0.4483,
      "step": 59000
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00014996172801386482,
      "loss": 0.4447,
      "step": 60000
    },
    {
      "epoch": 0.92,
      "eval_runtime": 1.2288,
      "eval_samples_per_second": 813.826,
      "eval_steps_per_second": 13.021,
      "step": 60000
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00014995369178303722,
      "loss": 0.4408,
      "step": 61000
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0001499448905498439,
      "loss": 0.4381,
      "step": 62000
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00014993532441053364,
      "loss": 0.434,
      "step": 63000
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0001499249934697203,
      "loss": 0.4316,
      "step": 64000
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0001499138978403813,
      "loss": 0.4275,
      "step": 65000
    },
    {
      "epoch": 0.99,
      "eval_runtime": 1.0345,
      "eval_samples_per_second": 966.655,
      "eval_steps_per_second": 15.466,
      "step": 65000
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00014990203764385677,
      "loss": 0.425,
      "step": 66000
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00014988941300984784,
      "loss": 0.422,
      "step": 67000
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0001498760240764155,
      "loss": 0.4191,
      "step": 68000
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.000149861870989979,
      "loss": 0.4164,
      "step": 69000
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0001498469539053142,
      "loss": 0.4138,
      "step": 70000
    },
    {
      "epoch": 1.07,
      "eval_runtime": 1.1341,
      "eval_samples_per_second": 881.784,
      "eval_steps_per_second": 14.109,
      "step": 70000
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00014983127298555198,
      "loss": 0.4114,
      "step": 71000
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00014981482840217632,
      "loss": 0.4086,
      "step": 72000
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.00014979762033502262,
      "loss": 0.4066,
      "step": 73000
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.00014977964897227547,
      "loss": 0.4042,
      "step": 74000
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00014976091451046687,
      "loss": 0.402,
      "step": 75000
    },
    {
      "epoch": 1.15,
      "eval_runtime": 1.0331,
      "eval_samples_per_second": 967.957,
      "eval_steps_per_second": 15.487,
      "step": 75000
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00014974141715447386,
      "loss": 0.3999,
      "step": 76000
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00014972115711751644,
      "loss": 0.398,
      "step": 77000
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00014970013462115505,
      "loss": 0.3971,
      "step": 78000
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.00014967834989528843,
      "loss": 0.3942,
      "step": 79000
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.00014965580317815078,
      "loss": 0.3926,
      "step": 80000
    },
    {
      "epoch": 1.22,
      "eval_runtime": 1.084,
      "eval_samples_per_second": 922.521,
      "eval_steps_per_second": 14.76,
      "step": 80000
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00014963249471630944,
      "loss": 0.3906,
      "step": 81000
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.000149608424764662,
      "loss": 0.391,
      "step": 82000
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0001495835935864336,
      "loss": 0.3875,
      "step": 83000
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00014955800145317397,
      "loss": 0.3861,
      "step": 84000
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.00014953164864475466,
      "loss": 0.3844,
      "step": 85000
    },
    {
      "epoch": 1.3,
      "eval_runtime": 1.0992,
      "eval_samples_per_second": 909.734,
      "eval_steps_per_second": 14.556,
      "step": 85000
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0001495045354493657,
      "loss": 0.3829,
      "step": 86000
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00014947666216351272,
      "loss": 0.3815,
      "step": 87000
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00014944802909201344,
      "loss": 0.38,
      "step": 88000
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.00014941863654799456,
      "loss": 0.3789,
      "step": 89000
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.00014938848485288825,
      "loss": 0.3785,
      "step": 90000
    },
    {
      "epoch": 1.37,
      "eval_runtime": 0.9266,
      "eval_samples_per_second": 1079.167,
      "eval_steps_per_second": 17.267,
      "step": 90000
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0001493575743364286,
      "loss": 0.3766,
      "step": 91000
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00014932590533664808,
      "loss": 0.3745,
      "step": 92000
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0001492934781998738,
      "loss": 0.3741,
      "step": 93000
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0001492602932807237,
      "loss": 0.3729,
      "step": 94000
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.00014922635094210277,
      "loss": 0.3709,
      "step": 95000
    },
    {
      "epoch": 1.45,
      "eval_runtime": 0.9895,
      "eval_samples_per_second": 1010.579,
      "eval_steps_per_second": 16.169,
      "step": 95000
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.000149191651555199,
      "loss": 0.3699,
      "step": 96000
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0001491561954994793,
      "loss": 0.3688,
      "step": 97000
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.00014911998316268537,
      "loss": 0.3678,
      "step": 98000
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00014908301494082963,
      "loss": 0.3666,
      "step": 99000
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.00014904529123819054,
      "loss": 0.3654,
      "step": 100000
    },
    {
      "epoch": 1.53,
      "eval_runtime": 1.0046,
      "eval_samples_per_second": 995.424,
      "eval_steps_per_second": 15.927,
      "step": 100000
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.00014900681246730852,
      "loss": 0.3643,
      "step": 101000
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00014896757904898125,
      "loss": 0.3646,
      "step": 102000
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00014892759141225904,
      "loss": 0.3628,
      "step": 103000
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00014888684999444035,
      "loss": 0.3616,
      "step": 104000
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00014884535524106675,
      "loss": 0.3604,
      "step": 105000
    },
    {
      "epoch": 1.6,
      "eval_runtime": 1.0499,
      "eval_samples_per_second": 952.499,
      "eval_steps_per_second": 15.24,
      "step": 105000
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00014880310760591824,
      "loss": 0.3594,
      "step": 106000
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0001487601075510082,
      "loss": 0.3597,
      "step": 107000
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0001487163555465783,
      "loss": 0.3583,
      "step": 108000
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0001486718520710935,
      "loss": 0.3583,
      "step": 109000
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00014862659761123663,
      "loss": 0.3558,
      "step": 110000
    },
    {
      "epoch": 1.68,
      "eval_runtime": 1.0153,
      "eval_samples_per_second": 984.91,
      "eval_steps_per_second": 15.759,
      "step": 110000
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00014858059266190327,
      "loss": 0.3552,
      "step": 111000
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00014853383772619612,
      "loss": 0.3544,
      "step": 112000
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.00014848633331541967,
      "loss": 0.3537,
      "step": 113000
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0001484380799490746,
      "loss": 0.3524,
      "step": 114000
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00014838907815485194,
      "loss": 0.3519,
      "step": 115000
    },
    {
      "epoch": 1.76,
      "eval_runtime": 1.003,
      "eval_samples_per_second": 997.001,
      "eval_steps_per_second": 15.952,
      "step": 115000
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.00014833932846862748,
      "loss": 0.3511,
      "step": 116000
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00014828883143445582,
      "loss": 0.3502,
      "step": 117000
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0001482375876045644,
      "loss": 0.3493,
      "step": 118000
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0001481855975393476,
      "loss": 0.3489,
      "step": 119000
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0001481328618073604,
      "loss": 0.3482,
      "step": 120000
    },
    {
      "epoch": 1.83,
      "eval_runtime": 1.0596,
      "eval_samples_per_second": 943.744,
      "eval_steps_per_second": 15.1,
      "step": 120000
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0001480793809853123,
      "loss": 0.3478,
      "step": 121000
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00014802515565806107,
      "loss": 0.3468,
      "step": 122000
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00014797018641860612,
      "loss": 0.346,
      "step": 123000
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0001479144738680823,
      "loss": 0.3474,
      "step": 124000
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00014785801861575312,
      "loss": 0.3447,
      "step": 125000
    },
    {
      "epoch": 1.91,
      "eval_runtime": 0.9375,
      "eval_samples_per_second": 1066.699,
      "eval_steps_per_second": 17.067,
      "step": 125000
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.00014780082127900416,
      "loss": 0.3439,
      "step": 126000
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00014774288248333635,
      "loss": 0.3436,
      "step": 127000
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00014768420286235908,
      "loss": 0.3429,
      "step": 128000
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00014762478305778328,
      "loss": 0.3422,
      "step": 129000
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0001475646237194144,
      "loss": 0.3414,
      "step": 130000
    },
    {
      "epoch": 1.99,
      "eval_runtime": 1.0085,
      "eval_samples_per_second": 991.553,
      "eval_steps_per_second": 15.865,
      "step": 130000
    }
  ],
  "max_steps": 1000000,
  "num_train_epochs": 16,
  "total_flos": 9.113027363263641e+21,
  "trial_name": null,
  "trial_params": null
}