{
  "best_metric": 1.0027852058410645,
  "best_model_checkpoint": "/kaggle/output/checkpoint-40000",
  "epoch": 1.6297262059973925,
  "eval_steps": 1000,
  "global_step": 40000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 2.7777777777777777e-11,
      "loss": 1.0788,
      "step": 1
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.7750000000000004e-08,
      "loss": 1.2063,
      "step": 1000
    },
    {
      "epoch": 0.04,
      "eval_accuracy": 0.32894211576846305,
      "eval_loss": 1.109653115272522,
      "eval_runtime": 50.6057,
      "eval_samples_per_second": 99.001,
      "eval_steps_per_second": 12.39,
      "step": 1000
    },
    {
      "epoch": 0.08,
      "learning_rate": 5.5527777777777784e-08,
      "loss": 1.1293,
      "step": 2000
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.32994011976047904,
      "eval_loss": 1.1056402921676636,
      "eval_runtime": 50.7169,
      "eval_samples_per_second": 98.784,
      "eval_steps_per_second": 12.363,
      "step": 2000
    },
    {
      "epoch": 0.12,
      "learning_rate": 8.327777777777778e-08,
      "loss": 1.1263,
      "step": 3000
    },
    {
      "epoch": 0.12,
      "eval_accuracy": 0.3347305389221557,
      "eval_loss": 1.1063188314437866,
      "eval_runtime": 50.8081,
      "eval_samples_per_second": 98.606,
      "eval_steps_per_second": 12.341,
      "step": 3000
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.1105555555555557e-07,
      "loss": 1.1211,
      "step": 4000
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.34291417165668664,
      "eval_loss": 1.1010773181915283,
      "eval_runtime": 50.6469,
      "eval_samples_per_second": 98.92,
      "eval_steps_per_second": 12.38,
      "step": 4000
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.3883333333333335e-07,
      "loss": 1.1176,
      "step": 5000
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.3395209580838323,
      "eval_loss": 1.099412441253662,
      "eval_runtime": 50.7266,
      "eval_samples_per_second": 98.765,
      "eval_steps_per_second": 12.36,
      "step": 5000
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.6658333333333335e-07,
      "loss": 1.1165,
      "step": 6000
    },
    {
      "epoch": 0.24,
      "eval_accuracy": 0.3596806387225549,
      "eval_loss": 1.0983413457870483,
      "eval_runtime": 50.7493,
      "eval_samples_per_second": 98.721,
      "eval_steps_per_second": 12.355,
      "step": 6000
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.9436111111111112e-07,
      "loss": 1.1127,
      "step": 7000
    },
    {
      "epoch": 0.29,
      "eval_accuracy": 0.35768463073852297,
      "eval_loss": 1.094804048538208,
      "eval_runtime": 50.8322,
      "eval_samples_per_second": 98.56,
      "eval_steps_per_second": 12.335,
      "step": 7000
    },
    {
      "epoch": 0.33,
      "learning_rate": 2.2213888888888891e-07,
      "loss": 1.1095,
      "step": 8000
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.36407185628742517,
      "eval_loss": 1.0937609672546387,
      "eval_runtime": 50.8609,
      "eval_samples_per_second": 98.504,
      "eval_steps_per_second": 12.328,
      "step": 8000
    },
    {
      "epoch": 0.37,
      "learning_rate": 2.499166666666667e-07,
      "loss": 1.1074,
      "step": 9000
    },
    {
      "epoch": 0.37,
      "eval_accuracy": 0.34770459081836325,
      "eval_loss": 1.1025313138961792,
      "eval_runtime": 50.8052,
      "eval_samples_per_second": 98.612,
      "eval_steps_per_second": 12.341,
      "step": 9000
    },
    {
      "epoch": 0.41,
      "learning_rate": 2.776666666666667e-07,
      "loss": 1.1065,
      "step": 10000
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.38163672654690617,
      "eval_loss": 1.0920748710632324,
      "eval_runtime": 50.8964,
      "eval_samples_per_second": 98.435,
      "eval_steps_per_second": 12.319,
      "step": 10000
    },
    {
      "epoch": 0.45,
      "learning_rate": 3.054444444444444e-07,
      "loss": 1.1026,
      "step": 11000
    },
    {
      "epoch": 0.45,
      "eval_accuracy": 0.37944111776447104,
      "eval_loss": 1.0899525880813599,
      "eval_runtime": 50.6905,
      "eval_samples_per_second": 98.835,
      "eval_steps_per_second": 12.369,
      "step": 11000
    },
    {
      "epoch": 0.49,
      "learning_rate": 3.3319444444444444e-07,
      "loss": 1.1016,
      "step": 12000
    },
    {
      "epoch": 0.49,
      "eval_accuracy": 0.4,
      "eval_loss": 1.0867687463760376,
      "eval_runtime": 50.8282,
      "eval_samples_per_second": 98.567,
      "eval_steps_per_second": 12.336,
      "step": 12000
    },
    {
      "epoch": 0.53,
      "learning_rate": 3.6094444444444446e-07,
      "loss": 1.0999,
      "step": 13000
    },
    {
      "epoch": 0.53,
      "eval_accuracy": 0.3467065868263473,
      "eval_loss": 1.0919947624206543,
      "eval_runtime": 50.7953,
      "eval_samples_per_second": 98.631,
      "eval_steps_per_second": 12.344,
      "step": 13000
    },
    {
      "epoch": 0.57,
      "learning_rate": 3.8872222222222223e-07,
      "loss": 1.0976,
      "step": 14000
    },
    {
      "epoch": 0.57,
      "eval_accuracy": 0.38003992015968063,
      "eval_loss": 1.0882444381713867,
      "eval_runtime": 50.8866,
      "eval_samples_per_second": 98.454,
      "eval_steps_per_second": 12.322,
      "step": 14000
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.1650000000000006e-07,
      "loss": 1.0951,
      "step": 15000
    },
    {
      "epoch": 0.61,
      "eval_accuracy": 0.4229540918163673,
      "eval_loss": 1.0766451358795166,
      "eval_runtime": 50.7597,
      "eval_samples_per_second": 98.7,
      "eval_steps_per_second": 12.352,
      "step": 15000
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.4425e-07,
      "loss": 1.087,
      "step": 16000
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.4339321357285429,
      "eval_loss": 1.0621711015701294,
      "eval_runtime": 50.7559,
      "eval_samples_per_second": 98.708,
      "eval_steps_per_second": 12.353,
      "step": 16000
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.7202777777777785e-07,
      "loss": 1.0769,
      "step": 17000
    },
    {
      "epoch": 0.69,
      "eval_accuracy": 0.4357285429141717,
      "eval_loss": 1.0624364614486694,
      "eval_runtime": 50.8608,
      "eval_samples_per_second": 98.504,
      "eval_steps_per_second": 12.328,
      "step": 17000
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.998055555555556e-07,
      "loss": 1.0798,
      "step": 18000
    },
    {
      "epoch": 0.73,
      "eval_accuracy": 0.4311377245508982,
      "eval_loss": 1.0611094236373901,
      "eval_runtime": 50.7255,
      "eval_samples_per_second": 98.767,
      "eval_steps_per_second": 12.361,
      "step": 18000
    },
    {
      "epoch": 0.77,
      "learning_rate": 5.275555555555556e-07,
      "loss": 1.0717,
      "step": 19000
    },
    {
      "epoch": 0.77,
      "eval_accuracy": 0.43852295409181635,
      "eval_loss": 1.056412935256958,
      "eval_runtime": 50.8955,
      "eval_samples_per_second": 98.437,
      "eval_steps_per_second": 12.319,
      "step": 19000
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.553333333333334e-07,
      "loss": 1.0682,
      "step": 20000
    },
    {
      "epoch": 0.81,
      "eval_accuracy": 0.43313373253493015,
      "eval_loss": 1.0576339960098267,
      "eval_runtime": 50.6675,
      "eval_samples_per_second": 98.88,
      "eval_steps_per_second": 12.375,
      "step": 20000
    },
    {
      "epoch": 0.86,
      "learning_rate": 5.830833333333334e-07,
      "loss": 1.0674,
      "step": 21000
    },
    {
      "epoch": 0.86,
      "eval_accuracy": 0.4305389221556886,
      "eval_loss": 1.057605504989624,
      "eval_runtime": 50.8082,
      "eval_samples_per_second": 98.606,
      "eval_steps_per_second": 12.341,
      "step": 21000
    },
    {
      "epoch": 0.9,
      "learning_rate": 6.108611111111111e-07,
      "loss": 1.0689,
      "step": 22000
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.4351297405189621,
      "eval_loss": 1.0509696006774902,
      "eval_runtime": 50.7595,
      "eval_samples_per_second": 98.701,
      "eval_steps_per_second": 12.352,
      "step": 22000
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.386388888888889e-07,
      "loss": 1.065,
      "step": 23000
    },
    {
      "epoch": 0.94,
      "eval_accuracy": 0.4363273453093812,
      "eval_loss": 1.0513038635253906,
      "eval_runtime": 50.685,
      "eval_samples_per_second": 98.846,
      "eval_steps_per_second": 12.371,
      "step": 23000
    },
    {
      "epoch": 0.98,
      "learning_rate": 6.663888888888889e-07,
      "loss": 1.0638,
      "step": 24000
    },
    {
      "epoch": 0.98,
      "eval_accuracy": 0.4429141716566866,
      "eval_loss": 1.04515540599823,
      "eval_runtime": 50.9215,
      "eval_samples_per_second": 98.387,
      "eval_steps_per_second": 12.313,
      "step": 24000
    },
    {
      "epoch": 1.02,
      "learning_rate": 6.941666666666667e-07,
      "loss": 1.0607,
      "step": 25000
    },
    {
      "epoch": 1.02,
      "eval_accuracy": 0.4471057884231537,
      "eval_loss": 1.0432593822479248,
      "eval_runtime": 50.8338,
      "eval_samples_per_second": 98.556,
      "eval_steps_per_second": 12.334,
      "step": 25000
    },
    {
      "epoch": 1.06,
      "learning_rate": 7.219444444444444e-07,
      "loss": 1.0578,
      "step": 26000
    },
    {
      "epoch": 1.06,
      "eval_accuracy": 0.4407185628742515,
      "eval_loss": 1.044198751449585,
      "eval_runtime": 50.8791,
      "eval_samples_per_second": 98.469,
      "eval_steps_per_second": 12.323,
      "step": 26000
    },
    {
      "epoch": 1.1,
      "learning_rate": 7.496944444444444e-07,
      "loss": 1.0602,
      "step": 27000
    },
    {
      "epoch": 1.1,
      "eval_accuracy": 0.4363273453093812,
      "eval_loss": 1.045542597770691,
      "eval_runtime": 50.8933,
      "eval_samples_per_second": 98.441,
      "eval_steps_per_second": 12.32,
      "step": 27000
    },
    {
      "epoch": 1.14,
      "learning_rate": 7.774722222222223e-07,
      "loss": 1.0566,
      "step": 28000
    },
    {
      "epoch": 1.14,
      "eval_accuracy": 0.437125748502994,
      "eval_loss": 1.038167953491211,
      "eval_runtime": 50.8894,
      "eval_samples_per_second": 98.449,
      "eval_steps_per_second": 12.321,
      "step": 28000
    },
    {
      "epoch": 1.18,
      "learning_rate": 8.052222222222223e-07,
      "loss": 1.0528,
      "step": 29000
    },
    {
      "epoch": 1.18,
      "eval_accuracy": 0.45069860279441115,
      "eval_loss": 1.0337190628051758,
      "eval_runtime": 50.8393,
      "eval_samples_per_second": 98.546,
      "eval_steps_per_second": 12.333,
      "step": 29000
    },
    {
      "epoch": 1.22,
      "learning_rate": 8.330000000000001e-07,
      "loss": 1.0507,
      "step": 30000
    },
    {
      "epoch": 1.22,
      "eval_accuracy": 0.4497005988023952,
      "eval_loss": 1.0287295579910278,
      "eval_runtime": 50.8361,
      "eval_samples_per_second": 98.552,
      "eval_steps_per_second": 12.334,
      "step": 30000
    },
    {
      "epoch": 1.26,
      "learning_rate": 8.607777777777779e-07,
      "loss": 1.0553,
      "step": 31000
    },
    {
      "epoch": 1.26,
      "eval_accuracy": 0.45209580838323354,
      "eval_loss": 1.0318636894226074,
      "eval_runtime": 50.8224,
      "eval_samples_per_second": 98.578,
      "eval_steps_per_second": 12.337,
      "step": 31000
    },
    {
      "epoch": 1.3,
      "learning_rate": 8.885277777777779e-07,
      "loss": 1.0509,
      "step": 32000
    },
    {
      "epoch": 1.3,
      "eval_accuracy": 0.4530938123752495,
      "eval_loss": 1.0215647220611572,
      "eval_runtime": 50.8903,
      "eval_samples_per_second": 98.447,
      "eval_steps_per_second": 12.321,
      "step": 32000
    },
    {
      "epoch": 1.34,
      "learning_rate": 9.163055555555556e-07,
      "loss": 1.0444,
      "step": 33000
    },
    {
      "epoch": 1.34,
      "eval_accuracy": 0.47025948103792414,
      "eval_loss": 1.0156230926513672,
      "eval_runtime": 50.8879,
      "eval_samples_per_second": 98.452,
      "eval_steps_per_second": 12.321,
      "step": 33000
    },
    {
      "epoch": 1.39,
      "learning_rate": 9.440833333333335e-07,
      "loss": 1.0383,
      "step": 34000
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.4686626746506986,
      "eval_loss": 1.0187312364578247,
      "eval_runtime": 50.7712,
      "eval_samples_per_second": 98.678,
      "eval_steps_per_second": 12.35,
      "step": 34000
    },
    {
      "epoch": 1.43,
      "learning_rate": 9.718333333333334e-07,
      "loss": 1.043,
      "step": 35000
    },
    {
      "epoch": 1.43,
      "eval_accuracy": 0.4748502994011976,
      "eval_loss": 1.0167384147644043,
      "eval_runtime": 50.6223,
      "eval_samples_per_second": 98.968,
      "eval_steps_per_second": 12.386,
      "step": 35000
    },
    {
      "epoch": 1.47,
      "learning_rate": 9.996111111111111e-07,
      "loss": 1.0396,
      "step": 36000
    },
    {
      "epoch": 1.47,
      "eval_accuracy": 0.47964071856287427,
      "eval_loss": 1.0082147121429443,
      "eval_runtime": 50.671,
      "eval_samples_per_second": 98.873,
      "eval_steps_per_second": 12.374,
      "step": 36000
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.0273888888888891e-06,
      "loss": 1.0373,
      "step": 37000
    },
    {
      "epoch": 1.51,
      "eval_accuracy": 0.4802395209580838,
      "eval_loss": 1.0071511268615723,
      "eval_runtime": 50.5903,
      "eval_samples_per_second": 99.031,
      "eval_steps_per_second": 12.394,
      "step": 37000
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.0551666666666669e-06,
      "loss": 1.0371,
      "step": 38000
    },
    {
      "epoch": 1.55,
      "eval_accuracy": 0.49481037924151694,
      "eval_loss": 1.0086288452148438,
      "eval_runtime": 50.7157,
      "eval_samples_per_second": 98.786,
      "eval_steps_per_second": 12.363,
      "step": 38000
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.082888888888889e-06,
      "loss": 1.0367,
      "step": 39000
    },
    {
      "epoch": 1.59,
      "eval_accuracy": 0.4884231536926148,
      "eval_loss": 1.0103219747543335,
      "eval_runtime": 50.6401,
      "eval_samples_per_second": 98.934,
      "eval_steps_per_second": 12.382,
      "step": 39000
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.1106666666666668e-06,
      "loss": 1.0311,
      "step": 40000
    },
    {
      "epoch": 1.63,
      "eval_accuracy": 0.4846307385229541,
      "eval_loss": 1.0027852058410645,
      "eval_runtime": 50.6939,
      "eval_samples_per_second": 98.828,
      "eval_steps_per_second": 12.368,
      "step": 40000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 10000000,
  "num_train_epochs": 408,
  "save_steps": 1000,
  "total_flos": 8.36139439203287e+16,
  "trial_name": null,
  "trial_params": null
}