{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 1,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "grad_norm": 36.0623664855957,
      "learning_rate": 2.5e-05,
      "loss": 1.4971,
      "step": 1
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.52,
      "eval_loss": 1.3626638650894165,
      "eval_runtime": 1.0371,
      "eval_samples_per_second": 96.426,
      "eval_steps_per_second": 2.893,
      "step": 1
    },
    {
      "epoch": 0.4,
      "grad_norm": 35.2884407043457,
      "learning_rate": 5e-05,
      "loss": 1.5356,
      "step": 2
    },
    {
      "epoch": 0.4,
      "eval_accuracy": 0.52,
      "eval_loss": 1.2972925901412964,
      "eval_runtime": 1.0426,
      "eval_samples_per_second": 95.917,
      "eval_steps_per_second": 2.878,
      "step": 2
    },
    {
      "epoch": 0.6,
      "grad_norm": 40.47761154174805,
      "learning_rate": 4.8958333333333335e-05,
      "loss": 1.5671,
      "step": 3
    },
    {
      "epoch": 0.6,
      "eval_accuracy": 0.52,
      "eval_loss": 1.121543049812317,
      "eval_runtime": 1.0354,
      "eval_samples_per_second": 96.579,
      "eval_steps_per_second": 2.897,
      "step": 3
    },
    {
      "epoch": 0.8,
      "grad_norm": 29.40814208984375,
      "learning_rate": 4.791666666666667e-05,
      "loss": 1.1053,
      "step": 4
    },
    {
      "epoch": 0.8,
      "eval_accuracy": 0.52,
      "eval_loss": 0.9736524820327759,
      "eval_runtime": 1.0346,
      "eval_samples_per_second": 96.655,
      "eval_steps_per_second": 2.9,
      "step": 4
    },
    {
      "epoch": 1.0,
      "grad_norm": 32.18461227416992,
      "learning_rate": 4.6875e-05,
      "loss": 1.1256,
      "step": 5
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.52,
      "eval_loss": 0.8558984398841858,
      "eval_runtime": 1.0357,
      "eval_samples_per_second": 96.556,
      "eval_steps_per_second": 2.897,
      "step": 5
    },
    {
      "epoch": 1.2,
      "grad_norm": 16.461198806762695,
      "learning_rate": 4.5833333333333334e-05,
      "loss": 0.828,
      "step": 6
    },
    {
      "epoch": 1.2,
      "eval_accuracy": 0.53,
      "eval_loss": 0.7607324123382568,
      "eval_runtime": 1.036,
      "eval_samples_per_second": 96.527,
      "eval_steps_per_second": 2.896,
      "step": 6
    },
    {
      "epoch": 1.4,
      "grad_norm": 8.941577911376953,
      "learning_rate": 4.4791666666666673e-05,
      "loss": 0.7714,
      "step": 7
    },
    {
      "epoch": 1.4,
      "eval_accuracy": 0.5,
      "eval_loss": 0.7198632955551147,
      "eval_runtime": 1.0366,
      "eval_samples_per_second": 96.473,
      "eval_steps_per_second": 2.894,
      "step": 7
    },
    {
      "epoch": 1.6,
      "grad_norm": 16.71706771850586,
      "learning_rate": 4.375e-05,
      "loss": 0.7838,
      "step": 8
    },
    {
      "epoch": 1.6,
      "eval_accuracy": 0.44,
      "eval_loss": 0.7371875047683716,
      "eval_runtime": 1.0343,
      "eval_samples_per_second": 96.684,
      "eval_steps_per_second": 2.901,
      "step": 8
    },
    {
      "epoch": 1.8,
      "grad_norm": 5.3311238288879395,
      "learning_rate": 4.270833333333333e-05,
      "loss": 0.7412,
      "step": 9
    },
    {
      "epoch": 1.8,
      "eval_accuracy": 0.46,
      "eval_loss": 0.7655664682388306,
      "eval_runtime": 1.0353,
      "eval_samples_per_second": 96.592,
      "eval_steps_per_second": 2.898,
      "step": 9
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.812692642211914,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.741,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.47,
      "eval_loss": 0.7885058522224426,
      "eval_runtime": 1.0364,
      "eval_samples_per_second": 96.49,
      "eval_steps_per_second": 2.895,
      "step": 10
    },
    {
      "epoch": 2.2,
      "grad_norm": 19.197566986083984,
      "learning_rate": 4.0625000000000005e-05,
      "loss": 0.8213,
      "step": 11
    },
    {
      "epoch": 2.2,
      "eval_accuracy": 0.47,
      "eval_loss": 0.7977539300918579,
      "eval_runtime": 1.0399,
      "eval_samples_per_second": 96.162,
      "eval_steps_per_second": 2.885,
      "step": 11
    },
    {
      "epoch": 2.4,
      "grad_norm": 6.057560443878174,
      "learning_rate": 3.958333333333333e-05,
      "loss": 0.676,
      "step": 12
    },
    {
      "epoch": 2.4,
      "eval_accuracy": 0.47,
      "eval_loss": 0.7945410013198853,
      "eval_runtime": 1.0431,
      "eval_samples_per_second": 95.872,
      "eval_steps_per_second": 2.876,
      "step": 12
    },
    {
      "epoch": 2.6,
      "grad_norm": 7.396554470062256,
      "learning_rate": 3.854166666666667e-05,
      "loss": 0.7126,
      "step": 13
    },
    {
      "epoch": 2.6,
      "eval_accuracy": 0.47,
      "eval_loss": 0.7799025177955627,
      "eval_runtime": 1.0445,
      "eval_samples_per_second": 95.742,
      "eval_steps_per_second": 2.872,
      "step": 13
    },
    {
      "epoch": 2.8,
      "grad_norm": 2.042255163192749,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.7063,
      "step": 14
    },
    {
      "epoch": 2.8,
      "eval_accuracy": 0.47,
      "eval_loss": 0.7568800449371338,
      "eval_runtime": 1.0397,
      "eval_samples_per_second": 96.179,
      "eval_steps_per_second": 2.885,
      "step": 14
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.283554792404175,
      "learning_rate": 3.6458333333333336e-05,
      "loss": 0.6571,
      "step": 15
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.48,
      "eval_loss": 0.7332909107208252,
      "eval_runtime": 1.0437,
      "eval_samples_per_second": 95.812,
      "eval_steps_per_second": 2.874,
      "step": 15
    },
    {
      "epoch": 3.2,
      "grad_norm": 5.649877548217773,
      "learning_rate": 3.541666666666667e-05,
      "loss": 0.6836,
      "step": 16
    },
    {
      "epoch": 3.2,
      "eval_accuracy": 0.49,
      "eval_loss": 0.7175489664077759,
      "eval_runtime": 1.0384,
      "eval_samples_per_second": 96.304,
      "eval_steps_per_second": 2.889,
      "step": 16
    },
    {
      "epoch": 3.4,
      "grad_norm": 10.291581153869629,
      "learning_rate": 3.4375e-05,
      "loss": 0.7174,
      "step": 17
    },
    {
      "epoch": 3.4,
      "eval_accuracy": 0.5,
      "eval_loss": 0.710507869720459,
      "eval_runtime": 1.0395,
      "eval_samples_per_second": 96.202,
      "eval_steps_per_second": 2.886,
      "step": 17
    },
    {
      "epoch": 3.6,
      "grad_norm": 9.05163860321045,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.741,
      "step": 18
    },
    {
      "epoch": 3.6,
      "eval_accuracy": 0.54,
      "eval_loss": 0.7027149796485901,
      "eval_runtime": 1.0393,
      "eval_samples_per_second": 96.222,
      "eval_steps_per_second": 2.887,
      "step": 18
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.545870304107666,
      "learning_rate": 3.229166666666667e-05,
      "loss": 0.7073,
      "step": 19
    },
    {
      "epoch": 3.8,
      "eval_accuracy": 0.52,
      "eval_loss": 0.7002732753753662,
      "eval_runtime": 1.0416,
      "eval_samples_per_second": 96.007,
      "eval_steps_per_second": 2.88,
      "step": 19
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.6787610054016113,
      "learning_rate": 3.125e-05,
      "loss": 0.7142,
      "step": 20
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.55,
      "eval_loss": 0.6979882717132568,
      "eval_runtime": 1.0399,
      "eval_samples_per_second": 96.16,
      "eval_steps_per_second": 2.885,
      "step": 20
    },
    {
      "epoch": 4.2,
      "grad_norm": 4.416342258453369,
      "learning_rate": 3.0208333333333334e-05,
      "loss": 0.6872,
      "step": 21
    },
    {
      "epoch": 4.2,
      "eval_accuracy": 0.55,
      "eval_loss": 0.6990429759025574,
      "eval_runtime": 1.0375,
      "eval_samples_per_second": 96.382,
      "eval_steps_per_second": 2.891,
      "step": 21
    },
    {
      "epoch": 4.4,
      "grad_norm": 9.770723342895508,
      "learning_rate": 2.916666666666667e-05,
      "loss": 0.7075,
      "step": 22
    },
    {
      "epoch": 4.4,
      "eval_accuracy": 0.55,
      "eval_loss": 0.7003124952316284,
      "eval_runtime": 1.0384,
      "eval_samples_per_second": 96.298,
      "eval_steps_per_second": 2.889,
      "step": 22
    },
    {
      "epoch": 4.6,
      "grad_norm": 5.248044013977051,
      "learning_rate": 2.8125000000000003e-05,
      "loss": 0.6618,
      "step": 23
    },
    {
      "epoch": 4.6,
      "eval_accuracy": 0.55,
      "eval_loss": 0.6986132860183716,
      "eval_runtime": 1.0403,
      "eval_samples_per_second": 96.123,
      "eval_steps_per_second": 2.884,
      "step": 23
    },
    {
      "epoch": 4.8,
      "grad_norm": 3.608055353164673,
      "learning_rate": 2.7083333333333332e-05,
      "loss": 0.6821,
      "step": 24
    },
    {
      "epoch": 4.8,
      "eval_accuracy": 0.54,
      "eval_loss": 0.6961914300918579,
      "eval_runtime": 1.04,
      "eval_samples_per_second": 96.158,
      "eval_steps_per_second": 2.885,
      "step": 24
    },
    {
      "epoch": 5.0,
      "grad_norm": 8.092246055603027,
      "learning_rate": 2.604166666666667e-05,
      "loss": 0.7073,
      "step": 25
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.54,
      "eval_loss": 0.6970117092132568,
      "eval_runtime": 1.0439,
      "eval_samples_per_second": 95.792,
      "eval_steps_per_second": 2.874,
      "step": 25
    },
    {
      "epoch": 5.2,
      "grad_norm": 6.533432483673096,
      "learning_rate": 2.5e-05,
      "loss": 0.7104,
      "step": 26
    },
    {
      "epoch": 5.2,
      "eval_accuracy": 0.49,
      "eval_loss": 0.6976756453514099,
      "eval_runtime": 1.0385,
      "eval_samples_per_second": 96.296,
      "eval_steps_per_second": 2.889,
      "step": 26
    },
    {
      "epoch": 5.4,
      "grad_norm": 3.886502742767334,
      "learning_rate": 2.3958333333333334e-05,
      "loss": 0.6792,
      "step": 27
    },
    {
      "epoch": 5.4,
      "eval_accuracy": 0.51,
      "eval_loss": 0.7020704746246338,
      "eval_runtime": 1.037,
      "eval_samples_per_second": 96.43,
      "eval_steps_per_second": 2.893,
      "step": 27
    },
    {
      "epoch": 5.6,
      "grad_norm": 9.063314437866211,
      "learning_rate": 2.2916666666666667e-05,
      "loss": 0.666,
      "step": 28
    },
    {
      "epoch": 5.6,
      "eval_accuracy": 0.5,
      "eval_loss": 0.7069726586341858,
      "eval_runtime": 1.0397,
      "eval_samples_per_second": 96.181,
      "eval_steps_per_second": 2.885,
      "step": 28
    },
    {
      "epoch": 5.8,
      "grad_norm": 3.9558804035186768,
      "learning_rate": 2.1875e-05,
      "loss": 0.7274,
      "step": 29
    },
    {
      "epoch": 5.8,
      "eval_accuracy": 0.45,
      "eval_loss": 0.7108399271965027,
      "eval_runtime": 1.0373,
      "eval_samples_per_second": 96.402,
      "eval_steps_per_second": 2.892,
      "step": 29
    },
    {
      "epoch": 6.0,
      "grad_norm": 7.0909600257873535,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.6817,
      "step": 30
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.48,
      "eval_loss": 0.7125291228294373,
      "eval_runtime": 1.041,
      "eval_samples_per_second": 96.059,
      "eval_steps_per_second": 2.882,
      "step": 30
    },
    {
      "epoch": 6.2,
      "grad_norm": 2.2684037685394287,
      "learning_rate": 1.9791666666666665e-05,
      "loss": 0.7132,
      "step": 31
    },
    {
      "epoch": 6.2,
      "eval_accuracy": 0.49,
      "eval_loss": 0.7147753834724426,
      "eval_runtime": 1.0438,
      "eval_samples_per_second": 95.804,
      "eval_steps_per_second": 2.874,
      "step": 31
    },
    {
      "epoch": 6.4,
      "grad_norm": 3.9249820709228516,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.6927,
      "step": 32
    },
    {
      "epoch": 6.4,
      "eval_accuracy": 0.48,
      "eval_loss": 0.7158203125,
      "eval_runtime": 1.0439,
      "eval_samples_per_second": 95.792,
      "eval_steps_per_second": 2.874,
      "step": 32
    },
    {
      "epoch": 6.6,
      "grad_norm": 2.211745500564575,
      "learning_rate": 1.7708333333333335e-05,
      "loss": 0.6594,
      "step": 33
    },
    {
      "epoch": 6.6,
      "eval_accuracy": 0.5,
      "eval_loss": 0.7170214653015137,
      "eval_runtime": 1.0378,
      "eval_samples_per_second": 96.354,
      "eval_steps_per_second": 2.891,
      "step": 33
    },
    {
      "epoch": 6.8,
      "grad_norm": 2.334580659866333,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.6982,
      "step": 34
    },
    {
      "epoch": 6.8,
      "eval_accuracy": 0.55,
      "eval_loss": 0.7200292944908142,
      "eval_runtime": 1.0374,
      "eval_samples_per_second": 96.394,
      "eval_steps_per_second": 2.892,
      "step": 34
    },
    {
      "epoch": 7.0,
      "grad_norm": 7.557387828826904,
      "learning_rate": 1.5625e-05,
      "loss": 0.6886,
      "step": 35
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.52,
      "eval_loss": 0.7210448980331421,
      "eval_runtime": 1.0426,
      "eval_samples_per_second": 95.918,
      "eval_steps_per_second": 2.878,
      "step": 35
    },
    {
      "epoch": 7.2,
      "grad_norm": 2.4877591133117676,
      "learning_rate": 1.4583333333333335e-05,
      "loss": 0.7269,
      "step": 36
    },
    {
      "epoch": 7.2,
      "eval_accuracy": 0.52,
      "eval_loss": 0.721484363079071,
      "eval_runtime": 1.041,
      "eval_samples_per_second": 96.06,
      "eval_steps_per_second": 2.882,
      "step": 36
    },
    {
      "epoch": 7.4,
      "grad_norm": 8.11579704284668,
      "learning_rate": 1.3541666666666666e-05,
      "loss": 0.6664,
      "step": 37
    },
    {
      "epoch": 7.4,
      "eval_accuracy": 0.48,
      "eval_loss": 0.7200976014137268,
      "eval_runtime": 1.0385,
      "eval_samples_per_second": 96.296,
      "eval_steps_per_second": 2.889,
      "step": 37
    },
    {
      "epoch": 7.6,
      "grad_norm": 4.135441303253174,
      "learning_rate": 1.25e-05,
      "loss": 0.7207,
      "step": 38
    },
    {
      "epoch": 7.6,
      "eval_accuracy": 0.48,
      "eval_loss": 0.7217872738838196,
      "eval_runtime": 1.0493,
      "eval_samples_per_second": 95.297,
      "eval_steps_per_second": 2.859,
      "step": 38
    },
    {
      "epoch": 7.8,
      "grad_norm": 3.225309133529663,
      "learning_rate": 1.1458333333333333e-05,
      "loss": 0.6954,
      "step": 39
    },
    {
      "epoch": 7.8,
      "eval_accuracy": 0.47,
      "eval_loss": 0.7211818099021912,
      "eval_runtime": 1.0489,
      "eval_samples_per_second": 95.336,
      "eval_steps_per_second": 2.86,
      "step": 39
    },
    {
      "epoch": 8.0,
      "grad_norm": 10.74594783782959,
      "learning_rate": 1.0416666666666668e-05,
      "loss": 0.7275,
      "step": 40
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.51,
      "eval_loss": 0.7183593511581421,
      "eval_runtime": 1.0511,
      "eval_samples_per_second": 95.137,
      "eval_steps_per_second": 2.854,
      "step": 40
    },
    {
      "epoch": 8.2,
      "grad_norm": 4.628167152404785,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.6705,
      "step": 41
    },
    {
      "epoch": 8.2,
      "eval_accuracy": 0.49,
      "eval_loss": 0.7179295420646667,
      "eval_runtime": 1.0442,
      "eval_samples_per_second": 95.77,
      "eval_steps_per_second": 2.873,
      "step": 41
    },
    {
      "epoch": 8.4,
      "grad_norm": 6.414822101593018,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.6446,
      "step": 42
    },
    {
      "epoch": 8.4,
      "eval_accuracy": 0.51,
      "eval_loss": 0.7144432067871094,
      "eval_runtime": 1.0424,
      "eval_samples_per_second": 95.931,
      "eval_steps_per_second": 2.878,
      "step": 42
    },
    {
      "epoch": 8.6,
      "grad_norm": 2.6695168018341064,
      "learning_rate": 7.2916666666666674e-06,
      "loss": 0.7056,
      "step": 43
    },
    {
      "epoch": 8.6,
      "eval_accuracy": 0.46,
      "eval_loss": 0.7158788442611694,
      "eval_runtime": 1.0443,
      "eval_samples_per_second": 95.757,
      "eval_steps_per_second": 2.873,
      "step": 43
    },
    {
      "epoch": 8.8,
      "grad_norm": 4.791749954223633,
      "learning_rate": 6.25e-06,
      "loss": 0.685,
      "step": 44
    },
    {
      "epoch": 8.8,
      "eval_accuracy": 0.51,
      "eval_loss": 0.7120408415794373,
      "eval_runtime": 1.0511,
      "eval_samples_per_second": 95.14,
      "eval_steps_per_second": 2.854,
      "step": 44
    },
    {
      "epoch": 9.0,
      "grad_norm": 3.8585619926452637,
      "learning_rate": 5.208333333333334e-06,
      "loss": 0.7205,
      "step": 45
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.47,
      "eval_loss": 0.7118456959724426,
      "eval_runtime": 1.041,
      "eval_samples_per_second": 96.059,
      "eval_steps_per_second": 2.882,
      "step": 45
    },
    {
      "epoch": 9.2,
      "grad_norm": 2.295781135559082,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.687,
      "step": 46
    },
    {
      "epoch": 9.2,
      "eval_accuracy": 0.48,
      "eval_loss": 0.7108007073402405,
      "eval_runtime": 1.0379,
      "eval_samples_per_second": 96.349,
      "eval_steps_per_second": 2.89,
      "step": 46
    },
    {
      "epoch": 9.4,
      "grad_norm": 2.4573535919189453,
      "learning_rate": 3.125e-06,
      "loss": 0.6797,
      "step": 47
    },
    {
      "epoch": 9.4,
      "eval_accuracy": 0.47,
      "eval_loss": 0.7099804878234863,
      "eval_runtime": 1.04,
      "eval_samples_per_second": 96.155,
      "eval_steps_per_second": 2.885,
      "step": 47
    },
    {
      "epoch": 9.6,
      "grad_norm": 3.169243097305298,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 0.6795,
      "step": 48
    },
    {
      "epoch": 9.6,
      "eval_accuracy": 0.47,
      "eval_loss": 0.7097461819648743,
      "eval_runtime": 1.0509,
      "eval_samples_per_second": 95.155,
      "eval_steps_per_second": 2.855,
      "step": 48
    },
    {
      "epoch": 9.8,
      "grad_norm": 5.209898948669434,
      "learning_rate": 1.0416666666666667e-06,
      "loss": 0.698,
      "step": 49
    },
    {
      "epoch": 9.8,
      "eval_accuracy": 0.48,
      "eval_loss": 0.7098827958106995,
      "eval_runtime": 1.0539,
      "eval_samples_per_second": 94.889,
      "eval_steps_per_second": 2.847,
      "step": 49
    },
    {
      "epoch": 10.0,
      "grad_norm": 2.3503174781799316,
      "learning_rate": 0.0,
      "loss": 0.7015,
      "step": 50
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.45,
      "eval_loss": 0.7111523151397705,
      "eval_runtime": 1.0519,
      "eval_samples_per_second": 95.065,
      "eval_steps_per_second": 2.852,
      "step": 50
    },
    {
      "epoch": 10.0,
      "step": 50,
      "total_flos": 8024398363623424.0,
      "train_loss": 0.7720831573009491,
      "train_runtime": 187.2255,
      "train_samples_per_second": 21.365,
      "train_steps_per_second": 0.267
    }
  ],
  "logging_steps": 1,
  "max_steps": 50,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 8024398363623424.0,
  "train_batch_size": 10,
  "trial_name": null,
  "trial_params": null
}