{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "learning_rate": 0.0002954545454545454,
      "loss": 0.71,
      "step": 250
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00028787878787878786,
      "loss": 0.5325,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5033329725265503,
      "eval_runtime": 9.0003,
      "eval_samples_per_second": 888.861,
      "eval_steps_per_second": 27.777,
      "step": 500
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002803030303030303,
      "loss": 0.4322,
      "step": 750
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002727272727272727,
      "loss": 0.441,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.4973089098930359,
      "eval_runtime": 9.0771,
      "eval_samples_per_second": 881.341,
      "eval_steps_per_second": 27.542,
      "step": 1000
    },
    {
      "epoch": 2.5,
      "learning_rate": 0.0002651515151515151,
      "loss": 0.352,
      "step": 1250
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.00025757575757575756,
      "loss": 0.372,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.5114700198173523,
      "eval_runtime": 9.3118,
      "eval_samples_per_second": 859.12,
      "eval_steps_per_second": 26.848,
      "step": 1500
    },
    {
      "epoch": 3.5,
      "learning_rate": 0.00025,
      "loss": 0.2921,
      "step": 1750
    },
    {
      "epoch": 4.0,
      "learning_rate": 0.0002424242424242424,
      "loss": 0.3107,
      "step": 2000
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.5385135412216187,
      "eval_runtime": 9.7573,
      "eval_samples_per_second": 819.896,
      "eval_steps_per_second": 25.622,
      "step": 2000
    },
    {
      "epoch": 4.5,
      "learning_rate": 0.00023484848484848483,
      "loss": 0.2439,
      "step": 2250
    },
    {
      "epoch": 5.0,
      "learning_rate": 0.00022727272727272725,
      "loss": 0.2619,
      "step": 2500
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.5607308149337769,
      "eval_runtime": 9.0142,
      "eval_samples_per_second": 887.492,
      "eval_steps_per_second": 27.734,
      "step": 2500
    },
    {
      "epoch": 5.5,
      "learning_rate": 0.00021969696969696969,
      "loss": 0.2105,
      "step": 2750
    },
    {
      "epoch": 6.0,
      "learning_rate": 0.0002121212121212121,
      "loss": 0.2257,
      "step": 3000
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.5820582509040833,
      "eval_runtime": 22.0785,
      "eval_samples_per_second": 362.343,
      "eval_steps_per_second": 11.323,
      "step": 3000
    },
    {
      "epoch": 6.5,
      "learning_rate": 0.0002045454545454545,
      "loss": 0.1812,
      "step": 3250
    },
    {
      "epoch": 7.0,
      "learning_rate": 0.00019696969696969695,
      "loss": 0.1947,
      "step": 3500
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.6063645482063293,
      "eval_runtime": 22.1706,
      "eval_samples_per_second": 360.838,
      "eval_steps_per_second": 11.276,
      "step": 3500
    },
    {
      "epoch": 7.5,
      "learning_rate": 0.00018939393939393937,
      "loss": 0.1597,
      "step": 3750
    },
    {
      "epoch": 8.0,
      "learning_rate": 0.0001818181818181818,
      "loss": 0.1715,
      "step": 4000
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.6192964911460876,
      "eval_runtime": 22.493,
      "eval_samples_per_second": 355.666,
      "eval_steps_per_second": 11.115,
      "step": 4000
    },
    {
      "epoch": 8.5,
      "learning_rate": 0.00017424242424242422,
      "loss": 0.1439,
      "step": 4250
    },
    {
      "epoch": 9.0,
      "learning_rate": 0.00016666666666666666,
      "loss": 0.1535,
      "step": 4500
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.6339118480682373,
      "eval_runtime": 17.7948,
      "eval_samples_per_second": 449.57,
      "eval_steps_per_second": 14.049,
      "step": 4500
    },
    {
      "epoch": 9.5,
      "learning_rate": 0.00015909090909090907,
      "loss": 0.1322,
      "step": 4750
    },
    {
      "epoch": 10.0,
      "learning_rate": 0.00015151515151515152,
      "loss": 0.1403,
      "step": 5000
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.648122251033783,
      "eval_runtime": 22.0949,
      "eval_samples_per_second": 362.074,
      "eval_steps_per_second": 11.315,
      "step": 5000
    },
    {
      "epoch": 10.5,
      "learning_rate": 0.00014393939393939393,
      "loss": 0.1229,
      "step": 5250
    },
    {
      "epoch": 11.0,
      "learning_rate": 0.00013636363636363634,
      "loss": 0.1297,
      "step": 5500
    },
    {
      "epoch": 11.0,
      "eval_loss": 0.6578399538993835,
      "eval_runtime": 22.4161,
      "eval_samples_per_second": 356.887,
      "eval_steps_per_second": 11.153,
      "step": 5500
    },
    {
      "epoch": 11.5,
      "learning_rate": 0.00012878787878787878,
      "loss": 0.1158,
      "step": 5750
    },
    {
      "epoch": 12.0,
      "learning_rate": 0.0001212121212121212,
      "loss": 0.1217,
      "step": 6000
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.6666226983070374,
      "eval_runtime": 17.8264,
      "eval_samples_per_second": 448.773,
      "eval_steps_per_second": 14.024,
      "step": 6000
    },
    {
      "epoch": 12.5,
      "learning_rate": 0.00011363636363636362,
      "loss": 0.1106,
      "step": 6250
    },
    {
      "epoch": 13.0,
      "learning_rate": 0.00010606060606060605,
      "loss": 0.1152,
      "step": 6500
    },
    {
      "epoch": 13.0,
      "eval_loss": 0.6757141351699829,
      "eval_runtime": 15.3201,
      "eval_samples_per_second": 522.191,
      "eval_steps_per_second": 16.318,
      "step": 6500
    },
    {
      "epoch": 13.5,
      "learning_rate": 9.848484848484848e-05,
      "loss": 0.1061,
      "step": 6750
    },
    {
      "epoch": 14.0,
      "learning_rate": 9.09090909090909e-05,
      "loss": 0.1101,
      "step": 7000
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.687734842300415,
      "eval_runtime": 21.8339,
      "eval_samples_per_second": 366.402,
      "eval_steps_per_second": 11.45,
      "step": 7000
    },
    {
      "epoch": 14.5,
      "learning_rate": 8.333333333333333e-05,
      "loss": 0.1024,
      "step": 7250
    },
    {
      "epoch": 15.0,
      "learning_rate": 7.575757575757576e-05,
      "loss": 0.1059,
      "step": 7500
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.6898844242095947,
      "eval_runtime": 22.4671,
      "eval_samples_per_second": 356.076,
      "eval_steps_per_second": 11.127,
      "step": 7500
    },
    {
      "epoch": 15.5,
      "learning_rate": 6.818181818181817e-05,
      "loss": 0.0991,
      "step": 7750
    },
    {
      "epoch": 16.0,
      "learning_rate": 6.06060606060606e-05,
      "loss": 0.1022,
      "step": 8000
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.6962941884994507,
      "eval_runtime": 17.8764,
      "eval_samples_per_second": 447.518,
      "eval_steps_per_second": 13.985,
      "step": 8000
    },
    {
      "epoch": 16.5,
      "learning_rate": 5.3030303030303025e-05,
      "loss": 0.0963,
      "step": 8250
    },
    {
      "epoch": 17.0,
      "learning_rate": 4.545454545454545e-05,
      "loss": 0.0991,
      "step": 8500
    },
    {
      "epoch": 17.0,
      "eval_loss": 0.702376127243042,
      "eval_runtime": 21.7707,
      "eval_samples_per_second": 367.466,
      "eval_steps_per_second": 11.483,
      "step": 8500
    },
    {
      "epoch": 17.5,
      "learning_rate": 3.790909090909091e-05,
      "loss": 0.0939,
      "step": 8750
    },
    {
      "epoch": 18.0,
      "learning_rate": 3.033333333333333e-05,
      "loss": 0.0961,
      "step": 9000
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.7053780555725098,
      "eval_runtime": 22.1712,
      "eval_samples_per_second": 360.829,
      "eval_steps_per_second": 11.276,
      "step": 9000
    },
    {
      "epoch": 18.5,
      "learning_rate": 2.2757575757575757e-05,
      "loss": 0.0914,
      "step": 9250
    },
    {
      "epoch": 19.0,
      "learning_rate": 1.518181818181818e-05,
      "loss": 0.0933,
      "step": 9500
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.7104560732841492,
      "eval_runtime": 22.4548,
      "eval_samples_per_second": 356.271,
      "eval_steps_per_second": 11.133,
      "step": 9500
    },
    {
      "epoch": 19.5,
      "learning_rate": 7.6060606060606056e-06,
      "loss": 0.0893,
      "step": 9750
    },
    {
      "epoch": 20.0,
      "learning_rate": 3.03030303030303e-08,
      "loss": 0.0901,
      "step": 10000
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.7170030474662781,
      "eval_runtime": 17.8822,
      "eval_samples_per_second": 447.372,
      "eval_steps_per_second": 13.98,
      "step": 10000
    }
  ],
  "max_steps": 10000,
  "num_train_epochs": 20,
  "total_flos": 4.180672512e+16,
  "trial_name": null,
  "trial_params": null
}