{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 15.0,
  "eval_steps": 500,
  "global_step": 4770,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9968553459119497,
      "grad_norm": 0.7076742053031921,
      "learning_rate": 0.0002560077952502327,
      "loss": 0.1862,
      "step": 317
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8941935483870967,
      "eval_loss": 0.0539756715297699,
      "eval_runtime": 5.3793,
      "eval_samples_per_second": 576.28,
      "eval_steps_per_second": 12.083,
      "step": 318
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 0.1704549938440323,
      "learning_rate": 0.00023778312175049682,
      "loss": 0.0474,
      "step": 634
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9209677419354839,
      "eval_loss": 0.04036521539092064,
      "eval_runtime": 5.4418,
      "eval_samples_per_second": 569.666,
      "eval_steps_per_second": 11.945,
      "step": 636
    },
    {
      "epoch": 2.990566037735849,
      "grad_norm": 0.20852865278720856,
      "learning_rate": 0.000219558448250761,
      "loss": 0.0303,
      "step": 951
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9358064516129032,
      "eval_loss": 0.03348854184150696,
      "eval_runtime": 5.4145,
      "eval_samples_per_second": 572.536,
      "eval_steps_per_second": 12.005,
      "step": 954
    },
    {
      "epoch": 3.9874213836477987,
      "grad_norm": 0.28731900453567505,
      "learning_rate": 0.00020133377475102513,
      "loss": 0.0248,
      "step": 1268
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9316129032258065,
      "eval_loss": 0.032027412205934525,
      "eval_runtime": 5.4049,
      "eval_samples_per_second": 573.553,
      "eval_steps_per_second": 12.026,
      "step": 1272
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.08666377514600754,
      "learning_rate": 0.00018310910125128927,
      "loss": 0.0225,
      "step": 1585
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9374193548387096,
      "eval_loss": 0.027690911665558815,
      "eval_runtime": 5.4267,
      "eval_samples_per_second": 571.248,
      "eval_steps_per_second": 11.978,
      "step": 1590
    },
    {
      "epoch": 5.981132075471698,
      "grad_norm": 0.09478049725294113,
      "learning_rate": 0.0001648844277515534,
      "loss": 0.0202,
      "step": 1902
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9409677419354838,
      "eval_loss": 0.02713106758892536,
      "eval_runtime": 5.4468,
      "eval_samples_per_second": 569.137,
      "eval_steps_per_second": 11.934,
      "step": 1908
    },
    {
      "epoch": 6.977987421383648,
      "grad_norm": 0.060874972492456436,
      "learning_rate": 0.00014665975425181757,
      "loss": 0.0183,
      "step": 2219
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.94,
      "eval_loss": 0.02618584595620632,
      "eval_runtime": 5.3927,
      "eval_samples_per_second": 574.854,
      "eval_steps_per_second": 12.053,
      "step": 2226
    },
    {
      "epoch": 7.9748427672955975,
      "grad_norm": 0.06319313496351242,
      "learning_rate": 0.0001284350807520817,
      "loss": 0.0171,
      "step": 2536
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9406451612903226,
      "eval_loss": 0.02482638880610466,
      "eval_runtime": 5.3996,
      "eval_samples_per_second": 574.119,
      "eval_steps_per_second": 12.038,
      "step": 2544
    },
    {
      "epoch": 8.971698113207546,
      "grad_norm": 0.08764627575874329,
      "learning_rate": 0.00011021040725234585,
      "loss": 0.0162,
      "step": 2853
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9425806451612904,
      "eval_loss": 0.024511994794011116,
      "eval_runtime": 5.4361,
      "eval_samples_per_second": 570.265,
      "eval_steps_per_second": 11.957,
      "step": 2862
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.05495918542146683,
      "learning_rate": 9.198573375261e-05,
      "loss": 0.0157,
      "step": 3170
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.947741935483871,
      "eval_loss": 0.02288609929382801,
      "eval_runtime": 5.4756,
      "eval_samples_per_second": 566.152,
      "eval_steps_per_second": 11.871,
      "step": 3180
    },
    {
      "epoch": 10.965408805031446,
      "grad_norm": 0.07047451287508011,
      "learning_rate": 7.376106025287414e-05,
      "loss": 0.0145,
      "step": 3487
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9454838709677419,
      "eval_loss": 0.022273289039731026,
      "eval_runtime": 5.434,
      "eval_samples_per_second": 570.487,
      "eval_steps_per_second": 11.962,
      "step": 3498
    },
    {
      "epoch": 11.962264150943396,
      "grad_norm": 0.04967096447944641,
      "learning_rate": 5.553638675313829e-05,
      "loss": 0.0138,
      "step": 3804
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9451612903225807,
      "eval_loss": 0.02146817557513714,
      "eval_runtime": 5.4207,
      "eval_samples_per_second": 571.884,
      "eval_steps_per_second": 11.991,
      "step": 3816
    },
    {
      "epoch": 12.959119496855346,
      "grad_norm": 0.05987446755170822,
      "learning_rate": 3.731171325340243e-05,
      "loss": 0.0133,
      "step": 4121
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9483870967741935,
      "eval_loss": 0.020876701921224594,
      "eval_runtime": 5.4416,
      "eval_samples_per_second": 569.687,
      "eval_steps_per_second": 11.945,
      "step": 4134
    },
    {
      "epoch": 13.955974842767295,
      "grad_norm": 0.04805705323815346,
      "learning_rate": 1.9087039753666574e-05,
      "loss": 0.0128,
      "step": 4438
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9474193548387096,
      "eval_loss": 0.02010694146156311,
      "eval_runtime": 5.4483,
      "eval_samples_per_second": 568.99,
      "eval_steps_per_second": 11.93,
      "step": 4452
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.0513463132083416,
      "learning_rate": 8.623662539307187e-07,
      "loss": 0.0124,
      "step": 4755
    }
  ],
  "logging_steps": 317,
  "max_steps": 4770,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 1000000000.0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1236646073993904.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.2421834167342251,
    "learning_rate": 0.00027423246874996854,
    "lr_scheduler_type": "linear",
    "num_train_epochs": 15,
    "temperature": 8.720536143796046,
    "weight_decay": 0.0693510968632614
  }
}