{
  "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": 1.04072105884552,
      "learning_rate": 0.00041128900119572176,
      "loss": 0.2584,
      "step": 317
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7422580645161291,
      "eval_loss": 0.1071159690618515,
      "eval_runtime": 5.4205,
      "eval_samples_per_second": 571.899,
      "eval_steps_per_second": 11.991,
      "step": 318
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 0.5266364216804504,
      "learning_rate": 0.0003820101749260061,
      "loss": 0.0823,
      "step": 634
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8493548387096774,
      "eval_loss": 0.06963900476694107,
      "eval_runtime": 5.4463,
      "eval_samples_per_second": 569.198,
      "eval_steps_per_second": 11.935,
      "step": 636
    },
    {
      "epoch": 2.990566037735849,
      "grad_norm": 0.6715505123138428,
      "learning_rate": 0.0003527313486562904,
      "loss": 0.0581,
      "step": 951
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8670967741935484,
      "eval_loss": 0.0621778629720211,
      "eval_runtime": 5.5165,
      "eval_samples_per_second": 561.951,
      "eval_steps_per_second": 11.783,
      "step": 954
    },
    {
      "epoch": 3.9874213836477987,
      "grad_norm": 0.15568938851356506,
      "learning_rate": 0.00032345252238657476,
      "loss": 0.0477,
      "step": 1268
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8770967741935484,
      "eval_loss": 0.06203094497323036,
      "eval_runtime": 5.4775,
      "eval_samples_per_second": 565.955,
      "eval_steps_per_second": 11.867,
      "step": 1272
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.11771262437105179,
      "learning_rate": 0.00029417369611685915,
      "loss": 0.0359,
      "step": 1585
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9083870967741936,
      "eval_loss": 0.04885668307542801,
      "eval_runtime": 5.4322,
      "eval_samples_per_second": 570.673,
      "eval_steps_per_second": 11.966,
      "step": 1590
    },
    {
      "epoch": 5.981132075471698,
      "grad_norm": 0.19639770686626434,
      "learning_rate": 0.0002648948698471435,
      "loss": 0.0304,
      "step": 1902
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9012903225806451,
      "eval_loss": 0.05035402998328209,
      "eval_runtime": 5.4681,
      "eval_samples_per_second": 566.923,
      "eval_steps_per_second": 11.887,
      "step": 1908
    },
    {
      "epoch": 6.977987421383648,
      "grad_norm": 0.09110367298126221,
      "learning_rate": 0.00023561604357742784,
      "loss": 0.0274,
      "step": 2219
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9058064516129032,
      "eval_loss": 0.04763167351484299,
      "eval_runtime": 5.4566,
      "eval_samples_per_second": 568.117,
      "eval_steps_per_second": 11.912,
      "step": 2226
    },
    {
      "epoch": 7.9748427672955975,
      "grad_norm": 0.07694189995527267,
      "learning_rate": 0.00020633721730771218,
      "loss": 0.025,
      "step": 2536
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9103225806451613,
      "eval_loss": 0.04542699456214905,
      "eval_runtime": 5.4199,
      "eval_samples_per_second": 571.971,
      "eval_steps_per_second": 11.993,
      "step": 2544
    },
    {
      "epoch": 8.971698113207546,
      "grad_norm": 0.09620998799800873,
      "learning_rate": 0.00017705839103799654,
      "loss": 0.0231,
      "step": 2853
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9174193548387096,
      "eval_loss": 0.04218007251620293,
      "eval_runtime": 5.4278,
      "eval_samples_per_second": 571.134,
      "eval_steps_per_second": 11.975,
      "step": 2862
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.06652583926916122,
      "learning_rate": 0.0001477795647682809,
      "loss": 0.0205,
      "step": 3170
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9167741935483871,
      "eval_loss": 0.040490638464689255,
      "eval_runtime": 5.4856,
      "eval_samples_per_second": 565.113,
      "eval_steps_per_second": 11.849,
      "step": 3180
    },
    {
      "epoch": 10.965408805031446,
      "grad_norm": 0.1630864441394806,
      "learning_rate": 0.00011850073849856523,
      "loss": 0.0191,
      "step": 3487
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9241935483870968,
      "eval_loss": 0.03858407959342003,
      "eval_runtime": 5.4468,
      "eval_samples_per_second": 569.139,
      "eval_steps_per_second": 11.934,
      "step": 3498
    },
    {
      "epoch": 11.962264150943396,
      "grad_norm": 0.053046438843011856,
      "learning_rate": 8.922191222884958e-05,
      "loss": 0.0178,
      "step": 3804
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9225806451612903,
      "eval_loss": 0.03706495836377144,
      "eval_runtime": 5.4181,
      "eval_samples_per_second": 572.151,
      "eval_steps_per_second": 11.997,
      "step": 3816
    },
    {
      "epoch": 12.959119496855346,
      "grad_norm": 0.06040259450674057,
      "learning_rate": 5.9943085959133934e-05,
      "loss": 0.0169,
      "step": 4121
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9241935483870968,
      "eval_loss": 0.03619401156902313,
      "eval_runtime": 5.4534,
      "eval_samples_per_second": 568.455,
      "eval_steps_per_second": 11.919,
      "step": 4134
    },
    {
      "epoch": 13.955974842767295,
      "grad_norm": 0.05235395208001137,
      "learning_rate": 3.066425968941828e-05,
      "loss": 0.0159,
      "step": 4438
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9241935483870968,
      "eval_loss": 0.0354730561375618,
      "eval_runtime": 5.4401,
      "eval_samples_per_second": 569.842,
      "eval_steps_per_second": 11.948,
      "step": 4452
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.051196809858083725,
      "learning_rate": 1.3854334197026332e-06,
      "loss": 0.0154,
      "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": 0.015638508087186143,
    "learning_rate": 0.00044056782746543737,
    "lr_scheduler_type": "linear",
    "num_train_epochs": 15,
    "temperature": 5.955400780245458,
    "weight_decay": 0.08329448851063012
  }
}