{
  "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": 0.23767763376235962,
      "learning_rate": 0.00016296356874677235,
      "loss": 0.191,
      "step": 317
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9019354838709678,
      "eval_loss": 0.0458562970161438,
      "eval_runtime": 5.4882,
      "eval_samples_per_second": 564.845,
      "eval_steps_per_second": 11.844,
      "step": 318
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 0.1802085041999817,
      "learning_rate": 0.0001576746953443942,
      "loss": 0.0439,
      "step": 634
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9232258064516129,
      "eval_loss": 0.03263504058122635,
      "eval_runtime": 5.4169,
      "eval_samples_per_second": 572.287,
      "eval_steps_per_second": 12.0,
      "step": 636
    },
    {
      "epoch": 2.990566037735849,
      "grad_norm": 0.2029435783624649,
      "learning_rate": 0.000149115506750554,
      "loss": 0.03,
      "step": 951
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9361290322580645,
      "eval_loss": 0.02502182498574257,
      "eval_runtime": 5.3728,
      "eval_samples_per_second": 576.983,
      "eval_steps_per_second": 12.098,
      "step": 954
    },
    {
      "epoch": 3.9874213836477987,
      "grad_norm": 0.3161197900772095,
      "learning_rate": 0.00013765774012707753,
      "loss": 0.0239,
      "step": 1268
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9358064516129032,
      "eval_loss": 0.025332694873213768,
      "eval_runtime": 5.4466,
      "eval_samples_per_second": 569.16,
      "eval_steps_per_second": 11.934,
      "step": 1272
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.14759434759616852,
      "learning_rate": 0.00012379902180837435,
      "loss": 0.0213,
      "step": 1585
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9403225806451613,
      "eval_loss": 0.021437037736177444,
      "eval_runtime": 5.4616,
      "eval_samples_per_second": 567.598,
      "eval_steps_per_second": 11.901,
      "step": 1590
    },
    {
      "epoch": 5.981132075471698,
      "grad_norm": 0.11553900688886642,
      "learning_rate": 0.00010814125471524764,
      "loss": 0.0191,
      "step": 1902
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9412903225806452,
      "eval_loss": 0.020237039774656296,
      "eval_runtime": 5.4187,
      "eval_samples_per_second": 572.088,
      "eval_steps_per_second": 11.995,
      "step": 1908
    },
    {
      "epoch": 6.977987421383648,
      "grad_norm": 0.07789044082164764,
      "learning_rate": 9.136447689521929e-05,
      "loss": 0.0178,
      "step": 2219
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9458064516129032,
      "eval_loss": 0.019201194867491722,
      "eval_runtime": 5.3973,
      "eval_samples_per_second": 574.359,
      "eval_steps_per_second": 12.043,
      "step": 2226
    },
    {
      "epoch": 7.9748427672955975,
      "grad_norm": 0.06594134867191315,
      "learning_rate": 7.419732654839714e-05,
      "loss": 0.0163,
      "step": 2536
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9432258064516129,
      "eval_loss": 0.018832536414265633,
      "eval_runtime": 5.4136,
      "eval_samples_per_second": 572.635,
      "eval_steps_per_second": 12.007,
      "step": 2544
    },
    {
      "epoch": 8.971698113207546,
      "grad_norm": 0.0778631940484047,
      "learning_rate": 5.738539628284173e-05,
      "loss": 0.0154,
      "step": 2853
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9448387096774193,
      "eval_loss": 0.017736049368977547,
      "eval_runtime": 5.4576,
      "eval_samples_per_second": 568.014,
      "eval_steps_per_second": 11.91,
      "step": 2862
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.054840076714754105,
      "learning_rate": 4.1658851017015924e-05,
      "loss": 0.0145,
      "step": 3170
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9483870967741935,
      "eval_loss": 0.01691010408103466,
      "eval_runtime": 5.4183,
      "eval_samples_per_second": 572.136,
      "eval_steps_per_second": 11.996,
      "step": 3180
    },
    {
      "epoch": 10.965408805031446,
      "grad_norm": 0.06288731843233109,
      "learning_rate": 2.7700715927704587e-05,
      "loss": 0.0138,
      "step": 3487
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9487096774193549,
      "eval_loss": 0.01642891950905323,
      "eval_runtime": 5.4182,
      "eval_samples_per_second": 572.15,
      "eval_steps_per_second": 11.997,
      "step": 3498
    },
    {
      "epoch": 11.962264150943396,
      "grad_norm": 0.054928138852119446,
      "learning_rate": 1.6117211740808655e-05,
      "loss": 0.0133,
      "step": 3804
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9487096774193549,
      "eval_loss": 0.016290361061692238,
      "eval_runtime": 5.4138,
      "eval_samples_per_second": 572.606,
      "eval_steps_per_second": 12.006,
      "step": 3816
    },
    {
      "epoch": 12.959119496855346,
      "grad_norm": 0.06059136986732483,
      "learning_rate": 7.411425743542737e-06,
      "loss": 0.013,
      "step": 4121
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.95,
      "eval_loss": 0.015878545120358467,
      "eval_runtime": 5.4661,
      "eval_samples_per_second": 567.13,
      "eval_steps_per_second": 11.891,
      "step": 4134
    },
    {
      "epoch": 13.955974842767295,
      "grad_norm": 0.05395493656396866,
      "learning_rate": 1.961462021662692e-06,
      "loss": 0.0127,
      "step": 4438
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9509677419354838,
      "eval_loss": 0.01581409014761448,
      "eval_runtime": 5.422,
      "eval_samples_per_second": 571.75,
      "eval_steps_per_second": 11.988,
      "step": 4452
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.047738075256347656,
      "learning_rate": 4.019886535859987e-09,
      "loss": 0.0126,
      "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": 0.9884583860107803,
    "learning_rate": 0.00016475242401724032,
    "lr_scheduler_type": "cosine",
    "num_train_epochs": 15,
    "temperature": 11.128466952956192,
    "weight_decay": 0.28960673719057545
  }
}