{
  "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": 0.2736971080303192,
      "learning_rate": 0.0001710898087245655,
      "loss": 0.1871,
      "step": 317
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9045161290322581,
      "eval_loss": 0.0472712367773056,
      "eval_runtime": 5.4187,
      "eval_samples_per_second": 572.094,
      "eval_steps_per_second": 11.996,
      "step": 318
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 0.2054843306541443,
      "learning_rate": 0.0001589102737221655,
      "loss": 0.0426,
      "step": 634
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9264516129032258,
      "eval_loss": 0.031125523149967194,
      "eval_runtime": 5.417,
      "eval_samples_per_second": 572.268,
      "eval_steps_per_second": 11.999,
      "step": 636
    },
    {
      "epoch": 2.990566037735849,
      "grad_norm": 0.25806155800819397,
      "learning_rate": 0.00014673073871976548,
      "loss": 0.0293,
      "step": 951
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9374193548387096,
      "eval_loss": 0.024875458329916,
      "eval_runtime": 5.4169,
      "eval_samples_per_second": 572.284,
      "eval_steps_per_second": 12.0,
      "step": 954
    },
    {
      "epoch": 3.9874213836477987,
      "grad_norm": 0.19356748461723328,
      "learning_rate": 0.00013455120371736547,
      "loss": 0.0233,
      "step": 1268
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9416129032258065,
      "eval_loss": 0.022784613072872162,
      "eval_runtime": 5.4349,
      "eval_samples_per_second": 570.385,
      "eval_steps_per_second": 11.96,
      "step": 1272
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.06689193099737167,
      "learning_rate": 0.00012237166871496545,
      "loss": 0.0209,
      "step": 1585
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9438709677419355,
      "eval_loss": 0.020542802289128304,
      "eval_runtime": 5.4272,
      "eval_samples_per_second": 571.199,
      "eval_steps_per_second": 11.977,
      "step": 1590
    },
    {
      "epoch": 5.981132075471698,
      "grad_norm": 0.13203346729278564,
      "learning_rate": 0.00011019213371256543,
      "loss": 0.0189,
      "step": 1902
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9419354838709677,
      "eval_loss": 0.020823262631893158,
      "eval_runtime": 5.4235,
      "eval_samples_per_second": 571.582,
      "eval_steps_per_second": 11.985,
      "step": 1908
    },
    {
      "epoch": 6.977987421383648,
      "grad_norm": 0.09329136461019516,
      "learning_rate": 9.801259871016543e-05,
      "loss": 0.0176,
      "step": 2219
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9438709677419355,
      "eval_loss": 0.019519412890076637,
      "eval_runtime": 5.4089,
      "eval_samples_per_second": 573.13,
      "eval_steps_per_second": 12.017,
      "step": 2226
    },
    {
      "epoch": 7.9748427672955975,
      "grad_norm": 0.07326487451791763,
      "learning_rate": 8.583306370776541e-05,
      "loss": 0.0165,
      "step": 2536
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.944516129032258,
      "eval_loss": 0.019586728885769844,
      "eval_runtime": 5.3939,
      "eval_samples_per_second": 574.721,
      "eval_steps_per_second": 12.051,
      "step": 2544
    },
    {
      "epoch": 8.971698113207546,
      "grad_norm": 0.07820314168930054,
      "learning_rate": 7.36535287053654e-05,
      "loss": 0.0157,
      "step": 2853
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9458064516129032,
      "eval_loss": 0.01853923685848713,
      "eval_runtime": 5.4292,
      "eval_samples_per_second": 570.989,
      "eval_steps_per_second": 11.972,
      "step": 2862
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.05801521614193916,
      "learning_rate": 6.147399370296537e-05,
      "loss": 0.015,
      "step": 3170
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9464516129032258,
      "eval_loss": 0.01768723502755165,
      "eval_runtime": 5.4502,
      "eval_samples_per_second": 568.786,
      "eval_steps_per_second": 11.926,
      "step": 3180
    },
    {
      "epoch": 10.965408805031446,
      "grad_norm": 0.062277697026729584,
      "learning_rate": 4.9294458700565364e-05,
      "loss": 0.0143,
      "step": 3487
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9487096774193549,
      "eval_loss": 0.017170609906315804,
      "eval_runtime": 5.4238,
      "eval_samples_per_second": 571.551,
      "eval_steps_per_second": 11.984,
      "step": 3498
    },
    {
      "epoch": 11.962264150943396,
      "grad_norm": 0.055214520543813705,
      "learning_rate": 3.7114923698165344e-05,
      "loss": 0.0136,
      "step": 3804
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9448387096774193,
      "eval_loss": 0.017127474769949913,
      "eval_runtime": 5.4107,
      "eval_samples_per_second": 572.934,
      "eval_steps_per_second": 12.013,
      "step": 3816
    },
    {
      "epoch": 12.959119496855346,
      "grad_norm": 0.06277412921190262,
      "learning_rate": 2.493538869576533e-05,
      "loss": 0.0132,
      "step": 4121
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.947741935483871,
      "eval_loss": 0.016646217554807663,
      "eval_runtime": 5.4524,
      "eval_samples_per_second": 568.559,
      "eval_steps_per_second": 11.921,
      "step": 4134
    },
    {
      "epoch": 13.955974842767295,
      "grad_norm": 0.05029454454779625,
      "learning_rate": 1.2755853693365315e-05,
      "loss": 0.0127,
      "step": 4438
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.947741935483871,
      "eval_loss": 0.015961091965436935,
      "eval_runtime": 5.418,
      "eval_samples_per_second": 572.166,
      "eval_steps_per_second": 11.997,
      "step": 4452
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.04746360704302788,
      "learning_rate": 5.763186909653004e-07,
      "loss": 0.0123,
      "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": 0.8512267071251705,
    "learning_rate": 0.00018326934372696552,
    "lr_scheduler_type": "linear",
    "num_train_epochs": 15,
    "temperature": 10.15370255750726,
    "weight_decay": 0.29327015385065763
  }
}