{
  "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": 0.42379891872406006,
      "learning_rate": 0.00013801466204863748,
      "loss": 0.2007,
      "step": 317
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8958064516129032,
      "eval_loss": 0.04874483495950699,
      "eval_runtime": 5.4128,
      "eval_samples_per_second": 572.716,
      "eval_steps_per_second": 12.009,
      "step": 318
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 0.2129104882478714,
      "learning_rate": 0.0001281896793696754,
      "loss": 0.0453,
      "step": 634
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9332258064516129,
      "eval_loss": 0.031727299094200134,
      "eval_runtime": 5.5097,
      "eval_samples_per_second": 562.647,
      "eval_steps_per_second": 11.797,
      "step": 636
    },
    {
      "epoch": 2.990566037735849,
      "grad_norm": 0.2193145602941513,
      "learning_rate": 0.00011836469669071336,
      "loss": 0.0301,
      "step": 951
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9374193548387096,
      "eval_loss": 0.023718850687146187,
      "eval_runtime": 5.4248,
      "eval_samples_per_second": 571.454,
      "eval_steps_per_second": 11.982,
      "step": 954
    },
    {
      "epoch": 3.9874213836477987,
      "grad_norm": 0.14876721799373627,
      "learning_rate": 0.00010853971401175128,
      "loss": 0.0243,
      "step": 1268
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9406451612903226,
      "eval_loss": 0.021157393231987953,
      "eval_runtime": 5.4343,
      "eval_samples_per_second": 570.447,
      "eval_steps_per_second": 11.961,
      "step": 1272
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.07504376024007797,
      "learning_rate": 9.871473133278923e-05,
      "loss": 0.0215,
      "step": 1585
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9429032258064516,
      "eval_loss": 0.020281489938497543,
      "eval_runtime": 5.5277,
      "eval_samples_per_second": 560.811,
      "eval_steps_per_second": 11.759,
      "step": 1590
    },
    {
      "epoch": 5.981132075471698,
      "grad_norm": 0.11960328370332718,
      "learning_rate": 8.888974865382715e-05,
      "loss": 0.0197,
      "step": 1902
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9454838709677419,
      "eval_loss": 0.019814670085906982,
      "eval_runtime": 5.4509,
      "eval_samples_per_second": 568.713,
      "eval_steps_per_second": 11.925,
      "step": 1908
    },
    {
      "epoch": 6.977987421383648,
      "grad_norm": 0.08081384003162384,
      "learning_rate": 7.906476597486509e-05,
      "loss": 0.0184,
      "step": 2219
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9435483870967742,
      "eval_loss": 0.01884547807276249,
      "eval_runtime": 5.4199,
      "eval_samples_per_second": 571.969,
      "eval_steps_per_second": 11.993,
      "step": 2226
    },
    {
      "epoch": 7.9748427672955975,
      "grad_norm": 0.07167127728462219,
      "learning_rate": 6.923978329590303e-05,
      "loss": 0.0173,
      "step": 2536
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9451612903225807,
      "eval_loss": 0.018296979367733,
      "eval_runtime": 5.4972,
      "eval_samples_per_second": 563.924,
      "eval_steps_per_second": 11.824,
      "step": 2544
    },
    {
      "epoch": 8.971698113207546,
      "grad_norm": 0.08106838166713715,
      "learning_rate": 5.941480061694095e-05,
      "loss": 0.0164,
      "step": 2853
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9419354838709677,
      "eval_loss": 0.018038731068372726,
      "eval_runtime": 5.4283,
      "eval_samples_per_second": 571.085,
      "eval_steps_per_second": 11.974,
      "step": 2862
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.05913994088768959,
      "learning_rate": 4.9589817937978886e-05,
      "loss": 0.0157,
      "step": 3170
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9487096774193549,
      "eval_loss": 0.017092861235141754,
      "eval_runtime": 5.4238,
      "eval_samples_per_second": 571.555,
      "eval_steps_per_second": 11.984,
      "step": 3180
    },
    {
      "epoch": 10.965408805031446,
      "grad_norm": 0.0745585560798645,
      "learning_rate": 3.976483525901682e-05,
      "loss": 0.015,
      "step": 3487
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9480645161290323,
      "eval_loss": 0.01634313352406025,
      "eval_runtime": 5.4783,
      "eval_samples_per_second": 565.874,
      "eval_steps_per_second": 11.865,
      "step": 3498
    },
    {
      "epoch": 11.962264150943396,
      "grad_norm": 0.05740603059530258,
      "learning_rate": 2.9939852580054752e-05,
      "loss": 0.0144,
      "step": 3804
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9474193548387096,
      "eval_loss": 0.016295555979013443,
      "eval_runtime": 5.4332,
      "eval_samples_per_second": 570.57,
      "eval_steps_per_second": 11.964,
      "step": 3816
    },
    {
      "epoch": 12.959119496855346,
      "grad_norm": 0.07131999731063843,
      "learning_rate": 2.0114869901092685e-05,
      "loss": 0.0139,
      "step": 4121
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9470967741935484,
      "eval_loss": 0.015728095546364784,
      "eval_runtime": 5.4211,
      "eval_samples_per_second": 571.838,
      "eval_steps_per_second": 11.99,
      "step": 4134
    },
    {
      "epoch": 13.955974842767295,
      "grad_norm": 0.05492794141173363,
      "learning_rate": 1.0289887222130618e-05,
      "loss": 0.0134,
      "step": 4438
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9493548387096774,
      "eval_loss": 0.01537332870066166,
      "eval_runtime": 5.4466,
      "eval_samples_per_second": 569.159,
      "eval_steps_per_second": 11.934,
      "step": 4452
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.04989900812506676,
      "learning_rate": 4.6490454316855204e-07,
      "loss": 0.0131,
      "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": 0.37288483884424406,
    "learning_rate": 0.00014783964472759955,
    "lr_scheduler_type": "linear",
    "num_train_epochs": 15,
    "temperature": 9.186082756077344,
    "weight_decay": 0.11993053192781417
  }
}