{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 2862,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9968553459119497,
      "grad_norm": 0.7500613927841187,
      "learning_rate": 1.7784765897973445e-05,
      "loss": 0.574,
      "step": 317
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.645483870967742,
      "eval_loss": 0.2891772985458374,
      "eval_runtime": 1.5814,
      "eval_samples_per_second": 1960.332,
      "eval_steps_per_second": 41.104,
      "step": 318
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 0.5959137678146362,
      "learning_rate": 1.556953179594689e-05,
      "loss": 0.2286,
      "step": 634
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8493548387096774,
      "eval_loss": 0.11736558377742767,
      "eval_runtime": 1.4156,
      "eval_samples_per_second": 2189.928,
      "eval_steps_per_second": 45.918,
      "step": 636
    },
    {
      "epoch": 2.990566037735849,
      "grad_norm": 0.5286552309989929,
      "learning_rate": 1.3354297693920338e-05,
      "loss": 0.1243,
      "step": 951
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8941935483870967,
      "eval_loss": 0.06792479753494263,
      "eval_runtime": 1.6008,
      "eval_samples_per_second": 1936.498,
      "eval_steps_per_second": 40.604,
      "step": 954
    },
    {
      "epoch": 3.9874213836477987,
      "grad_norm": 0.4791622459888458,
      "learning_rate": 1.1139063591893781e-05,
      "loss": 0.086,
      "step": 1268
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9119354838709678,
      "eval_loss": 0.04815125837922096,
      "eval_runtime": 1.4165,
      "eval_samples_per_second": 2188.476,
      "eval_steps_per_second": 45.887,
      "step": 1272
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.36969393491744995,
      "learning_rate": 8.923829489867226e-06,
      "loss": 0.0682,
      "step": 1585
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9212903225806451,
      "eval_loss": 0.039599157869815826,
      "eval_runtime": 1.6105,
      "eval_samples_per_second": 1924.896,
      "eval_steps_per_second": 40.361,
      "step": 1590
    },
    {
      "epoch": 5.981132075471698,
      "grad_norm": 0.41220352053642273,
      "learning_rate": 6.708595387840672e-06,
      "loss": 0.0589,
      "step": 1902
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9248387096774193,
      "eval_loss": 0.034682855010032654,
      "eval_runtime": 1.4265,
      "eval_samples_per_second": 2173.216,
      "eval_steps_per_second": 45.567,
      "step": 1908
    },
    {
      "epoch": 6.977987421383648,
      "grad_norm": 0.2867417633533478,
      "learning_rate": 4.4933612858141165e-06,
      "loss": 0.0535,
      "step": 2219
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9303225806451613,
      "eval_loss": 0.032181933522224426,
      "eval_runtime": 1.6126,
      "eval_samples_per_second": 1922.335,
      "eval_steps_per_second": 40.307,
      "step": 2226
    },
    {
      "epoch": 7.9748427672955975,
      "grad_norm": 0.30042725801467896,
      "learning_rate": 2.2781271837875614e-06,
      "loss": 0.0505,
      "step": 2536
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9306451612903226,
      "eval_loss": 0.030586862936615944,
      "eval_runtime": 1.4231,
      "eval_samples_per_second": 2178.305,
      "eval_steps_per_second": 45.674,
      "step": 2544
    },
    {
      "epoch": 8.971698113207546,
      "grad_norm": 0.3094331622123718,
      "learning_rate": 6.289308176100629e-08,
      "loss": 0.0486,
      "step": 2853
    }
  ],
  "logging_steps": 317,
  "max_steps": 2862,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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": 742988563029384.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8610448302361252,
    "num_train_epochs": 9,
    "temperature": 3
  }
}