{
  "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": 0.5192151069641113,
      "learning_rate": 5.2799658937698636e-05,
      "loss": 0.263,
      "step": 317
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8558064516129033,
      "eval_loss": 0.07722163945436478,
      "eval_runtime": 5.3713,
      "eval_samples_per_second": 577.147,
      "eval_steps_per_second": 12.101,
      "step": 318
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 0.3630720376968384,
      "learning_rate": 5.1086081394216154e-05,
      "loss": 0.0714,
      "step": 634
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9187096774193548,
      "eval_loss": 0.03628459945321083,
      "eval_runtime": 5.4431,
      "eval_samples_per_second": 569.53,
      "eval_steps_per_second": 11.942,
      "step": 636
    },
    {
      "epoch": 2.990566037735849,
      "grad_norm": 0.2893732786178589,
      "learning_rate": 4.8312932511839625e-05,
      "loss": 0.0429,
      "step": 951
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.932258064516129,
      "eval_loss": 0.025179537013173103,
      "eval_runtime": 5.4106,
      "eval_samples_per_second": 572.946,
      "eval_steps_per_second": 12.013,
      "step": 954
    },
    {
      "epoch": 3.9874213836477987,
      "grad_norm": 0.19381985068321228,
      "learning_rate": 4.4600653905279676e-05,
      "loss": 0.0333,
      "step": 1268
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9409677419354838,
      "eval_loss": 0.02168211154639721,
      "eval_runtime": 5.42,
      "eval_samples_per_second": 571.958,
      "eval_steps_per_second": 11.993,
      "step": 1272
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.12074659019708633,
      "learning_rate": 4.011047486607245e-05,
      "loss": 0.0286,
      "step": 1585
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.94,
      "eval_loss": 0.019550619646906853,
      "eval_runtime": 5.4026,
      "eval_samples_per_second": 573.8,
      "eval_steps_per_second": 12.031,
      "step": 1590
    },
    {
      "epoch": 5.981132075471698,
      "grad_norm": 0.1383192092180252,
      "learning_rate": 3.5037409955916656e-05,
      "loss": 0.0255,
      "step": 1902
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9370967741935484,
      "eval_loss": 0.018356727436184883,
      "eval_runtime": 5.4519,
      "eval_samples_per_second": 568.612,
      "eval_steps_per_second": 11.923,
      "step": 1908
    },
    {
      "epoch": 6.977987421383648,
      "grad_norm": 0.12834681570529938,
      "learning_rate": 2.9601789259934643e-05,
      "loss": 0.0234,
      "step": 2219
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9409677419354838,
      "eval_loss": 0.017373906448483467,
      "eval_runtime": 5.4536,
      "eval_samples_per_second": 568.43,
      "eval_steps_per_second": 11.919,
      "step": 2226
    },
    {
      "epoch": 7.9748427672955975,
      "grad_norm": 0.10002776980400085,
      "learning_rate": 2.403968915243833e-05,
      "loss": 0.0221,
      "step": 2536
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9383870967741935,
      "eval_loss": 0.016409363597631454,
      "eval_runtime": 5.3751,
      "eval_samples_per_second": 576.738,
      "eval_steps_per_second": 12.093,
      "step": 2544
    },
    {
      "epoch": 8.971698113207546,
      "grad_norm": 0.10179214179515839,
      "learning_rate": 1.8592679180012948e-05,
      "loss": 0.0209,
      "step": 2853
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9409677419354838,
      "eval_loss": 0.015916811302304268,
      "eval_runtime": 5.4101,
      "eval_samples_per_second": 573.001,
      "eval_steps_per_second": 12.015,
      "step": 2862
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.09086760133504868,
      "learning_rate": 1.3497330368683434e-05,
      "loss": 0.0201,
      "step": 3170
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9441935483870968,
      "eval_loss": 0.015536032617092133,
      "eval_runtime": 5.4274,
      "eval_samples_per_second": 571.174,
      "eval_steps_per_second": 11.976,
      "step": 3180
    },
    {
      "epoch": 10.965408805031446,
      "grad_norm": 0.11637966334819794,
      "learning_rate": 8.974940623603928e-06,
      "loss": 0.0194,
      "step": 3487
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9435483870967742,
      "eval_loss": 0.015223735943436623,
      "eval_runtime": 5.4108,
      "eval_samples_per_second": 572.927,
      "eval_steps_per_second": 12.013,
      "step": 3498
    },
    {
      "epoch": 11.962264150943396,
      "grad_norm": 0.08573295176029205,
      "learning_rate": 5.221923461087825e-06,
      "loss": 0.0189,
      "step": 3804
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9416129032258065,
      "eval_loss": 0.01511952094733715,
      "eval_runtime": 5.3754,
      "eval_samples_per_second": 576.706,
      "eval_steps_per_second": 12.092,
      "step": 3816
    },
    {
      "epoch": 12.959119496855346,
      "grad_norm": 0.11835107207298279,
      "learning_rate": 2.4012775033738117e-06,
      "loss": 0.0185,
      "step": 4121
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9416129032258065,
      "eval_loss": 0.014925230294466019,
      "eval_runtime": 5.4176,
      "eval_samples_per_second": 572.21,
      "eval_steps_per_second": 11.998,
      "step": 4134
    },
    {
      "epoch": 13.955974842767295,
      "grad_norm": 0.07725111395120621,
      "learning_rate": 6.355072275323512e-07,
      "loss": 0.0184,
      "step": 4438
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9412903225806452,
      "eval_loss": 0.014896313659846783,
      "eval_runtime": 5.4395,
      "eval_samples_per_second": 569.905,
      "eval_steps_per_second": 11.95,
      "step": 4452
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.07756518572568893,
      "learning_rate": 1.302429982933581e-09,
      "loss": 0.0182,
      "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": 0.0990358007931369,
    "learning_rate": 5.337924214697641e-05,
    "lr_scheduler_type": "cosine",
    "num_train_epochs": 15,
    "temperature": 13.220869405817513,
    "weight_decay": 0.027535903658101313
  }
}