{
  "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": 0.36009669303894043,
      "learning_rate": 0.00018598534310647178,
      "loss": 0.1867,
      "step": 317
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8990322580645161,
      "eval_loss": 0.04464339092373848,
      "eval_runtime": 5.5111,
      "eval_samples_per_second": 562.5,
      "eval_steps_per_second": 11.794,
      "step": 318
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 0.34826424717903137,
      "learning_rate": 0.00017994931344688276,
      "loss": 0.0427,
      "step": 634
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.927741935483871,
      "eval_loss": 0.03450721129775047,
      "eval_runtime": 5.4344,
      "eval_samples_per_second": 570.443,
      "eval_steps_per_second": 11.961,
      "step": 636
    },
    {
      "epoch": 2.990566037735849,
      "grad_norm": 0.19493499398231506,
      "learning_rate": 0.0001701809729547082,
      "loss": 0.0298,
      "step": 951
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.94,
      "eval_loss": 0.02594602108001709,
      "eval_runtime": 5.4515,
      "eval_samples_per_second": 568.654,
      "eval_steps_per_second": 11.923,
      "step": 954
    },
    {
      "epoch": 3.9874213836477987,
      "grad_norm": 0.12020555883646011,
      "learning_rate": 0.00015710457389761306,
      "loss": 0.0235,
      "step": 1268
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9393548387096774,
      "eval_loss": 0.023855317384004593,
      "eval_runtime": 5.4198,
      "eval_samples_per_second": 571.973,
      "eval_steps_per_second": 11.993,
      "step": 1272
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.23387576639652252,
      "learning_rate": 0.00014128804201050683,
      "loss": 0.0209,
      "step": 1585
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9441935483870968,
      "eval_loss": 0.02296462096273899,
      "eval_runtime": 5.3896,
      "eval_samples_per_second": 575.184,
      "eval_steps_per_second": 12.06,
      "step": 1590
    },
    {
      "epoch": 5.981132075471698,
      "grad_norm": 0.0913698598742485,
      "learning_rate": 0.00012341831071110512,
      "loss": 0.0193,
      "step": 1902
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9429032258064516,
      "eval_loss": 0.0211862213909626,
      "eval_runtime": 5.4057,
      "eval_samples_per_second": 573.472,
      "eval_steps_per_second": 12.024,
      "step": 1908
    },
    {
      "epoch": 6.977987421383648,
      "grad_norm": 0.07358141243457794,
      "learning_rate": 0.0001042714866505231,
      "loss": 0.0176,
      "step": 2219
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9435483870967742,
      "eval_loss": 0.02031989023089409,
      "eval_runtime": 5.4481,
      "eval_samples_per_second": 569.006,
      "eval_steps_per_second": 11.931,
      "step": 2226
    },
    {
      "epoch": 7.9748427672955975,
      "grad_norm": 0.06623157858848572,
      "learning_rate": 8.46791423494761e-05,
      "loss": 0.0164,
      "step": 2536
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9438709677419355,
      "eval_loss": 0.01943557895720005,
      "eval_runtime": 5.431,
      "eval_samples_per_second": 570.797,
      "eval_steps_per_second": 11.968,
      "step": 2544
    },
    {
      "epoch": 8.971698113207546,
      "grad_norm": 0.07008218765258789,
      "learning_rate": 6.549219987658473e-05,
      "loss": 0.0152,
      "step": 2853
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9470967741935484,
      "eval_loss": 0.018487755209207535,
      "eval_runtime": 5.4068,
      "eval_samples_per_second": 573.354,
      "eval_steps_per_second": 12.022,
      "step": 2862
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.053510431200265884,
      "learning_rate": 4.754397414958767e-05,
      "loss": 0.0143,
      "step": 3170
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9496774193548387,
      "eval_loss": 0.017804542556405067,
      "eval_runtime": 5.4089,
      "eval_samples_per_second": 573.127,
      "eval_steps_per_second": 12.017,
      "step": 3180
    },
    {
      "epoch": 10.965408805031446,
      "grad_norm": 0.06439082324504852,
      "learning_rate": 3.161398093898263e-05,
      "loss": 0.0136,
      "step": 3487
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.95,
      "eval_loss": 0.017272653058171272,
      "eval_runtime": 5.4681,
      "eval_samples_per_second": 566.924,
      "eval_steps_per_second": 11.887,
      "step": 3498
    },
    {
      "epoch": 11.962264150943396,
      "grad_norm": 0.05270359292626381,
      "learning_rate": 1.8394081441551166e-05,
      "loss": 0.0131,
      "step": 3804
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9480645161290323,
      "eval_loss": 0.01707734353840351,
      "eval_runtime": 5.4779,
      "eval_samples_per_second": 565.907,
      "eval_steps_per_second": 11.866,
      "step": 3816
    },
    {
      "epoch": 12.959119496855346,
      "grad_norm": 0.05668620765209198,
      "learning_rate": 8.458433810828254e-06,
      "loss": 0.0128,
      "step": 4121
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.947741935483871,
      "eval_loss": 0.016805246472358704,
      "eval_runtime": 5.433,
      "eval_samples_per_second": 570.591,
      "eval_steps_per_second": 11.964,
      "step": 4134
    },
    {
      "epoch": 13.955974842767295,
      "grad_norm": 0.05400459095835686,
      "learning_rate": 2.2385566902754444e-06,
      "loss": 0.0125,
      "step": 4438
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9490322580645161,
      "eval_loss": 0.016659492626786232,
      "eval_runtime": 5.4145,
      "eval_samples_per_second": 572.537,
      "eval_steps_per_second": 12.005,
      "step": 4452
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.04551490768790245,
      "learning_rate": 4.587773711453001e-09,
      "loss": 0.0124,
      "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": 1240673718275628.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8625117920088357,
    "learning_rate": 0.0001880269089840747,
    "lr_scheduler_type": "cosine",
    "num_train_epochs": 15,
    "temperature": 9.794825072054605,
    "weight_decay": 0.2689344090657866
  }
}