{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 16.0,
  "eval_steps": 500,
  "global_step": 5088,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9968553459119497,
      "grad_norm": 0.381134957075119,
      "learning_rate": 0.00019797823879361723,
      "loss": 0.2559,
      "step": 317
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.89,
      "eval_loss": 0.058846741914749146,
      "eval_runtime": 2.6122,
      "eval_samples_per_second": 1186.735,
      "eval_steps_per_second": 24.883,
      "step": 318
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 0.21276384592056274,
      "learning_rate": 0.00018482395212466384,
      "loss": 0.0471,
      "step": 634
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9338709677419355,
      "eval_loss": 0.031198587268590927,
      "eval_runtime": 2.6181,
      "eval_samples_per_second": 1184.043,
      "eval_steps_per_second": 24.827,
      "step": 636
    },
    {
      "epoch": 2.990566037735849,
      "grad_norm": 0.2707769274711609,
      "learning_rate": 0.00017166966545571045,
      "loss": 0.0304,
      "step": 951
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9393548387096774,
      "eval_loss": 0.027465904131531715,
      "eval_runtime": 2.6194,
      "eval_samples_per_second": 1183.467,
      "eval_steps_per_second": 24.815,
      "step": 954
    },
    {
      "epoch": 3.9874213836477987,
      "grad_norm": 0.1308950036764145,
      "learning_rate": 0.00015851537878675706,
      "loss": 0.0243,
      "step": 1268
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9416129032258065,
      "eval_loss": 0.023094674572348595,
      "eval_runtime": 2.6149,
      "eval_samples_per_second": 1185.525,
      "eval_steps_per_second": 24.858,
      "step": 1272
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.0777081847190857,
      "learning_rate": 0.00014536109211780367,
      "loss": 0.0209,
      "step": 1585
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9435483870967742,
      "eval_loss": 0.02354435622692108,
      "eval_runtime": 2.6158,
      "eval_samples_per_second": 1185.084,
      "eval_steps_per_second": 24.849,
      "step": 1590
    },
    {
      "epoch": 5.981132075471698,
      "grad_norm": 0.0858517661690712,
      "learning_rate": 0.00013220680544885025,
      "loss": 0.019,
      "step": 1902
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9451612903225807,
      "eval_loss": 0.02090289629995823,
      "eval_runtime": 2.6518,
      "eval_samples_per_second": 1169.015,
      "eval_steps_per_second": 24.512,
      "step": 1908
    },
    {
      "epoch": 6.977987421383648,
      "grad_norm": 0.08468086272478104,
      "learning_rate": 0.00011905251877989686,
      "loss": 0.0176,
      "step": 2219
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.947741935483871,
      "eval_loss": 0.01986338384449482,
      "eval_runtime": 2.6294,
      "eval_samples_per_second": 1178.961,
      "eval_steps_per_second": 24.72,
      "step": 2226
    },
    {
      "epoch": 7.9748427672955975,
      "grad_norm": 0.06428312510251999,
      "learning_rate": 0.00010589823211094344,
      "loss": 0.0166,
      "step": 2536
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9470967741935484,
      "eval_loss": 0.019629037007689476,
      "eval_runtime": 2.6518,
      "eval_samples_per_second": 1169.031,
      "eval_steps_per_second": 24.512,
      "step": 2544
    },
    {
      "epoch": 8.971698113207546,
      "grad_norm": 0.07834560424089432,
      "learning_rate": 9.274394544199005e-05,
      "loss": 0.0173,
      "step": 2853
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9438709677419355,
      "eval_loss": 0.020628534257411957,
      "eval_runtime": 2.6779,
      "eval_samples_per_second": 1157.602,
      "eval_steps_per_second": 24.272,
      "step": 2862
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.06184606999158859,
      "learning_rate": 7.958965877303666e-05,
      "loss": 0.0155,
      "step": 3170
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9480645161290323,
      "eval_loss": 0.018417367711663246,
      "eval_runtime": 2.6615,
      "eval_samples_per_second": 1164.775,
      "eval_steps_per_second": 24.423,
      "step": 3180
    },
    {
      "epoch": 10.965408805031446,
      "grad_norm": 0.0588119812309742,
      "learning_rate": 6.643537210408326e-05,
      "loss": 0.0145,
      "step": 3487
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9496774193548387,
      "eval_loss": 0.017711853608489037,
      "eval_runtime": 2.6508,
      "eval_samples_per_second": 1169.454,
      "eval_steps_per_second": 24.521,
      "step": 3498
    },
    {
      "epoch": 11.962264150943396,
      "grad_norm": 0.05123499035835266,
      "learning_rate": 5.328108543512986e-05,
      "loss": 0.0139,
      "step": 3804
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9483870967741935,
      "eval_loss": 0.017743250355124474,
      "eval_runtime": 2.6626,
      "eval_samples_per_second": 1164.291,
      "eval_steps_per_second": 24.413,
      "step": 3816
    },
    {
      "epoch": 12.959119496855346,
      "grad_norm": 0.05427718907594681,
      "learning_rate": 4.0126798766176454e-05,
      "loss": 0.0132,
      "step": 4121
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9493548387096774,
      "eval_loss": 0.01712816022336483,
      "eval_runtime": 2.6545,
      "eval_samples_per_second": 1167.815,
      "eval_steps_per_second": 24.486,
      "step": 4134
    },
    {
      "epoch": 13.955974842767295,
      "grad_norm": 0.0513085275888443,
      "learning_rate": 2.6972512097223057e-05,
      "loss": 0.0128,
      "step": 4438
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9503225806451613,
      "eval_loss": 0.016615070402622223,
      "eval_runtime": 2.6323,
      "eval_samples_per_second": 1177.661,
      "eval_steps_per_second": 24.693,
      "step": 4452
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.04954610392451286,
      "learning_rate": 1.3818225428269658e-05,
      "loss": 0.0123,
      "step": 4755
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.9509677419354838,
      "eval_loss": 0.01620567962527275,
      "eval_runtime": 2.6507,
      "eval_samples_per_second": 1169.516,
      "eval_steps_per_second": 24.522,
      "step": 4770
    },
    {
      "epoch": 15.949685534591195,
      "grad_norm": 0.05111037567257881,
      "learning_rate": 6.639387593162599e-07,
      "loss": 0.012,
      "step": 5072
    }
  ],
  "logging_steps": 317,
  "max_steps": 5088,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 16,
  "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": 1319419533829824.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.33188902619419924,
    "fp16": false,
    "learning_rate": 0.00020387069528254905,
    "lr_scheduler": "cosine",
    "num_train_epochs": 16,
    "temperature": 11,
    "warmup_steps": 175,
    "weight_decay": 0.15922831030843926
  }
}