{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 12.0,
  "eval_steps": 500,
  "global_step": 3816,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9968553459119497,
      "grad_norm": 1.7771599292755127,
      "learning_rate": 1.8338574423480085e-05,
      "loss": 1.7697,
      "step": 317
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6706451612903226,
      "eval_loss": 0.9693816304206848,
      "eval_runtime": 1.5802,
      "eval_samples_per_second": 1961.726,
      "eval_steps_per_second": 41.133,
      "step": 318
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 1.420722484588623,
      "learning_rate": 1.6677148846960168e-05,
      "loss": 0.6861,
      "step": 634
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8280645161290323,
      "eval_loss": 0.2716347873210907,
      "eval_runtime": 1.5732,
      "eval_samples_per_second": 1970.448,
      "eval_steps_per_second": 41.316,
      "step": 636
    },
    {
      "epoch": 2.990566037735849,
      "grad_norm": 1.0248616933822632,
      "learning_rate": 1.5015723270440253e-05,
      "loss": 0.2314,
      "step": 951
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9064516129032258,
      "eval_loss": 0.1180456280708313,
      "eval_runtime": 1.5744,
      "eval_samples_per_second": 1969.056,
      "eval_steps_per_second": 41.287,
      "step": 954
    },
    {
      "epoch": 3.9874213836477987,
      "grad_norm": 0.9902074337005615,
      "learning_rate": 1.3354297693920338e-05,
      "loss": 0.114,
      "step": 1268
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9245161290322581,
      "eval_loss": 0.0809452086687088,
      "eval_runtime": 1.5761,
      "eval_samples_per_second": 1966.881,
      "eval_steps_per_second": 41.241,
      "step": 1272
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.5610605478286743,
      "learning_rate": 1.169287211740042e-05,
      "loss": 0.0779,
      "step": 1585
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.932258064516129,
      "eval_loss": 0.06483655422925949,
      "eval_runtime": 1.5806,
      "eval_samples_per_second": 1961.319,
      "eval_steps_per_second": 41.124,
      "step": 1590
    },
    {
      "epoch": 5.981132075471698,
      "grad_norm": 0.8640622496604919,
      "learning_rate": 1.0031446540880504e-05,
      "loss": 0.0623,
      "step": 1902
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9341935483870968,
      "eval_loss": 0.05672295391559601,
      "eval_runtime": 1.5839,
      "eval_samples_per_second": 1957.197,
      "eval_steps_per_second": 41.038,
      "step": 1908
    },
    {
      "epoch": 6.977987421383648,
      "grad_norm": 0.6649580001831055,
      "learning_rate": 8.370020964360588e-06,
      "loss": 0.0543,
      "step": 2219
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9370967741935484,
      "eval_loss": 0.051882579922676086,
      "eval_runtime": 1.5711,
      "eval_samples_per_second": 1973.095,
      "eval_steps_per_second": 41.371,
      "step": 2226
    },
    {
      "epoch": 7.9748427672955975,
      "grad_norm": 0.3989936113357544,
      "learning_rate": 6.708595387840672e-06,
      "loss": 0.0484,
      "step": 2536
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9383870967741935,
      "eval_loss": 0.04967164620757103,
      "eval_runtime": 1.5874,
      "eval_samples_per_second": 1952.855,
      "eval_steps_per_second": 40.947,
      "step": 2544
    },
    {
      "epoch": 8.971698113207546,
      "grad_norm": 0.5839671492576599,
      "learning_rate": 5.047169811320756e-06,
      "loss": 0.0452,
      "step": 2853
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9354838709677419,
      "eval_loss": 0.04606916755437851,
      "eval_runtime": 1.5836,
      "eval_samples_per_second": 1957.599,
      "eval_steps_per_second": 41.046,
      "step": 2862
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.3575935661792755,
      "learning_rate": 3.385744234800839e-06,
      "loss": 0.0428,
      "step": 3170
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9374193548387096,
      "eval_loss": 0.04456347972154617,
      "eval_runtime": 1.5756,
      "eval_samples_per_second": 1967.538,
      "eval_steps_per_second": 41.255,
      "step": 3180
    },
    {
      "epoch": 10.965408805031446,
      "grad_norm": 0.3484518229961395,
      "learning_rate": 1.7243186582809226e-06,
      "loss": 0.0411,
      "step": 3487
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9380645161290323,
      "eval_loss": 0.04368600994348526,
      "eval_runtime": 1.6136,
      "eval_samples_per_second": 1921.224,
      "eval_steps_per_second": 40.284,
      "step": 3498
    },
    {
      "epoch": 11.962264150943396,
      "grad_norm": 0.39103928208351135,
      "learning_rate": 6.289308176100629e-08,
      "loss": 0.0402,
      "step": 3804
    }
  ],
  "logging_steps": 317,
  "max_steps": 3816,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 12,
  "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": 989687612074644.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.46453592863533577,
    "num_train_epochs": 12,
    "temperature": 1
  }
}