{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 7.861635220125786,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6780645161290323,
      "eval_loss": 0.40519121289253235,
      "eval_runtime": 2.041,
      "eval_samples_per_second": 1518.881,
      "eval_steps_per_second": 31.848,
      "step": 318
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.8578532934188843,
      "learning_rate": 1.650593990216632e-05,
      "loss": 0.6413,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8396774193548387,
      "eval_loss": 0.14012566208839417,
      "eval_runtime": 2.0415,
      "eval_samples_per_second": 1518.497,
      "eval_steps_per_second": 31.839,
      "step": 636
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8964516129032258,
      "eval_loss": 0.07267240434885025,
      "eval_runtime": 2.0914,
      "eval_samples_per_second": 1482.244,
      "eval_steps_per_second": 31.079,
      "step": 954
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.5715540051460266,
      "learning_rate": 1.3011879804332637e-05,
      "loss": 0.1635,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9145161290322581,
      "eval_loss": 0.05143408849835396,
      "eval_runtime": 2.5486,
      "eval_samples_per_second": 1216.367,
      "eval_steps_per_second": 25.504,
      "step": 1272
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.5522882342338562,
      "learning_rate": 9.517819706498952e-06,
      "loss": 0.0817,
      "step": 1500
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9235483870967742,
      "eval_loss": 0.04265019670128822,
      "eval_runtime": 2.1204,
      "eval_samples_per_second": 1461.984,
      "eval_steps_per_second": 30.655,
      "step": 1590
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9267741935483871,
      "eval_loss": 0.03797636553645134,
      "eval_runtime": 2.0762,
      "eval_samples_per_second": 1493.112,
      "eval_steps_per_second": 31.307,
      "step": 1908
    },
    {
      "epoch": 6.29,
      "grad_norm": 0.3155914843082428,
      "learning_rate": 6.02375960866527e-06,
      "loss": 0.0617,
      "step": 2000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9345161290322581,
      "eval_loss": 0.03546394035220146,
      "eval_runtime": 2.0483,
      "eval_samples_per_second": 1513.483,
      "eval_steps_per_second": 31.734,
      "step": 2226
    },
    {
      "epoch": 7.86,
      "grad_norm": 0.31727784872055054,
      "learning_rate": 2.5296995108315863e-06,
      "loss": 0.054,
      "step": 2500
    }
  ],
  "logging_steps": 500,
  "max_steps": 2862,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 651155886807636.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.980055715011544,
    "num_train_epochs": 9,
    "temperature": 2
  }
}