{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.71313672922252,
  "eval_steps": 500,
  "global_step": 6500,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.6702412868632708,
      "grad_norm": 17.923147201538086,
      "learning_rate": 1.3001640168623715e-05,
      "loss": 0.5514,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7683539986610413,
      "eval_loss": 0.46452510356903076,
      "eval_runtime": 8.9706,
      "eval_samples_per_second": 332.531,
      "eval_steps_per_second": 20.846,
      "step": 746
    },
    {
      "epoch": 1.3404825737265416,
      "grad_norm": 19.32151222229004,
      "learning_rate": 1.2067614294440976e-05,
      "loss": 0.3969,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7643312215805054,
      "eval_loss": 0.5671606063842773,
      "eval_runtime": 9.0896,
      "eval_samples_per_second": 328.177,
      "eval_steps_per_second": 20.573,
      "step": 1492
    },
    {
      "epoch": 2.0107238605898123,
      "grad_norm": 17.702476501464844,
      "learning_rate": 1.1133588420258238e-05,
      "loss": 0.2998,
      "step": 1500
    },
    {
      "epoch": 2.680965147453083,
      "grad_norm": 6.516891956329346,
      "learning_rate": 1.0199562546075501e-05,
      "loss": 0.1319,
      "step": 2000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7770700454711914,
      "eval_loss": 0.7936307191848755,
      "eval_runtime": 9.0031,
      "eval_samples_per_second": 331.332,
      "eval_steps_per_second": 20.771,
      "step": 2238
    },
    {
      "epoch": 3.351206434316354,
      "grad_norm": 0.19944968819618225,
      "learning_rate": 9.265536671892763e-06,
      "loss": 0.0932,
      "step": 2500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7750586867332458,
      "eval_loss": 1.1848183870315552,
      "eval_runtime": 9.064,
      "eval_samples_per_second": 329.104,
      "eval_steps_per_second": 20.631,
      "step": 2984
    },
    {
      "epoch": 4.021447721179625,
      "grad_norm": 0.03240065276622772,
      "learning_rate": 8.331510797710023e-06,
      "loss": 0.0618,
      "step": 3000
    },
    {
      "epoch": 4.6916890080428955,
      "grad_norm": 23.173229217529297,
      "learning_rate": 7.397484923527286e-06,
      "loss": 0.0314,
      "step": 3500
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7707006335258484,
      "eval_loss": 1.372557520866394,
      "eval_runtime": 9.0505,
      "eval_samples_per_second": 329.594,
      "eval_steps_per_second": 20.662,
      "step": 3730
    },
    {
      "epoch": 5.361930294906166,
      "grad_norm": 0.01009325310587883,
      "learning_rate": 6.463459049344548e-06,
      "loss": 0.0281,
      "step": 4000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.777405321598053,
      "eval_loss": 1.31714928150177,
      "eval_runtime": 9.0862,
      "eval_samples_per_second": 328.299,
      "eval_steps_per_second": 20.581,
      "step": 4476
    },
    {
      "epoch": 6.032171581769437,
      "grad_norm": 0.6762986779212952,
      "learning_rate": 5.529433175161809e-06,
      "loss": 0.0202,
      "step": 4500
    },
    {
      "epoch": 6.702412868632708,
      "grad_norm": 0.610667884349823,
      "learning_rate": 4.595407300979072e-06,
      "loss": 0.0144,
      "step": 5000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7767348289489746,
      "eval_loss": 1.4879707098007202,
      "eval_runtime": 8.9951,
      "eval_samples_per_second": 331.626,
      "eval_steps_per_second": 20.789,
      "step": 5222
    },
    {
      "epoch": 7.372654155495979,
      "grad_norm": 0.01202669832855463,
      "learning_rate": 3.6613814267963338e-06,
      "loss": 0.0084,
      "step": 5500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7763996124267578,
      "eval_loss": 1.6179240942001343,
      "eval_runtime": 9.0802,
      "eval_samples_per_second": 328.518,
      "eval_steps_per_second": 20.594,
      "step": 5968
    },
    {
      "epoch": 8.04289544235925,
      "grad_norm": 0.007873074151575565,
      "learning_rate": 2.7273555526135954e-06,
      "loss": 0.0046,
      "step": 6000
    },
    {
      "epoch": 8.71313672922252,
      "grad_norm": 0.029083162546157837,
      "learning_rate": 1.7933296784308572e-06,
      "loss": 0.0044,
      "step": 6500
    }
  ],
  "logging_steps": 500,
  "max_steps": 7460,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1.5429010800414744e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.3935666042806453e-05,
    "per_device_train_batch_size": 16
  }
}