{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.090909090909092,
  "eval_steps": 500,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18,
      "grad_norm": 220.0,
      "learning_rate": 4e-05,
      "loss": 24.9691,
      "step": 1
    },
    {
      "epoch": 0.91,
      "grad_norm": 21.75,
      "learning_rate": 0.0002,
      "loss": 20.4692,
      "step": 5
    },
    {
      "epoch": 0.91,
      "eval_loss": 7.439934253692627,
      "eval_runtime": 0.5593,
      "eval_samples_per_second": 3.576,
      "eval_steps_per_second": 1.788,
      "step": 5
    },
    {
      "epoch": 1.82,
      "grad_norm": 22.875,
      "learning_rate": 0.00019396926207859084,
      "loss": 12.9912,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_loss": 6.607400894165039,
      "eval_runtime": 0.5606,
      "eval_samples_per_second": 3.567,
      "eval_steps_per_second": 1.784,
      "step": 11
    },
    {
      "epoch": 2.73,
      "grad_norm": 3.96875,
      "learning_rate": 0.0001766044443118978,
      "loss": 10.1734,
      "step": 15
    },
    {
      "epoch": 2.91,
      "eval_loss": 6.02254581451416,
      "eval_runtime": 0.5738,
      "eval_samples_per_second": 3.485,
      "eval_steps_per_second": 1.743,
      "step": 16
    },
    {
      "epoch": 3.64,
      "grad_norm": 13.875,
      "learning_rate": 0.00015000000000000001,
      "loss": 9.8269,
      "step": 20
    },
    {
      "epoch": 4.0,
      "eval_loss": 3.550281047821045,
      "eval_runtime": 0.5671,
      "eval_samples_per_second": 3.527,
      "eval_steps_per_second": 1.763,
      "step": 22
    },
    {
      "epoch": 4.55,
      "grad_norm": 22.5,
      "learning_rate": 0.00011736481776669306,
      "loss": 5.2353,
      "step": 25
    },
    {
      "epoch": 4.91,
      "eval_loss": 1.6504679918289185,
      "eval_runtime": 0.5777,
      "eval_samples_per_second": 3.462,
      "eval_steps_per_second": 1.731,
      "step": 27
    },
    {
      "epoch": 5.45,
      "grad_norm": 3.078125,
      "learning_rate": 8.263518223330697e-05,
      "loss": 1.6367,
      "step": 30
    },
    {
      "epoch": 6.0,
      "eval_loss": 1.4912230968475342,
      "eval_runtime": 0.5649,
      "eval_samples_per_second": 3.54,
      "eval_steps_per_second": 1.77,
      "step": 33
    },
    {
      "epoch": 6.36,
      "grad_norm": 1.109375,
      "learning_rate": 5.000000000000002e-05,
      "loss": 1.4714,
      "step": 35
    },
    {
      "epoch": 6.91,
      "eval_loss": 1.420060157775879,
      "eval_runtime": 0.5812,
      "eval_samples_per_second": 3.441,
      "eval_steps_per_second": 1.72,
      "step": 38
    },
    {
      "epoch": 7.27,
      "grad_norm": 1.1171875,
      "learning_rate": 2.339555568810221e-05,
      "loss": 1.3916,
      "step": 40
    },
    {
      "epoch": 8.0,
      "eval_loss": 1.3933132886886597,
      "eval_runtime": 0.5657,
      "eval_samples_per_second": 3.536,
      "eval_steps_per_second": 1.768,
      "step": 44
    },
    {
      "epoch": 8.18,
      "grad_norm": 0.95703125,
      "learning_rate": 6.030737921409169e-06,
      "loss": 1.2832,
      "step": 45
    },
    {
      "epoch": 8.91,
      "eval_loss": 1.3881534337997437,
      "eval_runtime": 0.5859,
      "eval_samples_per_second": 3.414,
      "eval_steps_per_second": 1.707,
      "step": 49
    },
    {
      "epoch": 9.09,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0,
      "loss": 1.2863,
      "step": 50
    },
    {
      "epoch": 9.09,
      "eval_loss": 1.3843274116516113,
      "eval_runtime": 0.5652,
      "eval_samples_per_second": 3.539,
      "eval_steps_per_second": 1.769,
      "step": 50
    },
    {
      "epoch": 9.09,
      "step": 50,
      "total_flos": 7.66894376306606e+16,
      "train_loss": 6.66651237487793,
      "train_runtime": 191.2414,
      "train_samples_per_second": 4.602,
      "train_steps_per_second": 0.261
    }
  ],
  "logging_steps": 5,
  "max_steps": 50,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "total_flos": 7.66894376306606e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}