{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 64.93506493506493,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 6.4935064935064934,
      "grad_norm": 0.3268238604068756,
      "learning_rate": 0.0002,
      "loss": 1.6359,
      "step": 100
    },
    {
      "epoch": 12.987012987012987,
      "grad_norm": 0.4271763563156128,
      "learning_rate": 0.00019396926207859084,
      "loss": 1.0374,
      "step": 200
    },
    {
      "epoch": 19.48051948051948,
      "grad_norm": 0.715174674987793,
      "learning_rate": 0.0001766044443118978,
      "loss": 0.6054,
      "step": 300
    },
    {
      "epoch": 25.974025974025974,
      "grad_norm": 0.6491053700447083,
      "learning_rate": 0.00015000000000000001,
      "loss": 0.3787,
      "step": 400
    },
    {
      "epoch": 32.467532467532465,
      "grad_norm": 0.6284736394882202,
      "learning_rate": 0.00011736481776669306,
      "loss": 0.2367,
      "step": 500
    },
    {
      "epoch": 38.96103896103896,
      "grad_norm": 0.5069062113761902,
      "learning_rate": 8.263518223330697e-05,
      "loss": 0.1624,
      "step": 600
    },
    {
      "epoch": 45.45454545454545,
      "grad_norm": 0.4356766641139984,
      "learning_rate": 5.000000000000002e-05,
      "loss": 0.1212,
      "step": 700
    },
    {
      "epoch": 51.94805194805195,
      "grad_norm": 0.24417874217033386,
      "learning_rate": 2.339555568810221e-05,
      "loss": 0.1002,
      "step": 800
    },
    {
      "epoch": 58.44155844155844,
      "grad_norm": 0.18362776935100555,
      "learning_rate": 6.030737921409169e-06,
      "loss": 0.0948,
      "step": 900
    },
    {
      "epoch": 64.93506493506493,
      "grad_norm": 0.2180134356021881,
      "learning_rate": 0.0,
      "loss": 0.0945,
      "step": 1000
    }
  ],
  "logging_steps": 100,
  "max_steps": 1000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 67,
  "save_steps": 100,
  "total_flos": 1.4064060677803213e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}