{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8944543828264758,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.22361359570661896,
      "grad_norm": 55.74723434448242,
      "learning_rate": 3.899821109123435e-05,
      "logits/chosen": -0.6040188074111938,
      "logits/rejected": -0.7450945973396301,
      "logps/chosen": -420.3378601074219,
      "logps/rejected": -462.8780517578125,
      "loss": 1.797,
      "rewards/accuracies": 0.6399999856948853,
      "rewards/chosen": -8.708248138427734,
      "rewards/margins": 2.4897494316101074,
      "rewards/rejected": -11.197998046875,
      "step": 500
    },
    {
      "epoch": 0.4472271914132379,
      "grad_norm": 53.70317459106445,
      "learning_rate": 2.78175313059034e-05,
      "logits/chosen": -0.5758157968521118,
      "logits/rejected": -0.6371915936470032,
      "logps/chosen": -458.7496032714844,
      "logps/rejected": -508.36602783203125,
      "loss": 2.0075,
      "rewards/accuracies": 0.6168749928474426,
      "rewards/chosen": -12.504631042480469,
      "rewards/margins": 3.4706857204437256,
      "rewards/rejected": -15.975316047668457,
      "step": 1000
    },
    {
      "epoch": 0.6708407871198568,
      "grad_norm": 24.010499954223633,
      "learning_rate": 1.6636851520572452e-05,
      "logits/chosen": -0.9147199392318726,
      "logits/rejected": -0.9995958805084229,
      "logps/chosen": -451.2334899902344,
      "logps/rejected": -513.6981811523438,
      "loss": 1.7831,
      "rewards/accuracies": 0.6547499895095825,
      "rewards/chosen": -11.545605659484863,
      "rewards/margins": 4.830613613128662,
      "rewards/rejected": -16.376218795776367,
      "step": 1500
    },
    {
      "epoch": 0.8944543828264758,
      "grad_norm": 38.50263977050781,
      "learning_rate": 5.456171735241503e-06,
      "logits/chosen": -0.8216423392295837,
      "logits/rejected": -0.8953205347061157,
      "logps/chosen": -446.2303771972656,
      "logps/rejected": -521.4434814453125,
      "loss": 1.5334,
      "rewards/accuracies": 0.6894999742507935,
      "rewards/chosen": -11.270264625549316,
      "rewards/margins": 6.051136016845703,
      "rewards/rejected": -17.321399688720703,
      "step": 2000
    }
  ],
  "logging_steps": 500,
  "max_steps": 2236,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 250,
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}