{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.944,
  "eval_steps": 500,
  "global_step": 248,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.8,
      "grad_norm": 36.65541458129883,
      "learning_rate": 2.407258064516129e-06,
      "logits/chosen": NaN,
      "logits/rejected": NaN,
      "logps/chosen": -94.43272399902344,
      "logps/rejected": -152.86627197265625,
      "loss": 0.5541,
      "rewards/accuracies": 0.6068750023841858,
      "rewards/chosen": -0.667168140411377,
      "rewards/margins": 1.0208568572998047,
      "rewards/rejected": -1.688024878501892,
      "step": 50
    },
    {
      "epoch": 1.592,
      "grad_norm": 33.676387786865234,
      "learning_rate": 1.8024193548387098e-06,
      "logits/chosen": NaN,
      "logits/rejected": NaN,
      "logps/chosen": -92.77168273925781,
      "logps/rejected": -166.9850311279297,
      "loss": 0.3256,
      "rewards/accuracies": 0.7815656661987305,
      "rewards/chosen": -0.4749433994293213,
      "rewards/margins": 2.4482433795928955,
      "rewards/rejected": -2.923187017440796,
      "step": 100
    },
    {
      "epoch": 2.384,
      "grad_norm": 14.489395141601562,
      "learning_rate": 1.1975806451612903e-06,
      "logits/chosen": NaN,
      "logits/rejected": NaN,
      "logps/chosen": -94.82748413085938,
      "logps/rejected": -177.84359741210938,
      "loss": 0.2352,
      "rewards/accuracies": 0.8226010203361511,
      "rewards/chosen": -0.8623320460319519,
      "rewards/margins": 3.6384363174438477,
      "rewards/rejected": -4.500768661499023,
      "step": 150
    },
    {
      "epoch": 3.176,
      "grad_norm": 5.327777862548828,
      "learning_rate": 5.92741935483871e-07,
      "logits/chosen": NaN,
      "logits/rejected": NaN,
      "logps/chosen": -99.02391052246094,
      "logps/rejected": -195.1393280029297,
      "loss": 0.1656,
      "rewards/accuracies": 0.8529040217399597,
      "rewards/chosen": -1.0148193836212158,
      "rewards/margins": 4.8468122482299805,
      "rewards/rejected": -5.861631393432617,
      "step": 200
    }
  ],
  "logging_steps": 50,
  "max_steps": 248,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}