{
    "epoch": 1.0,
    "eval_logits/chosen": -0.2944760024547577,
    "eval_logits/rejected": 0.7734581232070923,
    "eval_logps/chosen": -369.4197998046875,
    "eval_logps/rejected": -469.3939208984375,
    "eval_loss": 0.5004380345344543,
    "eval_rewards/accuracies": 0.78515625,
    "eval_rewards/chosen": -1.0684473514556885,
    "eval_rewards/margins": 0.998650312423706,
    "eval_rewards/rejected": -2.0670976638793945,
    "eval_runtime": 39.5943,
    "eval_samples": 2000,
    "eval_samples_per_second": 50.512,
    "eval_steps_per_second": 0.808,
    "total_flos": 0.0,
    "train_loss": 0.5400827195355085,
    "train_runtime": 3932.0785,
    "train_samples": 61134,
    "train_samples_per_second": 15.548,
    "train_steps_per_second": 0.122
}