{
    "epoch": 1.0,
    "eval_logits/chosen": -1.684370994567871,
    "eval_logits/rejected": -1.4736462831497192,
    "eval_logps/chosen": -297.6668701171875,
    "eval_logps/rejected": -289.2203674316406,
    "eval_loss": 0.032715052366256714,
    "eval_rewards/accuracies": 0.7543103694915771,
    "eval_rewards/chosen": -0.1257677674293518,
    "eval_rewards/margins": 0.30121153593063354,
    "eval_rewards/rejected": -0.42697927355766296,
    "eval_runtime": 94.8006,
    "eval_samples": 1818,
    "eval_samples_per_second": 19.177,
    "eval_steps_per_second": 0.306,
    "total_flos": 0.0,
    "train_loss": 0.13218348616853767,
    "train_runtime": 11343.5766,
    "train_samples": 55758,
    "train_samples_per_second": 4.915,
    "train_steps_per_second": 0.038
}