{
    "epoch": 1.0,
    "eval_logits/chosen": -1.2603442668914795,
    "eval_logits/rejected": -1.2824182510375977,
    "eval_logps/chosen": -316.86859130859375,
    "eval_logps/rejected": -325.4167785644531,
    "eval_loss": 0.6635114550590515,
    "eval_rewards/accuracies": 0.6796875,
    "eval_rewards/chosen": 0.04844370484352112,
    "eval_rewards/margins": 0.0758848488330841,
    "eval_rewards/rejected": -0.02744114398956299,
    "eval_runtime": 91.2633,
    "eval_samples": 2000,
    "eval_samples_per_second": 21.915,
    "eval_steps_per_second": 0.351,
    "total_flos": 0.0,
    "train_loss": 0.6731121569996599,
    "train_runtime": 6739.8654,
    "train_samples": 61134,
    "train_samples_per_second": 9.071,
    "train_steps_per_second": 0.035
}