{
    "epoch": 1.0,
    "eval_logits/chosen": 0.9167268872261047,
    "eval_logits/rejected": 2.964365243911743,
    "eval_logps/chosen": -579.6903076171875,
    "eval_logps/rejected": -678.6295776367188,
    "eval_loss": 0.5096524953842163,
    "eval_rewards/accuracies": 0.7284482717514038,
    "eval_rewards/chosen": -2.956843376159668,
    "eval_rewards/margins": 1.3729774951934814,
    "eval_rewards/rejected": -4.3298211097717285,
    "eval_runtime": 91.8301,
    "eval_samples": 1818,
    "eval_samples_per_second": 19.797,
    "eval_steps_per_second": 0.316,
    "total_flos": 0.0,
    "train_loss": 0.5659637576943144,
    "train_runtime": 11398.0027,
    "train_samples": 55758,
    "train_samples_per_second": 4.892,
    "train_steps_per_second": 0.038
}