{
    "epoch": 1.0,
    "eval_logits/chosen": -2.8302695751190186,
    "eval_logits/rejected": -2.8291618824005127,
    "eval_logps/chosen": -283.5841369628906,
    "eval_logps/rejected": -264.64422607421875,
    "eval_loss": 0.5655555129051208,
    "eval_pred_label": 0.0,
    "eval_rewards/accuracies": 0.699999988079071,
    "eval_rewards/chosen": 0.06732505559921265,
    "eval_rewards/margins": 0.6017746329307556,
    "eval_rewards/rejected": -0.534449577331543,
    "eval_runtime": 457.1465,
    "eval_samples": 2000,
    "eval_samples_per_second": 4.375,
    "eval_steps_per_second": 0.273,
    "eval_use_label": 0.0
}