{
    "epoch": 1.971563981042654,
    "eval_logits/chosen": 153.10781860351562,
    "eval_logits/rejected": 153.006591796875,
    "eval_logps/chosen": -698.5294799804688,
    "eval_logps/rejected": -719.531005859375,
    "eval_loss": 0.4833243787288666,
    "eval_rewards/accuracies": 0.7395833134651184,
    "eval_rewards/chosen": -0.11214617639780045,
    "eval_rewards/margins": 1.1179896593093872,
    "eval_rewards/rejected": -1.2301357984542847,
    "eval_runtime": 116.3725,
    "eval_samples": 750,
    "eval_samples_per_second": 6.445,
    "eval_steps_per_second": 0.206
}