{
    "epoch": 0.9981298423724285,
    "eval_logits/chosen": -6.292795181274414,
    "eval_logits/rejected": -6.258862495422363,
    "eval_logps/chosen": -18846.923828125,
    "eval_logps/rejected": -18661.392578125,
    "eval_loss": 0.013096433132886887,
    "eval_rewards/accuracies": 0.4838709533214569,
    "eval_rewards/chosen": -187.58631896972656,
    "eval_rewards/margins": -1.8399263620376587,
    "eval_rewards/rejected": -185.7463836669922,
    "eval_runtime": 179.1207,
    "eval_samples_per_second": 10.948,
    "eval_steps_per_second": 0.173
}