{
    "epoch": 0.9988571428571429,
    "eval_logits/chosen": -0.5392131209373474,
    "eval_logits/rejected": 0.8578556776046753,
    "eval_logps/chosen": -395.783447265625,
    "eval_logps/rejected": -463.0950012207031,
    "eval_loss": 0.5202247500419617,
    "eval_rewards/accuracies": 0.7715517282485962,
    "eval_rewards/chosen": -1.2017946243286133,
    "eval_rewards/margins": 1.2383939027786255,
    "eval_rewards/rejected": -2.440188407897949,
    "eval_runtime": 91.4513,
    "eval_samples": 1831,
    "eval_samples_per_second": 20.022,
    "eval_steps_per_second": 0.317,
    "total_flos": 0.0,
    "train_loss": 0.5630035629534339,
    "train_runtime": 11387.5716,
    "train_samples": 55999,
    "train_samples_per_second": 4.918,
    "train_steps_per_second": 0.038
}