{
    "epoch": 0.9981298423724285,
    "eval_logits/chosen": 0.14491169154644012,
    "eval_logits/rejected": 0.15989559888839722,
    "eval_logps/chosen": -141.19911193847656,
    "eval_logps/rejected": -146.1117401123047,
    "eval_loss": 0.6417441368103027,
    "eval_rewards/accuracies": 0.5967742204666138,
    "eval_rewards/chosen": -0.5290672183036804,
    "eval_rewards/margins": 0.06448979675769806,
    "eval_rewards/rejected": -0.5935569405555725,
    "eval_runtime": 177.5006,
    "eval_samples_per_second": 11.048,
    "eval_steps_per_second": 0.175
}