eval_results.json

{
    "epoch": 0.9989462592202318,
    "eval_logits/chosen": -2.096877098083496,
    "eval_logits/rejected": -1.679472804069519,
    "eval_logps/chosen": -760.1741333007812,
    "eval_logps/rejected": -1507.9508056640625,
    "eval_loss": 0.20055127143859863,
    "eval_rewards/accuracies": 0.8687499761581421,
    "eval_rewards/chosen": -4.000043869018555,
    "eval_rewards/margins": 7.391962051391602,
    "eval_rewards/rejected": -11.392005920410156,
    "eval_runtime": 296.2738,
    "eval_samples": 6392,
    "eval_samples_per_second": 21.575,
    "eval_steps_per_second": 0.338
}