eval_results.json

{
    "epoch": 0.9984301412872841,
    "eval_logits/chosen": -2.3779289722442627,
    "eval_logits/rejected": -2.5323691368103027,
    "eval_logps/chosen": -509.05767822265625,
    "eval_logps/rejected": -558.6803588867188,
    "eval_loss": 0.577161967754364,
    "eval_rewards/accuracies": 0.7063491940498352,
    "eval_rewards/chosen": -1.0429552793502808,
    "eval_rewards/margins": 0.5840887427330017,
    "eval_rewards/rejected": -1.6270442008972168,
    "eval_runtime": 88.0288,
    "eval_samples": 2000,
    "eval_samples_per_second": 22.72,
    "eval_steps_per_second": 0.716
}