all_results.json

{
    "epoch": 11.984422327996539,
    "eval_logits/chosen": -0.263671875,
    "eval_logits/rejected": -0.1396484375,
    "eval_logps/chosen": -280.0,
    "eval_logps/rejected": -364.0,
    "eval_loss": 0.18390525877475739,
    "eval_rewards/accuracies": 0.9897540807723999,
    "eval_rewards/chosen": 3.3125,
    "eval_rewards/margins": 12.5,
    "eval_rewards/rejected": -9.1875,
    "eval_runtime": 105.6283,
    "eval_samples_per_second": 36.846,
    "eval_steps_per_second": 1.155
}