all_results.json

{
    "epoch": 0.9999343530493009,
    "eval_logits/chosen": -0.8205632567405701,
    "eval_logits/rejected": -0.9320290684700012,
    "eval_logps/chosen": -104.04210662841797,
    "eval_logps/rejected": -255.28575134277344,
    "eval_loss": 0.000569345080293715,
    "eval_rewards/accuracies": 1.0,
    "eval_rewards/chosen": -3.6049818992614746,
    "eval_rewards/margins": 14.385939598083496,
    "eval_rewards/rejected": -17.990922927856445,
    "eval_runtime": 161.6764,
    "eval_samples_per_second": 33.258,
    "eval_steps_per_second": 4.163
}