all_results.json

{
    "epoch": 0.9989701338825953,
    "eval_logits/chosen": -2.065095901489258,
    "eval_logits/rejected": -1.9689290523529053,
    "eval_logps/chosen": -335.3676452636719,
    "eval_logps/rejected": -427.32891845703125,
    "eval_loss": 0.5029041767120361,
    "eval_rewards/accuracies": 0.75,
    "eval_rewards/chosen": -1.1300084590911865,
    "eval_rewards/margins": 0.865437388420105,
    "eval_rewards/rejected": -1.995445728302002,
    "eval_runtime": 65.1127,
    "eval_samples_per_second": 15.358,
    "eval_steps_per_second": 0.246
}