eval_results.json

{
    "epoch": 1.971563981042654,
    "eval_logits/chosen": -13.29525375366211,
    "eval_logits/rejected": -12.665611267089844,
    "eval_logps/chosen": -371.9910583496094,
    "eval_logps/rejected": -373.3665466308594,
    "eval_loss": 0.596422553062439,
    "eval_rewards/accuracies": 0.65625,
    "eval_rewards/chosen": 0.3541353642940521,
    "eval_rewards/margins": 0.24235813319683075,
    "eval_rewards/rejected": 0.11177725344896317,
    "eval_runtime": 14.5207,
    "eval_samples": 750,
    "eval_samples_per_second": 51.65,
    "eval_steps_per_second": 1.653
}