all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -2.563148260116577,
    "eval_logits/rejected": -2.622511386871338,
    "eval_logps/chosen": -20.129846572875977,
    "eval_logps/rejected": -27.357025146484375,
    "eval_loss": 0.761045515537262,
    "eval_nll_loss": 0.0,
    "eval_rewards/accuracies": 0.7110000252723694,
    "eval_rewards/chosen": -2.01298451423645,
    "eval_rewards/margins": 0.7227182984352112,
    "eval_rewards/rejected": -2.7357029914855957,
    "eval_runtime": 34.5912,
    "eval_samples_per_second": 28.909,
    "eval_steps_per_second": 3.614
}