{
    "epoch": 1.0,
    "eval_logits/chosen": -2.617767333984375,
    "eval_logits/rejected": -2.721874952316284,
    "eval_logps/chosen": -238.54788208007812,
    "eval_logps/rejected": -388.59033203125,
    "eval_loss": 0.20815864205360413,
    "eval_rewards/accuracies": 0.9413930773735046,
    "eval_rewards/chosen": 1.3856867551803589,
    "eval_rewards/margins": 2.292266845703125,
    "eval_rewards/rejected": -0.9065799117088318,
    "eval_runtime": 2791.9998,
    "eval_samples": 2000,
    "eval_samples_per_second": 3.403,
    "eval_steps_per_second": 0.426,
    "total_flos": 0.0,
    "train_loss": 0.33413780781687524,
    "train_runtime": 91396.7242,
    "train_samples": 61966,
    "train_samples_per_second": 2.06,
    "train_steps_per_second": 0.016
}