{
    "epoch": 1.0,
    "eval_logits/chosen": -0.68359375,
    "eval_logits/rejected": -0.7265625,
    "eval_logps/chosen": -284.0,
    "eval_logps/rejected": -310.0,
    "eval_loss": 0.6820937395095825,
    "eval_rewards/accuracies": 0.390625,
    "eval_rewards/chosen": -0.10009765625,
    "eval_rewards/margins": 0.0028076171875,
    "eval_rewards/rejected": -0.1025390625,
    "eval_runtime": 8.7282,
    "eval_samples_per_second": 114.571,
    "eval_steps_per_second": 1.833
}