{
    "epoch": 1.0,
    "eval_logits/chosen": -2.818209171295166,
    "eval_logits/rejected": -2.815542459487915,
    "eval_logps/chosen": -31.4188232421875,
    "eval_logps/rejected": -35.04113006591797,
    "eval_loss": 23.746549606323242,
    "eval_rewards/accuracies": 0.5801494717597961,
    "eval_rewards/chosen": -0.013637068681418896,
    "eval_rewards/margins": 0.020660024136304855,
    "eval_rewards/rejected": -0.03429709002375603,
    "eval_runtime": 112.6126,
    "eval_samples": 343,
    "eval_samples_per_second": 3.046,
    "eval_steps_per_second": 0.382
}