{
    "epoch": 2.999098751126561,
    "eval_kl/chosen": 0.012016227468848228,
    "eval_kl/rejected": 0.000435983354691416,
    "eval_logits/chosen": -1.9296875,
    "eval_logits/rejected": -1.984375,
    "eval_logps/chosen": -338.0,
    "eval_logps/rejected": -326.0,
    "eval_loss": 1.6166386604309082,
    "eval_nll_loss": 1.1328125,
    "eval_ref_probs/chosen": 0.0038271306548267603,
    "eval_ref_probs/rejected": 0.0034569373819977045,
    "eval_rejected_term/max": 0.000435983354691416,
    "eval_rejected_term/min": 0.000435983354691416,
    "eval_rewards/accuracies": 0.7200000286102295,
    "eval_rewards/chosen": -1.65625,
    "eval_rewards/margins": 1.2109375,
    "eval_rewards/rejected": -2.859375,
    "eval_runtime": 25.4381,
    "eval_samples_per_second": 39.311,
    "eval_steps_per_second": 9.828
}