{
    "epoch": 0.9966024915062288,
    "eval_logits/chosen": -2.0022222995758057,
    "eval_logits/rejected": -1.888062834739685,
    "eval_logps/chosen": -342.24005126953125,
    "eval_logps/rejected": -327.4780578613281,
    "eval_loss": 0.6894429922103882,
    "eval_rewards/accuracies": 0.625,
    "eval_rewards/chosen": -0.8659123182296753,
    "eval_rewards/margins": -0.0024991072714328766,
    "eval_rewards/rejected": -0.8634131550788879,
    "eval_runtime": 1.2317,
    "eval_samples": 100,
    "eval_samples_per_second": 81.189,
    "eval_steps_per_second": 0.812
}