{
    "epoch": 1.0,
    "eval_logps/chosen": -37927.63671875,
    "eval_logps/rejected": -33295.5859375,
    "eval_loss": 0.013621850870549679,
    "eval_neglected": 256.0,
    "eval_rewards/accuracies": 0.454365074634552,
    "eval_rewards/chosen": -376.54638671875,
    "eval_rewards/margins": -46.1220817565918,
    "eval_rewards/rejected": -330.4242858886719,
    "eval_runtime": 968.279,
    "eval_samples": 2000,
    "eval_samples_per_second": 2.066,
    "eval_selected": 0.0,
    "eval_steps_per_second": 0.065,
    "train_loss": 0.11865393293933718,
    "train_runtime": 42177.7253,
    "train_samples": 61135,
    "train_samples_per_second": 1.449,
    "train_steps_per_second": 0.023
}