{
    "epoch": 1.0,
    "eval_logits/chosen": -1.9163823127746582,
    "eval_logits/rejected": -1.5818859338760376,
    "eval_logps/chosen": -168.4323272705078,
    "eval_logps/rejected": -2053.068603515625,
    "eval_loss": 0.04686330258846283,
    "eval_rewards/accuracies": 0.875,
    "eval_rewards/chosen": -0.778435230255127,
    "eval_rewards/margins": 13.74771499633789,
    "eval_rewards/rejected": -14.52614974975586,
    "eval_runtime": 4.8308,
    "eval_samples": 24,
    "eval_samples_per_second": 4.968,
    "eval_steps_per_second": 0.207,
    "train_loss": 0.23545521443209996,
    "train_runtime": 6883.5239,
    "train_samples": 20968,
    "train_samples_per_second": 3.046,
    "train_steps_per_second": 0.048
}