{
    "epoch": 0.9963674104826155,
    "eval_logits/chosen": -0.7187579870223999,
    "eval_logits/rejected": 0.8491049408912659,
    "eval_logps/chosen": -712.8595581054688,
    "eval_logps/rejected": -1311.8641357421875,
    "eval_loss": 0.16912925243377686,
    "eval_rewards/accuracies": 0.9240196347236633,
    "eval_rewards/chosen": -2.9091784954071045,
    "eval_rewards/margins": 5.781402111053467,
    "eval_rewards/rejected": -8.690580368041992,
    "eval_runtime": 602.9119,
    "eval_samples": 6491,
    "eval_samples_per_second": 10.766,
    "eval_steps_per_second": 0.169,
    "total_flos": 0.0,
    "train_loss": 0.250737202167511,
    "train_runtime": 29028.2374,
    "train_samples": 123309,
    "train_samples_per_second": 4.248,
    "train_steps_per_second": 0.008
}