{
    "epoch": 1.0,
    "eval_logits/chosen": -2.406076192855835,
    "eval_logits/rejected": -2.390552282333374,
    "eval_logps/chosen": -76.63563537597656,
    "eval_logps/rejected": -89.79125213623047,
    "eval_loss": 0.635656476020813,
    "eval_rewards/accuracies": 0.335317462682724,
    "eval_rewards/chosen": -0.07570639997720718,
    "eval_rewards/margins": 0.23013043403625488,
    "eval_rewards/rejected": -0.30583682656288147,
    "eval_runtime": 113.694,
    "eval_samples": 2000,
    "eval_samples_per_second": 17.591,
    "eval_steps_per_second": 0.554,
    "total_flos": 0.0,
    "train_loss": 0.6540692721803983,
    "train_runtime": 977.0227,
    "train_samples": 6113,
    "train_samples_per_second": 6.257,
    "train_steps_per_second": 0.098
}