{
    "epoch": 1.0,
    "eval_logits/chosen": -2.5527360439300537,
    "eval_logits/rejected": -2.5362329483032227,
    "eval_logps/chosen": -73.02584075927734,
    "eval_logps/rejected": -80.80119323730469,
    "eval_loss": 0.6450211405754089,
    "eval_rewards/accuracies": 0.3531745970249176,
    "eval_rewards/chosen": 0.6594291925430298,
    "eval_rewards/margins": 0.5122048258781433,
    "eval_rewards/rejected": 0.14722436666488647,
    "eval_runtime": 113.9424,
    "eval_samples": 2000,
    "eval_samples_per_second": 17.553,
    "eval_steps_per_second": 0.553,
    "total_flos": 0.0,
    "train_loss": 0.6522094729885407,
    "train_runtime": 3372.3295,
    "train_samples": 18340,
    "train_samples_per_second": 5.438,
    "train_steps_per_second": 0.085
}