{
    "epoch": 1.0,
    "eval_logits/chosen": -2.23917293548584,
    "eval_logits/rejected": -2.2205147743225098,
    "eval_logps/chosen": -87.08130645751953,
    "eval_logps/rejected": -104.99590301513672,
    "eval_loss": 0.6725901961326599,
    "eval_rewards/accuracies": 0.3313491940498352,
    "eval_rewards/chosen": -0.05929155647754669,
    "eval_rewards/margins": 0.05310999974608421,
    "eval_rewards/rejected": -0.1124015673995018,
    "eval_runtime": 115.2153,
    "eval_samples": 2000,
    "eval_samples_per_second": 17.359,
    "eval_steps_per_second": 0.547
}