{
    "epoch": 0.9981298423724285,
    "eval_logits/chosen": 0.13887186348438263,
    "eval_logits/rejected": 0.1539563536643982,
    "eval_logps/chosen": -139.28805541992188,
    "eval_logps/rejected": -144.10594177246094,
    "eval_loss": 0.6531820297241211,
    "eval_rewards/accuracies": 0.5887096524238586,
    "eval_rewards/chosen": -0.5099565386772156,
    "eval_rewards/margins": 0.06354229897260666,
    "eval_rewards/rejected": -0.5734988451004028,
    "eval_runtime": 165.6809,
    "eval_samples_per_second": 11.836,
    "eval_steps_per_second": 0.187
}