{
    "epoch": 0.9991111111111111,
    "eval_logits/chosen": 13.90869140625,
    "eval_logits/rejected": 12.107211112976074,
    "eval_logps/chosen": -421.04986572265625,
    "eval_logps/rejected": -403.05950927734375,
    "eval_loss": 0.6638615727424622,
    "eval_rewards/accuracies": 0.7420634627342224,
    "eval_rewards/chosen": -0.12479932606220245,
    "eval_rewards/margins": 0.06845071166753769,
    "eval_rewards/rejected": -0.19325004518032074,
    "eval_runtime": 89.6685,
    "eval_samples": 1000,
    "eval_samples_per_second": 11.152,
    "eval_steps_per_second": 0.703,
    "total_flos": 0.0,
    "train_loss": 0.6916975294143703,
    "train_runtime": 7520.1237,
    "train_samples": 36000,
    "train_samples_per_second": 4.787,
    "train_steps_per_second": 0.075
}