{ "epoch": 9.999098751126562, "eval_kl/chosen": 0.00043351567001082003, "eval_kl/rejected": 0.00033228081883862615, "eval_logits/chosen": -1.390625, "eval_logits/rejected": -1.359375, "eval_logps/chosen": -322.0, "eval_logps/rejected": -298.0, "eval_loss": 0.7175156474113464, "eval_nll_loss": 1.046875, "eval_ref_probs/chosen": 0.0038271306548267603, "eval_ref_probs/rejected": 0.0034569373819977045, "eval_rejected_term/max": 0.00033228081883862615, "eval_rejected_term/min": 0.00033228081883862615, "eval_rewards/accuracies": 0.5519999861717224, "eval_rewards/chosen": -0.053466796875, "eval_rewards/margins": 0.11083984375, "eval_rewards/rejected": -0.1640625, "eval_runtime": 25.3284, "eval_samples_per_second": 39.481, "eval_steps_per_second": 9.87 }