{ "epoch": 2.9906542056074765, "eval_logits/chosen": -2.1246237754821777, "eval_logits/rejected": -2.0873403549194336, "eval_logps/chosen": -237.59535217285156, "eval_logps/rejected": -242.21408081054688, "eval_loss": 0.5930217504501343, "eval_rewards/accuracies": 0.7916666865348816, "eval_rewards/chosen": 0.5294687747955322, "eval_rewards/margins": 2.4190399646759033, "eval_rewards/rejected": -1.8895710706710815, "eval_runtime": 153.3576, "eval_samples_per_second": 14.867, "eval_steps_per_second": 0.235 }