{ "epoch": 0.9999813887699838, "eval_logits/chosen": -3.890625, "eval_logits/rejected": -3.71875, "eval_logps/chosen": -912.0, "eval_logps/rejected": -1020.0, "eval_loss": 0.4581940472126007, "eval_rewards/accuracies": 0.7553164958953857, "eval_rewards/chosen": -7.09375, "eval_rewards/margins": 1.40625, "eval_rewards/rejected": -8.5, "eval_runtime": 6074.0773, "eval_samples": 191184, "eval_samples_per_second": 31.455, "eval_steps_per_second": 0.492 }