{ "epoch": 2.999098751126561, "eval_log_ratio_diff/mean": 0.5859375, "eval_logits/chosen": -2.65625, "eval_logits/rejected": -2.609375, "eval_logps/chosen": -1.984375, "eval_logps/rejected": -2.78125, "eval_loss": 4.600971221923828, "eval_nll_loss": 1.984375, "eval_ref_probs/chosen": 0.39426562190055847, "eval_ref_probs/rejected": 0.3336612582206726, "eval_rejected_term/max": -0.09504696726799011, "eval_rejected_term/min": -0.09504696726799011, "eval_rewards/accuracies": 0.7480000257492065, "eval_rewards/chosen": -0.09326171875, "eval_rewards/margins": 0.05859375, "eval_rewards/rejected": -0.1513671875, "eval_runtime": 25.2726, "eval_samples_per_second": 39.568, "eval_steps_per_second": 9.892 }