{ "epoch": 2.999098751126561, "eval_log_ratio_diff/mean": 5.0, "eval_logits/chosen": -1.65625, "eval_logits/rejected": -1.671875, "eval_logps/chosen": -324.0, "eval_logps/rejected": -304.0, "eval_loss": 0.6349183320999146, "eval_nll_loss": 1.078125, "eval_ref_probs/chosen": 0.0038271306548267603, "eval_ref_probs/rejected": 0.0034569373819977045, "eval_rejected_term/max": -0.0005586988409049809, "eval_rejected_term/min": -0.0005586988409049809, "eval_rewards/accuracies": 0.7160000205039978, "eval_rewards/chosen": -0.25, "eval_rewards/margins": 0.5, "eval_rewards/rejected": -0.75, "eval_runtime": 25.3311, "eval_samples_per_second": 39.477, "eval_steps_per_second": 9.869 }