{ "epoch": 0.9989528795811519, "eval_logits/chosen": -0.843854546546936, "eval_logits/rejected": -0.8338058590888977, "eval_logps/chosen": -410.1976013183594, "eval_logps/ref_chosen": -280.7076110839844, "eval_logps/ref_rejected": -280.1632385253906, "eval_logps/rejected": -484.2484130859375, "eval_loss": 0.5346022844314575, "eval_margin_dpo/margin_mean": 74.59521484375, "eval_margin_dpo/margin_std": 98.45130157470703, "eval_runtime": 50.8318, "eval_samples": 2000, "eval_samples_per_second": 39.345, "eval_steps_per_second": 0.63 }