{ "epoch": 1.0, "eval_logits/chosen": -0.6029719114303589, "eval_logits/rejected": -0.5759690403938293, "eval_logps/chosen": -106.27237701416016, "eval_logps/ref_chosen": -79.05104064941406, "eval_logps/ref_rejected": -86.79793548583984, "eval_logps/rejected": -135.82232666015625, "eval_loss": 0.40388280153274536, "eval_margin_dpo/beta": 0.10000000149011612, "eval_margin_dpo/beta_margin_grad_mean": -0.2565760016441345, "eval_margin_dpo/beta_margin_grad_std": 0.2537597417831421, "eval_margin_dpo/beta_margin_mean": 2.1803061962127686, "eval_margin_dpo/loss_margin_mean": 21.803062438964844, "eval_margin_dpo/margin_mean": 21.803062438964844, "eval_margin_dpo/margin_std": 26.34841537475586, "eval_runtime": 39.8235, "eval_samples": 2339, "eval_samples_per_second": 58.734, "eval_steps_per_second": 1.858 }