{ "epoch": 1.0, "eval_logits/chosen": -0.5923872590065002, "eval_logits/rejected": -0.5646775364875793, "eval_logps/chosen": -106.05509948730469, "eval_logps/ref_chosen": -79.05104064941406, "eval_logps/ref_rejected": -86.79793548583984, "eval_logps/rejected": -135.59478759765625, "eval_loss": 0.40463200211524963, "eval_margin_dpo/beta": 0.10000000149011612, "eval_margin_dpo/beta_margin_grad_mean": -0.25669676065444946, "eval_margin_dpo/beta_margin_grad_std": 0.2539878487586975, "eval_margin_dpo/beta_margin_mean": 2.1792807579040527, "eval_margin_dpo/beta_margin_std": 2.6363675594329834, "eval_margin_dpo/loss_margin_mean": 21.79280662536621, "eval_margin_dpo/margin_mean": 21.79280662536621, "eval_margin_dpo/margin_std": 26.363676071166992, "eval_runtime": 40.0533, "eval_samples": 2339, "eval_samples_per_second": 58.397, "eval_steps_per_second": 1.848 }