{ "epoch": 1.0, "eval_fcm_dpo/beta": 0.0024424130097031593, "eval_logits/chosen": -0.5151445865631104, "eval_logits/rejected": -0.5001552700996399, "eval_logps/chosen": -445.4465637207031, "eval_logps/ref_chosen": -79.05104064941406, "eval_logps/ref_rejected": -86.79793548583984, "eval_logps/rejected": -608.6779174804688, "eval_loss": 0.5710760354995728, "eval_margin_dpo/margin_mean": 155.484375, "eval_margin_dpo/margin_std": 264.51043701171875, "eval_runtime": 39.1561, "eval_samples": 2339, "eval_samples_per_second": 59.735, "eval_steps_per_second": 1.89, "total_flos": 0.0, "train_loss": 0.9836619676647382, "train_runtime": 1868.3255, "train_samples": 43598, "train_samples_per_second": 23.335, "train_steps_per_second": 0.364 }