{ "epoch": 1.0, "eval_logits/chosen": -2.7366418838500977, "eval_logits/rejected": -2.709143877029419, "eval_logps/chosen": -241.407470703125, "eval_logps/rejected": -256.4153747558594, "eval_loss": 0.4928562045097351, "eval_rewards/accuracies": 0.734375, "eval_rewards/chosen": 21.185983657836914, "eval_rewards/margins": 14.934151649475098, "eval_rewards/rejected": 6.251833915710449, "eval_runtime": 96.1252, "eval_samples": 2000, "eval_samples_per_second": 20.806, "eval_steps_per_second": 0.333 }