{ "epoch": 3.0, "eval_logits/chosen": -2.0662753582000732, "eval_logits/rejected": -1.9411602020263672, "eval_logps/chosen": -266.15380859375, "eval_logps/rejected": -228.28196716308594, "eval_loss": 0.5263338685035706, "eval_rewards/accuracies": 0.7480000257492065, "eval_rewards/chosen": -0.14929771423339844, "eval_rewards/margins": 0.7504671812057495, "eval_rewards/rejected": -0.899764895439148, "eval_runtime": 443.1683, "eval_samples": 2000, "eval_samples_per_second": 4.513, "eval_steps_per_second": 0.282 }