{ "epoch": 0.9999360981532366, "eval_logits/chosen": -3.390625, "eval_logits/rejected": -3.515625, "eval_logps/chosen": -330.0, "eval_logps/rejected": -588.0, "eval_loss": 0.302869975566864, "eval_rewards/accuracies": 0.8719757795333862, "eval_rewards/chosen": -2.078125, "eval_rewards/margins": 2.734375, "eval_rewards/rejected": -4.8125, "eval_runtime": 1880.847, "eval_samples": 55582, "eval_samples_per_second": 29.511, "eval_steps_per_second": 0.461 }