{ "epoch": 3.0, "eval_logits/chosen": -2.470487117767334, "eval_logits/rejected": -2.4498531818389893, "eval_logps/chosen": -86.45170593261719, "eval_logps/rejected": -97.09423065185547, "eval_loss": 0.8855458498001099, "eval_rewards/accuracies": 0.341269850730896, "eval_rewards/chosen": -3.889953851699829, "eval_rewards/margins": 1.3026951551437378, "eval_rewards/rejected": -5.1926493644714355, "eval_runtime": 137.5858, "eval_samples": 2000, "eval_samples_per_second": 14.536, "eval_steps_per_second": 0.458 }