{ "epoch": 1.0, "eval_logits/chosen": 0.675320029258728, "eval_logits/rejected": 1.6118428707122803, "eval_logps/chosen": -382.88262939453125, "eval_logps/rejected": -487.1402893066406, "eval_loss": 0.4968120753765106, "eval_rewards/accuracies": 0.76171875, "eval_rewards/chosen": -1.2028917074203491, "eval_rewards/margins": 1.0418392419815063, "eval_rewards/rejected": -2.2447309494018555, "eval_runtime": 96.2043, "eval_samples": 2000, "eval_samples_per_second": 20.789, "eval_steps_per_second": 0.333 }