{ "epoch": 0.9999360981532366, "eval_logits/chosen": -2.625, "eval_logits/rejected": -2.515625, "eval_logps/chosen": -496.0, "eval_logps/rejected": -788.0, "eval_loss": 0.27973371744155884, "eval_rewards/accuracies": 0.8794642686843872, "eval_rewards/chosen": -3.078125, "eval_rewards/margins": 3.125, "eval_rewards/rejected": -6.21875, "eval_runtime": 1539.2648, "eval_samples": 55582, "eval_samples_per_second": 36.06, "eval_steps_per_second": 0.564 }