{ "eval_logits/chosen": -0.5102726221084595, "eval_logits/rejected": -0.39576610922813416, "eval_logps/chosen": -266.3994445800781, "eval_logps/rejected": -356.5616760253906, "eval_loss": 6.111550331115723, "eval_rewards/chosen": -1.2513880729675293, "eval_rewards/rejected": 6.533991813659668, "eval_runtime": 11.4441, "eval_samples": 60, "eval_samples_per_second": 5.243, "eval_steps_per_second": 2.621 }