{ "epoch": 1.0, "eval_kl": 0.0, "eval_logits/chosen": -379691072.0, "eval_logits/rejected": -379400672.0, "eval_logps/chosen": -350.84775, "eval_logps/rejected": -419.8235625, "eval_loss": 0.43092331290245056, "eval_rewards/chosen": -0.629920166015625, "eval_rewards/margins": 0.89880419921875, "eval_rewards/rejected": -1.528724365234375, "eval_runtime": 171.8673, "eval_samples": 4000, "eval_samples_per_second": 23.274, "eval_steps_per_second": 0.727 }