{ "epoch": 1.0, "eval_logits/chosen": -2.839386224746704, "eval_logits/rejected": -2.8369719982147217, "eval_logps/chosen": -32.09164810180664, "eval_logps/rejected": -35.94130325317383, "eval_loss": 0.9567335247993469, "eval_rewards/accuracies": 0.6212624907493591, "eval_rewards/chosen": -0.08091981709003448, "eval_rewards/margins": 0.043395183980464935, "eval_rewards/rejected": -0.12431501597166061, "eval_runtime": 112.7556, "eval_samples": 343, "eval_samples_per_second": 3.042, "eval_steps_per_second": 0.381 }