{ "epoch": 3.0, "eval_logits/chosen": -2.3620400428771973, "eval_logits/rejected": -2.259032964706421, "eval_logps/chosen": -321.4129333496094, "eval_logps/rejected": -315.8341064453125, "eval_loss": 0.7384642958641052, "eval_rewards/accuracies": 0.7559999823570251, "eval_rewards/chosen": -4.756579875946045, "eval_rewards/margins": 3.8600614070892334, "eval_rewards/rejected": -8.6166410446167, "eval_runtime": 330.136, "eval_samples": 2000, "eval_samples_per_second": 6.058, "eval_steps_per_second": 0.379 }