{ "epoch": 1.0, "eval_logits/chosen": 0.6094364523887634, "eval_logits/rejected": 1.5124281644821167, "eval_logps/chosen": -364.74273681640625, "eval_logps/rejected": -463.2261962890625, "eval_loss": 0.5025366544723511, "eval_rewards/accuracies": 0.77734375, "eval_rewards/chosen": -1.021127700805664, "eval_rewards/margins": 0.9845134615898132, "eval_rewards/rejected": -2.005641222000122, "eval_runtime": 109.2669, "eval_samples": 2000, "eval_samples_per_second": 18.304, "eval_steps_per_second": 0.293 }