{ "epoch": 1.0, "eval_logits/chosen": -0.8487240672111511, "eval_logits/rejected": -0.7531492114067078, "eval_logps/chosen": -394.66204833984375, "eval_logps/rejected": -365.29083251953125, "eval_loss": 0.053853537887334824, "eval_rewards/accuracies": 0.429500013589859, "eval_rewards/chosen": 0.0016431997064501047, "eval_rewards/margins": -0.00030138425063341856, "eval_rewards/rejected": 0.0019445838406682014, "eval_runtime": 545.5213, "eval_samples": 2000, "eval_samples_per_second": 3.666, "eval_steps_per_second": 0.917 }