{ "epoch": 0.9999801520354088, "eval_logits/chosen": -2.8125, "eval_logits/rejected": -3.0, "eval_logps/chosen": -692.0, "eval_logps/rejected": -1072.0, "eval_loss": 0.26600757241249084, "eval_rewards/accuracies": 0.8884963989257812, "eval_rewards/chosen": -5.0625, "eval_rewards/margins": 4.0, "eval_rewards/rejected": -9.0625, "eval_runtime": 5409.4679, "eval_samples": 182266, "eval_samples_per_second": 32.653, "eval_steps_per_second": 0.51 }