{ "epoch": 1.0, "eval_logits/chosen": -1.2524774074554443, "eval_logits/rejected": -1.1317628622055054, "eval_logps/chosen": -545.8759765625, "eval_logps/rejected": -635.6185302734375, "eval_loss": 0.4863443076610565, "eval_rewards/accuracies": 0.7394999861717224, "eval_rewards/chosen": -2.812239646911621, "eval_rewards/margins": 1.0978591442108154, "eval_rewards/rejected": -3.9100985527038574, "eval_runtime": 384.8212, "eval_samples": 2000, "eval_samples_per_second": 5.197, "eval_steps_per_second": 0.65 }