{ "epoch": 1.0, "eval_logits/chosen": -1.4296875, "eval_logits/rejected": -1.3828125, "eval_logps/chosen": -464.0, "eval_logps/rejected": -420.0, "eval_loss": 0.6500468850135803, "eval_rewards/accuracies": 0.6170634627342224, "eval_rewards/chosen": -0.40234375, "eval_rewards/margins": 0.1513671875, "eval_rewards/rejected": -0.5546875, "eval_runtime": 52.2116, "eval_samples": 2000, "eval_samples_per_second": 38.306, "eval_steps_per_second": 1.207 }