{ "epoch": 1.0, "eval_logits/chosen": 98.7645034790039, "eval_logits/rejected": 98.74665832519531, "eval_logps/chosen": -32.56877517700195, "eval_logps/rejected": -36.22236251831055, "eval_loss": 0.7821906208992004, "eval_rewards/accuracies": 0.52491694688797, "eval_rewards/chosen": -0.10047715157270432, "eval_rewards/margins": 0.1041894406080246, "eval_rewards/rejected": -0.20466658473014832, "eval_runtime": 103.6205, "eval_samples": 343, "eval_samples_per_second": 3.31, "eval_steps_per_second": 0.415 }