{ "epoch": 1.0, "eval_logits/chosen": -2.4783902168273926, "eval_logits/rejected": -2.4373888969421387, "eval_logps/chosen": -251.0624237060547, "eval_logps/rejected": -269.08074951171875, "eval_loss": 0.5351515412330627, "eval_rewards/accuracies": 0.76171875, "eval_rewards/chosen": 0.5765520930290222, "eval_rewards/margins": 0.8972306251525879, "eval_rewards/rejected": -0.3206784725189209, "eval_runtime": 96.1948, "eval_samples": 2000, "eval_samples_per_second": 20.791, "eval_steps_per_second": 0.333 }