{ "epoch": 1.0, "eval_logits/chosen": -2.604215621948242, "eval_logits/rejected": -2.3780710697174072, "eval_logps/chosen": -260.56195068359375, "eval_logps/rejected": -1094.3392333984375, "eval_loss": 0.0027059800922870636, "eval_rewards/accuracies": 1.0, "eval_rewards/chosen": -1.9348751306533813, "eval_rewards/margins": 8.553033828735352, "eval_rewards/rejected": -10.487908363342285, "eval_runtime": 3.9022, "eval_samples": 5, "eval_samples_per_second": 1.281, "eval_steps_per_second": 0.256 }