{ "epoch": 2.0, "eval_logits/chosen": -2.3789620399475098, "eval_logits/rejected": -2.3597521781921387, "eval_logps/chosen": -80.83098602294922, "eval_logps/rejected": -96.45763397216797, "eval_loss": 0.6385183930397034, "eval_rewards/accuracies": 0.3333333432674408, "eval_rewards/chosen": -0.22302590310573578, "eval_rewards/margins": 0.31624171137809753, "eval_rewards/rejected": -0.5392675995826721, "eval_runtime": 115.9806, "eval_samples": 2000, "eval_samples_per_second": 17.244, "eval_steps_per_second": 0.543 }