从我以前炼的一个模型（https://huggingface.co/meishi/AiMeishi_cnXianxia/tree/main）中提取出来的Lora，效果可能比大模型稍微差一点，但从体积上看性价比颇高。

The Lora model, extracted from a model I trained previously (https://huggingface.co/meishi/AiMeishi_cnXianxia/tree/main), may have slightly lower performance than the large model, but it is quite cost-effective in terms of size.