似乎只学习到了风格，而没有学习到低像素的画风，需要 pixel 来模仿。 It seems that the lora only learned the chibi style, but not the low-pixel style, need pixel to imitate it. 补充说明：似乎跟人物lora不适配，会生成一个大人和一个小人 Additional note: It seems that it is not compatible with the character Lora, and it will generate a big and a little girl再次补充：如果提示词太详细，可能不会生成小人，这个时候加大lora和chibi的权重就行了。 One more addition: If the prompt words are too detailed, the chibi girl may not appear. At this time, just increase the lora weight and chibi weight.

Files with SHA256: 1920d5abbcc602e764ddd62acdfcf64ce11ff57f9efc8b2a7aceb95b936eef09