In terms of results: portrait > upper body > full body. Adding ADetailer improves facial details.

The clothing section may be less accurate, and the shoes—due to insufficient original training images—will be even less precise.

效果来说，肖像>上半身>全身。脸部需要加ADetailer效果比较好。

服装部分会弱一些，鞋子部分因为原始训练图就比较缺少，所以也比较弱。