V6：

新增可用场景：

fire,universe,sea,waves

/////////

这个lora来源于一张歪出来的图，然后通过自我迭代，目前第五版LORA算是可以拿出来看了。

lora效果：让立体的风景出现在女孩的衣服上。

使用方法：

trigger words:scenery_on_clothes。然后使用风景，1girl

可以通过arm,leg,feet,bare shoulders使女孩的手脚出现。

可以根据权重调整风景化的覆盖力度，一般0.5-0.9。我使用anything4.5，chilloutmix也可以。

请注意Hires. Fix的Denoising，如果你需要多风景化，可以加大，但是不建议超过0.7，太高了会出现畸形。如果只是想维持目前的程度单纯提高分辨率，Denoising建议低于0.4。

可以换衣服，目前试出来的衣服：dress,bikini,swimsuit

目前试出来有显著特征的风景：forest,house,building,ruins,tower,chinese town

相信这个lora还有更大的潜力未被挖掘。

欢迎尝试和上传图片，这个lora会继续迭代，我希望以后能有更大的泛用性。

不如点个like或者follow吧，更新的时候你会收到通知：）

This LORA originated from a distorted image and after self-iteration, the current fifth version of LORA is presentable.

The effect of LORA is to make a three-dimensional landscape appear on a girl's clothes.

Usage:

Trigger words: scenery_on_clothes. Then use scenery, 1girl.

You can make the girl's hands and feet appear by using arm, leg, feet, bare shoulders.

You can adjust the coverage intensity of scenery overlay by weight, usually between 0.5-0.9. I use anything4.5 and chilloutmix, but other models may work as well.

Please note the denoising of Hires. Fix. If you need more scenery, you can increase it, but it is not recommended to exceed 0.7, which may cause deformities. If you only want to maintain the current level and simply increase the resolution, the denoising is recommended to be less than 0.4.

You can change clothes. The clothes that have been tried are: dress, bikini, swimsuit.

Scenery with significant features that have been tried: forest, house, building, ruins, tower, Chinese town.

I believe this LORA still has untapped potential.

Feel free to try and upload images. This LORA will continue to iterate, and I hope it can have greater versatility in the future.

Why not give it a like or follow? You will receive notifications when it is updated :)