城市相关的全要素Lora，在大量的高分辨率效果图上进行分类训练，同时由多个模型融合而成，有较强的泛化能力，适用于大多数与建筑、城市设计相关的生成场景。对于特殊的场景搭配专用的lora和提示词表现会更好。

使用建议：

1. 建议生成尺寸较大的图像，长边768像素以上；

2. 可以搭配我公开的其他lora模型，但因为训练集部分的重合需要根据情况降低权重，一般不超过0.3，其他主题的lora则需要根据实际情况调整

3. 另外建议使用negative embedding 比如： 16-token-negative-deliberate-neg，realisticvision-negative-embedding

4. 实质上本模型是RealisticUrbanMix没有查分之前的版本，所以可以使用其他底模搭配RealisticUrbanMix获得相似的效果，或者直接搭配RealisticUrbanMix以达到调整风格强度的效果

The city-related full-element Lora is trained on a large number of high-resolution renderings, and is composed of multiple models at the same time. It has strong generalization ability and is suitable for most generation scenarios related to architecture and urban design. For special scenes, it will perform better with dedicated lora and prompt words.

Recommendations:

It is recommended to generate an image with a larger size, with a long side of 768 pixels or more;

It can be used with other lora models that I disclosed, but because of the overlap of the training set part, the weight needs to be reduced according to the situation, generally no more than 0.3, and the lora of other topics needs to be adjusted according to the actual situation In addition,

It is recommended to use negative embedding such as: 16-token-negative-deliberate-neg, realisticvision-negative-embedding

In essence, this model is the version before RealisticUrbanMix did not check the score, so you can use other base models with RealisticUrbanMix to achieve similar effects, or directly match with RealisticUrbanMix to achieve the effect of adjusting the style intensity