General-Bottle

General-Bottle

<Gallery />

Model description

We think that when we train some general bottle images, and they have same height and width. Which means we control input image parameters(only same bottle with different color and patterns) and we get one Lora which will not generate other bottles. They have similar shape, height and width. We tested our Lora and we found we could change some color prompts and got expected results. We could use this way to train some images. The result is that you could change one fixed style image but you do not need to redesign this.

我们将同样的瓶身固定，瓶身形状大小不变，只是瓶身颜色和图案不一致。我们输入一个关键提示词，将得到一个稳定的瓶身图案，大小一样，但瓶身图案颜色不一致。这也属于控制变量法，经过测试得到的结果还不错。因为一些原因，我们上传的图案都是AI生成的，但大家可以借鉴这个思路来训练模型。简单而言，你想修改一种固定的设计风格，但是又不需要重新绘制。

Trigger words

You should use Chinese-liquor-bottle to trigger the image generation.

Download model

Download them in the Files & versions tab.