一个实验性质的产物, 不好但很有趣. 因此我上传上来. 我相信通过更详细地设计和处理能够得到更好的效果, 不过我暂时封存这个想法.该lora会生成类似塔罗牌的图片. 请将前两个提示词分别设置为 magic card, The ***, 同时建议第三个提示词设置成1girl, 例如 magic card, The light, 1girl,. 生成的图像会试图表达第二个提示词的内容. 预设图片中有十个提示词, 分别是The Mystic, The Natural,The ICE,The Dark, The MAGMA,The AIR,The WATERY, The FIERY,The LIGHT,The STORM. 然而, 我发现即使使用其他词, lora仍然会试图表达那个概念, 尽管效果不一定好. 我的示例图片中给出了三个例子, 分别是chain, hypnosis, Succubus. 经测试过多的提示词可能会混淆想要表达的概念, 此外, 也不适用于和其他lora混用, 因此, 本lora的实用性并不高, 但我认为其功能是有趣的且可拓展的. machine translationA product with experimental properties, not good but very interesting. So I uploaded it. I believe that better results can be achieved through more detailed design and processing, but I have temporarily sealed this idea.The Lora will generate images similar to tarot cards. Please set the first two prompt words to "magic card, The ****", and suggest setting the third prompt word to "1girl", such as "magic card, The light, 1girl". The generated image will attempt to express the content of the second prompt word. There are ten prompt words in the train images:The Mystic, The Natural, The ICE, The Dark, The MAGMA, The AIR, The WATERY, The FIERY, The LIGHT, The STORM. However, I found that even if other words are used, Lora still tries to express that concept, although the effect may not be good. My example images provide three examples, chain, hyposis, and Succubus. After testing, too many prompt words may confuse the intended concept and are not suitable for mixing with other Loras. Therefore, the practicality of this Lora is not high, but I believe its functionality is interesting and expandable.

Files with SHA256: 3213b55669f7e34047d7fc49b5508cced5b9296367ee23a83e428d75d5a63508