Model Information






Description
NO LORA,NO Embeddings, NO Quality prompt ,NO Negative Prompt
现在你可以直观的感受到这个模型的表现
Now you can intuitively feel the performance of this model.
画风提取物 Painting style extract:https://civitai.com/models/579831/kmmodelmaterialstyle?modelVersionId=646720
基础提取物 Basic extract:https://civitai.com/models/579863/kmmodelmaterialbasic?modelVersionId=646748
建议资源添加失败,请点击查看其余的素材
Failed to add suggested resources,Please click to view the remaining materials.
在经过大量尝试后我终于找到了拆解模型的正确方式。我认为这些演示可以证明我的猜想。
After numerous attempts, I finally found the correct way to disassemble the model. I believe these demonstrations can prove my hypothesis.
我在第一次上传模型前发现了一个现象,模型混合时会出现信息浓缩或缺少而无法正常生成图片,我有目的的根据这个现象进行混合。但是我早期混合的产物都很不成熟并且使用了错误的混合方式,总之我刚完成提取流程的优化,虽然这些依旧不完美。
Before my first model upload, I discovered a phenomenon where the information would condense or be missing during model blending, resulting in the inability to generate images normally. I purposefully blended based on this phenomenon. However, the products of my early blending were immature and used incorrect blending methods. In any case, I have just completed the optimization of the extraction process, although it is still not perfect.
现在我有60个画风素材和30个基础素材,虽然有部分模型的特点重叠了还有部分有着提纯不完整等问题,但数量依旧很可观。好吧,看看我的人气,我想也没多少人在乎
Now I have 60 art style materials and 30 basic materials. Although there are some overlapping characteristics in some models and some have issues with incomplete purification, the quantity is still considerable. Well, let's see my popularity, I don't think many people care.
Later, I will upload the model materials that I have disassembled.
之后我会上传我拆解的模型素材
项目进度:
Lora混合再提取:无法控制获取结果
模型细节元素提取:难以直观感受到结果,需要大量测试
模型拆解:重复素材过多,需要更新素材库。但是这样做并没有多少意义
Project Progress
lora mixed re-extraction: unable to control the acquisition results
Model detail element extraction: It is difficult to intuitively perceive the results, requiring extensive testing
Model disassembly: Excessive repetition of materials, requiring an update of the material library. However, this approach holds little significance
寻找模型素材的过程是让人感到痛苦的,因为每一个模型都很不错
KM_Model_splicing_material_demonstration
Demonstration2
Model Details
- Type
- AI Model
- Task
- text-to-image
- Subtype
- Safetensors / Checkpoint AI Model
- Created
- Updated
- June 15, 2026