Sign In

Ukiyo-e FuYue Style Background-Mix

487
2.5k
6
Verified:
SafeTensor
Type
LoRA
Stats
2,270
Reviews
Published
Mar 27, 2023
Base Model
Other
Training
Epochs: 10
Trigger Words
none
Hash
AutoV2
51186E88B3
default creator card background decoration
LatentCat's Avatar
LatentCat

这是一个由富岳三十六景为基础训练的lora,我预期提取其中对于浪花,山,等景色的风格描述和色彩,并应用到其他模型的背景中去,无需提示词,详细的参考设置可以参考上传的图片信息,和对比长图,不同的层有不同的创作力,所以我没有拆分合并,你可以使用lora weight block插件自行拆层,期待你的作品。

参考图使用的模型来自https://civitai.com/models/19809/scmix,这是一个有着优秀的光影和泛化能力的合并模型。

The reference image uses a model from https://civitai.com/models/19809/scmix, which is a combined model with excellent lighting and colorization capabilities.

This is a Lora model trained based on the "Thirty-six Views of Mount Fuji" by Hokusai. I aim to extract the style descriptions and colors for scenes such as waves and mountains and apply them to the backgrounds of other models. There is no need for prompts. Please refer to the uploaded image information and the comparison chart for detailed reference settings. Different layers have different creative strengths, so I did not split or merge them. You can use the Lora weight block plugin to split the layers yourself. I am looking forward to your work.

Key Points

  • This is a Lora model trained based on "Thirty-six Views of Mount Fuji" by Hokusai.

  • The aim is to extract the style descriptions and colors for scenes such as waves and mountains and apply them to other models' backgrounds.

  • No need for prompts.

  • Refer to the uploaded image information and comparison chart for reference settings.

  • Different layers have different creative strengths, so they were not split or merged.

  • The Lora weight block plugin can be used to split the layers.

Explanation

The paragraph introduces a Lora model trained based on "Thirty-six Views of Mount Fuji" by Hokusai. The aim of the model is to extract the style descriptions and colors for scenes such as waves and mountains and apply them to other models' backgrounds. No prompts are necessary. The uploaded image information and comparison chart should be used for detailed reference settings. Different layers have different creative strengths, so the author did not split or merge them. The Lora weight block plugin can be used to split the layers. The author is looking forward to the results of this work.

more info and pic look here:https://shiny-stranger-af6.notion.site/lora-1be1e664868e4609aeee1fe920cedb8b