obsession

CHECKPOINT
Original


Updated:

The model already exists. If you are the creator, please contact us to claim it.
View duplicate model

Version Detail

Illustrious
1000
1
用“Euler a”会更清晰一些,用“Euler”会有绘画的凌乱美,根据画风按需选择。 正面词 Positive Prompt: masterpiece, best quality, amazing quality, very awa,absurdres,newest,very aesthetic, 反义词 Negative Prompt: bad quality,worst quality,worst detail,sketch,censor,low quality,logo,text,watermark,signature, Recommended settings: Steps: 15-35 CFG scale: 2-7 Sampler: Euler a or Euler 推荐我喜欢的分辨率:768X1344 832X1216 768X1280 704X1408 1024X1536 Model Recommended Resolution:768X1344 832X1216 768X1280 704X1408 1024X1536 小知识:eps噪声预测的问题是图像明暗度偏中性,很难画出非常暗的图片 对比度不够,XL一般都是eps,vpred噪声预测暗部表现优异,色彩对比度更加激进一些,提示词有的才会画出来,几乎不会矫正和随机,eps则反之,总之vpred比eps高一个门槛,eps模型简单好画一些,随机有惊喜,vpred模型认真写提示词会有惊喜。 Tips: The problem with eps noise prediction is that the brightness of the image tends to be neutral, making it difficult to draw very dark images. The contrast is not enough. XL usually uses eps, and vpred noise prediction performs excellently in the dark areas. The color contrast is more aggressive. Some keywords will be drawn out, and it's almost impossible to correct and randomize. Eps is the opposite. In short, vpred is one step higher than eps. The eps model is simpler and easier to draw, and there are surprises in randomness. The vpred model will have surprises if you write keywords carefully.

Project Permissions

    Use Permissions

  • Use in TENSOR Online

  • As a online training base model on TENSOR

  • Use without crediting me

  • Share merges of this model

  • Use different permissions on merges

    Commercial Use

  • Sell generated contents

  • Use on generation services

  • Sell this model or merges

Related Posts