Sign In

ZHMix-Fantasy

502
3.3k
70
Verified:
SafeTensor
Type
Checkpoint Merge
Stats
790
Reviews
Published
Mar 20, 2024
Base Model
SD 1.5
Training
Epochs: 1
Hash
AutoV2
8B21C77C69

Read the info below to get the high quality images

(click on show more)

阅读以下信息以获得高质量图像

(点击显示更多)

ZHMix-Fantasy-V4.0:

This version reduces the requirements for prompt words and only requires simple prompt words to have rich picture effects.

这个版本降低了对提示词的要求,只需要简单的提示词就可以拥有丰富的画面效果。

ZHMix-Fantasy-V3.0:

The style of this version is once again close to the 2.5D direction. Because of the Illustration model, it can meet the demand for 2D style, so this Fantasy model changes to the impasto style and NIJI style.

该版本的风格再一次靠近2.5D方向,因为有了Illustration这个模型,可以满足对2D风格的需求,所以这个Fantasy模型转变偏向与厚涂和NIJI风格。

--------------------

ZHMix-Fantasy-V2.0:

This version has fixed the color tendency problem, the painting style is more 2.5D, the texture has been significantly improved, and the support for LORA will be relatively stable.

该版本修复了色彩倾向问题,绘画风格更加2.5D,纹理有了明显提升,对LORA的支持也会相对稳定。

--------------------

ZHMix-Fantasy-V1.0:

This model has abandoned the realistic style and leans more towards a painterly texture, though it still retains a 2.5D art style to some extent. It significantly increases exploration in the direction of stylization. It excels in portraying flowing hair and clothing wrinkles, and it's particularly adept at creating lighting and atmospheric effects. However, its support for Lora may be somewhat limited. Please adjust the weight accordingly, as the model's visual style is already quite strong. Feel free to let your imagination run wild!

这个模型放弃了写实风格,更偏向于绘画质感,仅仅是偏向,还是保留2.5D的画风,

很大程度的增加风格化方向的探索,对飘逸的头发和衣服的褶皱有非常好的表现力,光线和氛围感的营造也很出彩,但是对lora的支持范围会比较差,请适当调低权重,因为模型本身的画面风格就很强烈了,请放飞你的想象力吧。

There may still be many immature aspects. Please give me more guidance and exchanges.

可能还有很多不成熟的地方,请大家多多指导和交流.

Proper use of facial restoration tools will give you a charming face,

正确使用面部修复工具会让你得到一个迷人的面孔,

Welcome to comment and exchange, your pictures are my biggest motivation for updating.

欢迎评论交流,你的返图就是我更新的最大动力。

--------------------

Read the info below to get the high quality images

(click on show more)

阅读以下信息以获得高质量图像

(点击显示更多)

--------------------

Recommended Parameters:

推荐参数:

VAE: use the 840000 ema pruned

Clip skip: 2

Upscaler: 4x-Ultrasharp

Sampler: DPM ++ SDE Karras

CFG: 6~9

Steps: 20~40

Height: 960px

Width: 720px

Adetailer for Faces

推荐面部修复:Face Editor

don't use Hi.Res fix

不推荐使用高清修复

Negative prompt:

负面提示词:

negative_hand-neg,

bad_prompt_version2-neg,

NG_deepnegative_v1_75t,

bad-hands-5,

FastNegativeV2,

realisticvision-negative-embedding,

--------------------

This model is free for personal use and free for personal merging.

For commercial use, please be sure to contact me.

You are solely responsible for any legal liability resulting from unethical use of this model.

该模型免费供个人使用,免费供个人合并,

如需商业用途,请务必联系我.

因不道德使用本模型而产生的任何法律责任由您自行承担.