中庸无为 | PoetAlmightyMix_2.5D版本v1.0 (ID: 159784)

PoetAlmightyMax_2.5D:

Specializing in the style is not particularly outstanding, as I gave the definition of the model, moderate and inactive, silent, inaction, but not inaction, so the eighteen martial arts are all good!

Because the bottom film has a fusion of MAJIC big brother's aesthetic V1, so in the photography and lighting effects of the picture quality will show good, of course, the picture quality is mainly expressed in 2.5D and real people between the content will be good. The main characteristics of the model are light and shadow, texture, tone, color temperature, color, shadow, highlights, saturation and tonal balance are just right ~~~~~ because I'm not what art or painting professional so the wording of the description may be inappropriate, but it's really good, you can refer to the example picture!

Instructions for use:

1. sample images are generated 512x768, tile + usd zoom 2 times, no lora no controlnet, 100% restorable!

2. VAE: normal use recommended vae-ft-mse-84000-ema-pruned.

3. Tip word please also free to play, no specific requirements

4. amplification algorithm recommended R-ESRGAN 4x + Anime6B

5. Sampling method recommended dpm++2m karras, step 30, cfg7.

6. More than half of the body is recommended to open high-resolution fix (Hires. fix)

---

PoetAlmightyMax_2.5D：

擅长风格没有特别出众的，就像我给模型的定义一样，中庸且无为，寂然不动，无为而不无为，所以十八般武艺样样精通！

因为底膜有融合MAJIC大佬的唯美V1，所以在摄影光效方面的画质都会表现的不错，当然画质主要表现在2.5D与真人之间，内容都会不错,模型主要表现的特点有光影，质感，色调，色温，色彩，阴影，高光，饱和度及色调平衡方面都恰到好处~~~~~因为我也不是什么美术或者绘画专业所以形容的措词或许有所不当，但是它真的很不错，具体可以参考例图！

使用说明：

1.示例图均是512x768生成，tile+usd放大2倍，无lora无controlnet，100%可还原

2.VAE：正常使用推荐vae-ft-mse-84000-ema-pruned，

3.提示词还请大家自由发挥，无特定要求

4.放大算法推荐R-ESRGAN 4x+ Anime6B

5.采样方式推荐dpm++2m karras，步数30，cfg7,

6.半身以上建议开高分辨率修复 (Hires. fix)

描述:

PoetAlmightyMax_2.5D:

Specializing in the style is not particularly outstanding, as I gave the definition of the model, moderate and inactive, silent, inaction, but not inaction, so the eighteen martial arts are all good!

Because the bottom film has a fusion of MAJIC big brother's aesthetic V1, so in the photography and lighting effects of the picture quality will show good, of course, the picture quality is mainly expressed in 2.5D and real people between the content will be good. The main characteristics of the model are light and shadow, texture, tone, color temperature, color, shadow, highlights, saturation and tonal balance are just right ~~~~~ because I'm not what art or painting professional so the wording of the description may be inappropriate, but it's really good, you can refer to the example picture!

Instructions for use:

1. sample images are generated 512x768, tile + usd zoom 2 times, no lora no controlnet, 100% restorable!

2. VAE: normal use recommended vae-ft-mse-84000-ema-pruned.

3. Tip word please also free to play, no specific requirements

4. amplification algorithm recommended R-ESRGAN 4x + Anime6B

5. Sampling method recommended dpm++2m karras, step 30, cfg7.

6. More than half of the body is recommended to open high-resolution fix (Hires. fix)

PoetAlmightyMax_2.5D：

擅长风格没有特别出众的，就像我给模型的定义一样，中庸且无为，寂然不动，无为而不无为，所以十八般武艺样样精通！

因为底膜有融合MAJIC大佬的唯美V1，所以在摄影光效方面的画质都会表现的不错，当然画质主要表现在2.5D与真人之间，内容都会不错。模型主要表现的特点有光影，质感，色调，色温，色彩，阴影，高光，饱和度及色调平衡方面都恰到好处~~~~~因为我也不是什么美术或者绘画专业所以形容的措词或许有所不当，但是它真的很不错，具体可以参考例图！

使用说明：

1.示例图均是512x768生成，tile+usd放大2倍，无lora无controlnet，100%可还原

2.VAE：正常使用推荐vae-ft-mse-84000-ema-pruned，

3.提示词还请大家自由发挥，无特定要求

4.放大算法推荐R-ESRGAN 4x+ Anime6B

5.采样方式推荐dpm++2m karras，步数30，cfg7。

6.半身以上建议开高分辨率修复 (Hires. fix)

训练词语: absurdres,ultra detailed,highres