I found a new AI tool Shakker, a best image to image tool. You can try it via https://www.shakker.ai ,it can help you:
-Remix: Upload a picture. Just switch the prompts, and you can create stunning images in the same style.
-Style Transfer: Shakker not only extracts the style,but also switches among various styles.
Besides, Shakker also offers Object Control,Composition Control,Collage Redrawing etc.
Niji-Expressive V2 launch!!
灵动优雅，张力尽显
本次升级泛用性，能在更多模型和采样方式上适用
选取500+优质且风格统一的Nijijourney图片，进行更高精度训练
推荐关键词：letterboxed, illustration,
目前仍然存在手部和肢体的bug（甚至更严重）
十分推荐使用“bad promt，easynegative，badhand”等负面embedding
======================
Vivid and elegant, full of tension
This upgrade is generalized and can be applied to more models and sampling methods
500+ Nijijourney pictures with high quality and uniform style were selected for higher precision training
Recommended keywords: letterboxed, illustration,
- There are still hand and limb bugs (even worse)
It is highly recommended to employ "bad promt, easynegative, badhand" and other negative embedding
Description
FAQ
Comments (21)
哇，好棒的模型！是上一代的新升级耶！
说的好，下次记得切小号😂
其实我有个思路，midjourney生成出来的难看的图（如果有的话）不要丢，直接练成negative embedding，搭配使用效果应该也不错。
@NotAPersonJustACat 哈哈哈，问题就是，mj很难出丑的图，相反sd这些负面embedding如果放正面会真的很丑
@Terry9993 这么强啊。你订阅的是哪个档位的？不知道一人使用$30的会不会不够
@NotAPersonJustACat 够的够的！而且30刀是无限出图，只不过时间慢一些，但绝对够
这次的lora线条感更强烈，更有插画风，尤其加强场景和植物
Recommended weight (0.8 ~ 0.9)
The biggest regret of this upgrade is that compared with v1, many good-looking characters do not have relevant prompt words. It is not clear whether you are intentional or careless, but it makes me feel very sad
Sorry, I don't know whether this is the problem of civitai or my pictures. All the uploaded sample pictures lacked prompt. In fact, the pictures I used in training were not much different from those in v1 version, only the training parameters were changed
The Blue Dragon
prompt
letterboxed,blue theme, male_focus, black_hair, weapon, solo, blue_eyes, sandals, 1boy, japanese_clothes, sword, illustration.media, ,action shot, <lora:NijiExpressV2:0.9>, masterpiece, best quality,
negative prompt
badhandv4 修手 verybadimagenegative_v1.3 easynegative, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry
Steps: 28, Sampler: DPM++ 2M Karras, CFG scale: 7, Seed: 743795245, Size: 584x912, Model hash: 7364c31aac, Model: facebombmix_v1Bakedvae, Denoising strength: 0.4, ENSD: 31337, Hires upscale: 2, Hires steps: 11, Hires upscaler: R-ESRGAN 4x+ Anime6B
（甚至更严重）
This model LOVES dragons haha
想问一下这个lora的话是会在哪方面改变图片a
画风
@Terry9993 好的谢谢你
Details
Files
NijiExpressV2.safetensors
Mirrors
nijiexpressV2_v20.safetensors
NijiExpress_v2.safetensors
NijiExpressV2.safetensors
NijiExpressV2.safetensors
nijiexpressV2_v20.safetensors
NijiExpressV2.safetensors
NijiExpressV2.safetensors
NijiExpressV2.safetensors
NijiExpressive_v2.safetensors
NijiExpressV2.safetensors
nijiexpressV2_v20.safetensors
31.safetensors
nijiexpressV2_v20.safetensors
NijiExpressive_V2_v1.0.safetensors
NijiExpressive_V2_v1.0.safetensors
NijiExpressV2.safetensors
NijiExpressV2.safetensors
NijiExpressV2.safetensors
nijiexpressV2_v20.safetensors
nijiexpressV2_v20.safetensors
NijiExpressV2.safetensors
NijiExpressive_V2_v1.0.safetensors