AI Roguelite

AI Roguelite

评价数不足
NoobAI-XL ComfyUI Workflow
   
奖励
收藏
已收藏
取消收藏
文件大小
发表于
更新日期
184.075 KB
7 月 5 日 下午 12:52
8 月 17 日 上午 4:41
2 项改动说明 ( 查看 )

订阅以下载
NoobAI-XL ComfyUI Workflow

描述
This ComfyUI custom workflow enables you to use the NoobAI-XL V-Prediction 1.0 image model.

You will need this model:
https://civitai.com/models/833294/noobai-xl-nai-xl

And this LoRA:
https://civitai.com/models/99619

Install them in your ComfyUI for this workflow to work correctly.

You can also download the workflow independently here:
https://files.catbox.moe/7alagf.png


RECOMMENDED GENERATION SETTINGS

1024x1024 at 28 steps
Negative prompt: worst quality, bad quality
All images set to max crop.


UPDATE CHANGELOG

Aug 17 2025: Updated the workflow to include the quality control LoRA by avokat. Greatly increases generation quality.

Jul 5 2025: Initial release.
5 条留言
Johnman  [作者] 8 月 24 日 上午 6:17 
The details are a bit too technical for me, but from what I gather from the LoRA's model card, it only affects the quality of the generation.
felinowottgald 8 月 19 日 上午 10:33 
ah, it updated, I was worried that after I tried to use it I was getting errors no matter the v-pred model I used, good to know that was only a lora I needed, does this lora affect only quality or also help the model to work on natural language?
Johnman  [作者] 8 月 17 日 上午 4:47 
Updated the mod to include avokat's quality control LoRA for NoobAI-XL. This improves the quality of the generations immensely. You need to make sure you have both the model and the LoRA installed.
Johnman  [作者] 7 月 31 日 上午 10:12 
@nkdi2211 You can now find the catboxed workflow in the description.
nkdi2211 7 月 30 日 下午 11:40 
I set it up, but I don't know if I mistakenly delete something. Can you upload your workflow? The image generation is so long, I don't know if it's because my crap GPU or because I just use to default on Comfyui.