相较于v1.0泛化性好了很多，某些概念过拟合的情况也得到了明显的改善，保留了底模构图的同时肢体控制和动作上会更稳定，在默认权重下风格会柔和一些，推荐控制在0.8-1.2之间来达到想要的画面效果。
如果人物在画面占比较少导致丢失细节，可以通过ADetailer来修复。
使用animagine角色列表中收录的触发词来还原人物，如：“1girl, ryoshu \(project moon\), limbus company”、“1girl, souryuu asuka langley, neon genesis evangelion”。
由于这个版本没有针对人物进行任何训练，完全依赖于animagine本身收录的角色，所以对于游戏内人物的身体比例、细节、还原度上不如上一个版本（比如良秀本来会随身携带的太刀），或许可以配合游戏内角色的lora来改善这个问题？
如果只是想还原游戏内立绘的服装、武器来达到类似假新闻的效果，可以使用1.0的版本，否则更推荐2.0的版本，无论是配合animagine收录的角色或者和其他的lora配合使用，都会有更好的效果。
如果对animagine的效果不满意，推荐尝试一下Ponymagine 9.1（质量提示词请使用animagine的版本），可能会有更好的表现。
gpt translate:
Compared to version 1.0, this version has much better generalization and has significantly improved the issue of overfitting certain concepts. While retaining the base model's composition, limb control and actions are more stable. The style is softer at the default weight, and it is recommended to adjust the weight between 0.8 and 1.2 to achieve the desired visual effect.
If the characters appear small in the image, causing a loss of detail, ADetailer can be used to fix this.
To recreate characters, use the trigger words included in the Animagine character list, such as: "1girl, ryoshu (project moon), limbus company" or "1girl, souryuu asuka langley, neon genesis evangelion"
Since no specific training was conducted for characters in this version and it completely relies on the characters included in Animagine, the body proportions, details and fidelity of in-game characters are not as good as the previous version (for example, Ryoshu originally carried a katana). Perhaps using LoRAs of in-game characters could improve this issue.
If you only want to recreate the costumes and weapons from in-game illustrations to achieve a fake news-like effect, you can use version 1.0. Otherwise, version 2.0 is recommended for better results, whether used with the characters included in Animagine or in conjunction with other LoRAs.
If you are not satisfied with the effects of Animagine, it is recommended to try Ponymagine 9.1 (for quality prompts, use the animagine version), which may perform better.
Description
FAQ
Comments (18)
Works for a pony,and maybe you can make a version for pony ?
Yes, I also have this idea, but I am still learning about this area. I will try to do it later.
I've tried, but the results weren't very good. Maybe you can give Ponymagine 9.1 a try?
@AvaToNight i used MFCG PDXL ,I just liked that your LoRa has characters from limbus. I would like a separate LORA.
How do I use this when the software doesn't seem to recognize it when I put it in the lora folder
Is the base model of the checkpoint you are using SDXL 1.0?
Cool
To make it easier:
1girl, ishmael \(project moon\), limbus company
1boy, male focus, sinclair \(project moon\), limbus company
1girl, don quixote \(project moon\), limbus company
1boy, male focus, yi sang \(project moon\), limbus company
1girl, faust \(project moon\), limbus company
1boy, male focus, hong lu \(project moon\), limbus company
1other, dante \(limbus company\), limbus company
1girl, ryoshu \(project moon\), limbus company
1girl, outis \(project moon\), limbus company
1boy, male focus, heathcliff \(project moon\), limbus company
1boy, male focus, meursault \(project moon\), limbus company
1girl, rodion \(project moon\), limbus company
1boy, male focus, gregor \(project moon\), limbus company
1girl, kromer \(project moon\), limbus company
1girl, yuri \(project moon\), limbus company
There was no all the characters in the file. Is the others included in the lora?
@HugMeIntoFace If you are referring to other characters in the game, they are not in lora and only the characters listed in characterfull.txt are supported.
@AvaToNight Thank you for the answer. :)
Project Moon has a beautiful art style and I thank Ava for posting this, but what is going on with the eyes in some of these pictures?
Asking if I use the art style myself one day.
Yes, there are indeed some noticeable issues with the eyes for the most part. Perhaps I should train specifically for that area. If I have the time, I'll try to improve upon it.
@AvaToNight Nice. In any case, I just used it to make a character wearing a blindfold so it did not matter for that one in particular. Going to be posting that to get the daily post bonus after I post the rest of my backlog.
@AvaToNight Seems like people were skipping out on version 2 due to the loss of aesthetic and because the eye problem got worse.
I may not have made any AI resources before, but can you think of any reason the eyes are getting messed up?
I am thinking the AI associated eye shape and color to the characters more than anything else, causing prompts without LC characters to be messed up. That's just me throwing ideas into the air though.
@DiamondLead When generating full-body images, the face occupies a small portion of the image, resulting in fewer pixels available to render detailed facial features. This might be the reason for issues with the eyes. Using ADetailer to redraw the face can effectively solve this problem.
@AvaToNight Ah, that also makes sense. I was thinking the eye problem was like how I said because only non-LC characters seem to encounter the problem. If that is the case, then it is probably drawing the eyes by way of drawing the character. Cool learn that.
@AvaToNight Ah, that also makes sense. I was thinking the eye problem was like how I said because only non-LC characters seem to encounter the problem. If that is the case, then it is probably drawing the eyes by way of drawing the character. Cool learn that.