Tags I used to train / Tags you should use:<pre><code>kyukim0n, pink body, yellow stomach, yellow neck, pink head, red markings, red tribals, blade arms, pink arms, pink legs, pink feet, white claws, 3 claws, 3 toes, blue hair, long hair, big hair, long tail, blue tail, black nose, pink ears, yellow inner ears, blue star marking on forehead, sickle in mouth, golden sickles, grey arm blades, red heel feathers, pink shoulder feathers, red tips on shoulder feathers</code></pre>v2.0: I retrained the lora with a few images removed and a few created with v1.0 added.Now styles should work better. Also furry checkpoints such as NovaFurry, and many others are compatible (I tested 6+)Known Issues: Hands might appear instead of blades. Especially if the character is leaning against something - be it a bed, a wall or the ground. It helps to add "arms up" or "blades in foreground" or play with negatives. / Also adding "white face, white lower jaws" can help if it should add colored jaws. The red markings are of course still not accurate, but that is most likely the most difficult thing with a small dataset.Inspiration: Digimon can change their shape with apps, functions and data cards etc... I added two action images where I added hands instead of blades to show you can play around with it. Also just having hands instead of blades can be nice with this character.You can also add sickles into the negative and remove it from the positive to have the character without it, which I also find nice to do. I mean I understand being ready to fight, but not all the time - if you know what I mean ;3Adetailer is of course still required. I still plan to keep making this lora better, but v2.0 is by far not perfect but a step forward in my opinion.v1.0: I did not see this Digimon anywhere, so I did this. This is the first version, and I plan to make a better version in the future. There are not a lot of good images you can use to train this character and most of them are low quality and very small resolution. Once there is more material created (with AI), I think I can make a better version soon. Don't expect the eyes to look good - aDetailer required. The red markings will not be accurate - yet.It is difficult to prevent hands to appear. Many furry typical tags and checkpoints add hands to the character. I'll try my best to find a better solution for version 2. For now you'll need to experiment a bit with negatives. I should add it as a concept and not just as a character.Those are all things I plan to make better next time. All example images had lora strength:1  NOTE: This model mainly exists to create a better training set for version 2. Please note, that most furry models make this lora hard to work with.

Kyukimon (Digimon) - Illustrious

It is now better compatible with many furry checkpoints. Read the main description for more details.

Files with SHA256: 07ea14080228bb6f2dabeb38392376bf207a7f5efbf2718a38018388bd3cf461