Illustrious
Use it with a weight of 0.8 to 1
Enjoy it!
V2.0
training was improved
the training dataset was improved
The error with the viewer was corrected (although it can still happen, but not as often)
Full body generation improved
the flexibility of the lora was improved
enjoy it!!
Note:
1. When it is detected that the input language is not English, the results automatically translated into English are displayed.
2. The translation results cannot be modified directly.
3. You can fill in the translation content into the prompt word input box with one click and modify it.
4. Supported objects: SD workspace, AI Tools use
Thanks, hope this will be much helpful❤️
V1.5
improved training
changed the basic training model
2D, 2.5D and realistic training now possible
for realistic place the weight between 0.9 and 1
for anime or 2.5D use the weight from 0.7 to 0.8
some errors depending on the model
sometimes it will come out without the eyes covered, the solution is to increase the weight.
the wings will sometimes come out wrong, solution is to regenerate the image.
these errors are not of the lora but of the models in some it will give better results and in others not.
I hope you enjoy it and have fun, share your generations!
V1.0
This loras has been trained with anything v5, with the recommended weight of 0.7 to 1
looking for digimon loras in civitai I realized that there was no kazemon (fairimon in Japanese), which seemed strange to me, so I decided to create one.
Share your generations in the comments and just enjoy it!!
Description
training was improved
the training dataset was improved
The error with the viewer was corrected (although it can still happen, but not as often)
Full body generation improved
the flexibility of the lora was improved
enjoy it!!