This is a lora trained on the style of Kanabun. I used 51-150 of their most recent images (at the time), so the style is a mix from that batch exclusively. After testing it for a while, I'm confident the lora replicates the style very well.
One thing to mention, the source material has a lot of censorship. So sadly, the pony version tends to censor more often than not.
Version 3.0 was trained on Anima V1.0.
The base for version 2.0 is Noob 1.0 V-Pred and Illustrious-XL v0.1.
The base for version 1.0 is Pony Diffusion V6 XL. You can also use Autism_Pony.
To get the results you see in the images above, these are the settings I used.
Visit My Site For Models I Can't Post On CivitAI: https://qtgurl.com/
Description
Trained on 51 images selected from the version 1.0 dataset. The base for this lora is Noob 1.0 V-Pred.
This lora uses slightly different and improved settings to accommodate for V-Prediction. V-Pred improves prompt comprehension and color range greatly. Noob has a nice range of concepts compare to Illustrious 0.1. You need to make sure you're using the proper UI and settings in order to utilize V-Pred. I don't expect this to work on Eps. models like Illustrious 0.1.
FAQ
Comments (6)
Why don't you just stop making your own models of the author's drawings, which you have publicly stated are banned for AI training?
It's fine to enjoy them in private, but why not make them publicly available for everyone to see
As far as I'm concern, if the images are publicly accessible on the internet, then they are open to be trained on and given to the public to use for personal use. At least I'm not charging money like a lot of companies and people are doing. From the public, to the public, is my creed. If the author doesn't want their worked trained on, they should lock it behind a pay wall. That way, even the companies can't use their images to make profit.
@Fish788
I understand your beliefs, but the author himself told you not to learn, so you should follow his advice.
I have been in contact with kanabun himself, and he has never agreed to make a learned model of his picture.
Please do not make him suffer any more. If possible, please make the model private immediately.
@nekoko2x948 If I complied with any take down request, I might as well take all my loras down. As well as all the lora on this site. As well as the big companies. None of us have "asked for permission" to use the data on the internet. If having this lora up is so damaging for the artist, then you or he should get in contact with CivitAi and force it removed. While you're at it, force every other lora/checkpoint to be removed as well.
They're not banned for training. You clearly have no idea how this works.
How about Konnyaku new style?
Details
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.