If you enjoyed this model, why don't you check out my other works? I put a lot of effort into correctly recreating artist styles and making the occasional character LoRA.๐
You can also buy me a Coffee โ at https://ko-fi.com/stickyricky
I am once again open for suggestions for other styles, SFW and NSFW.
You can always drop model suggestions in the comments, both for styles that are missing on civit or those that are outdated.
This model was trained using the collab/guide by holostrawberry [Guide] Make your own Loras, easy and free - colabs | Stable Diffusion Other | Civitai
About the artist:
At this point who doesn't know this iconic series...
If you somehow missed it, Berserk is one of the most influential (dark) fantasy manga of our time. Sadly, Miura wasn't able to finish it himself but nevertheless it continues, although in a shortened form.
Weights & Prompts
The only necessary prompt is:
kentaro_miura_berserkTry using
monochrome, greyscale, nib pen (medium)for scenery in particular.
Use a weight of 0.8-1.0, adjust to taste and depending on your checkpoint.
As usual I have included an XYZ plot for reference.
As I exclusively trained with monochrome manga panels it will prefer black and white image generation. This doesn't prevent you from creating colorful images but hampers the flexibility. You can also expect a certain drop in the accuracy of the model when forcing colors.
I have also never seen a model interact so weirdly with character LoRA before. I recommend experimenting a lot with the weights for those.
Please leave a review if you enjoyed this model or have problems with it. This helps me with improving future models and finetuning them in general. Give it a try and throw some images down below!
Description
FAQ
Comments (7)
You create such amazing LORA's.
If you are ever looking for suggestions, I would love to see your take on Derpixon & Hyung-Tae Kim..
Ya know.. just in case.. your bored or something.. :cough:
I was going to do these ones myself.. but after seeing how much better your speedy LORA turned out compared to mine..
I think I need to learn more first. x.x
Hey it's you again :D. Thanks for the praise - although this one turned out a bit worse than usual. In hindsight automatic tagging really struggled with the style so I'll have to manually tag 200 images and retry at some point ๐
Derpixon really might be worth doing. I remember downloading some models for their stuff early on. Absolutely terrible performance ๐ฅ
Hyung-Tae Kim looks interesting (older or current stuff?). I wonder how well SD will handle the expressive poses
Hiya, @StickyRicky! Big fan of your LoRAs! I was wondering what VAE you used in your example images?
For anyone who wants the model used in the example imgs, it's here: https://huggingface.co/a1079602570/animefull-final-pruned/blob/main/model.ckpt
Edit1: I'm using this one with pretty good success!
https://huggingface.co/a1079602570/animefull-final-pruned/blob/main/animevae.pt
Hey, thanks for adding the model. Still didn't get around to going through my old models.
I don't use any VAE in my testing. I like it when it works with minimal requirements.
This is incredible. Its as close as it gets to a Kentaro's Miura style and amosphere when it comes to landscapes.
The GOAT, together with Wayne Barlowe.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.









