These are LoRAs for cham22's styles, categorized by period.
If you are using an Anima checkpoint model, you can generate cham22's art style without a LoRA by entering "@chamchami" into the prompt. However, he has been active for a long time, and his art style has evolved over the years. Since the checkpoint model has learned all of his styles, it tends to generate an intermediate style that doesn't represent any specific period. I created these LoRAs to capture his art style from each individual period. I plan to create more for other periods later.
I will remove this if requested by the original artist.
Warning: Some of these LoRAs tend to generate images with incorrect colors. It is strongly recommended to set CFG to 4.0 or lower.
Early 2026 Version
Trigger words: @chamearly2026,
Recommended strength: Around 1.0
Later 2021 Version
Trigger words: @chamold,
Recommended strength: Around 1.0
Description
Trained based on Anima Preview v01.
FAQ
Comments (9)
Do you have this for Illustrious?
Read the description on this page properly!
Great work 👍 one small nitpick is. Do you know how to make the outer lines of the character less prominent?
The only thing I can think of as a way to deal with that is to enter "thick outlines, outline" into the negative prompt.
@toghashie441 That's actually a really good negative prompt. Thank you for that. I didn't think of that. 🤣🤣
Also how do you train a LORA with anima as I want to train my own style. Also if you have time could you do a big breasts/saggy breasts LORA especially with Lactation As I think this style is perfect for that. Honestly really great work. Well done.
Can you do one with Dikko? Model already knows him, but it is far from perfect and I'm pretty sure the preview model we have now is trained at low resolution
I checked his X bio and found that he says 'NO AI'. I can't make a LoRA for him.
Yo, can you post what training tool and parameters you use for style training anima?
I use kohya ss's sd-script with its GUI made by RedRayz. RedRayz provides recommended settings for technophobes like me, and I use those settings with only minor adjustments to create LoRAs. I can't explain the details much, but the settings are saved as metadata within the LoRA itself, so please check them yourself.