Trained on PonyDiffusion V6 with about 300 Japanese anime style NSFW images. This model replicates the SilvermoonMix01 on SD1.5: https://civarchive.com/models/53642/silvermoonmix01-or-delicate-skin-and-hair-texture
Recommended settings:
Prompts: score_9, score_8_up, score_7_up, score_6_up
Sampling method: Euler a
Recommend use ADetailer extension to enhance face details: https://github.com/Bing-su/adetailer
Change notes:
v2.0: Big improvements in background and limbs logic.
v1.1: Updated training parameters. This improves background logic and limbs.
Description
FAQ
Comments (8)
Hi, thx for this model, it's possible to get the list of the artist style ?
Hi, this model is designed for one specific style and doesn't need a trigger word.
If you are looking for a certain artist style, you can check my published LoRAs.
@silvermoong Okay ^^ because i see that you use 300 style images to train it so wonder if it's possible to mix them like on NAI3. Because using LoRa it's become a mess when i mix more than 2 LoRa style.
群友路过感谢银月大佬
看着有点像石惠的画风……
Nice Job Man :)