这个模型使用了超过180张高质量的原始图片进行训练。
这个模型拥有很好的画风模仿效果。TV版 《魔法骑士》 图片素材没有混入其中，得到纯净的CLAMP风格LORA。
illustriousXL 版本：https://civarchive.com/models/1361986
Pony XL版本：https://civarchive.com/models/480483
推荐权重：0.6~1.0
推荐模型：Animagine XL V3
--- --- --- --- --- --- --- ---
This model used more than 180 high-quality original pictures for training.
This model has a very good imitation effect of painting style.
All the training pictures come from the collection of "Magic knight Rayearth", and the TV version animation picture materials are not mixed in, so as to obtain the pure artist style LORA.
illustriousXL version：https://civarchive.com/models/1361986
Pony XL version：https://civarchive.com/models/480483
Recommendation Weight：0.6~1.0
Recommendation Model：Animagine XL V3
Description
FAQ
Comments (9)
Can you discuss your process? How do those eyes look near perfect?
In fact, there is no special process, it only requires a clear enough training set and detailed labels.
The learning rate should not be too high, epochs should be 10-15.
@flyx3 Does the fact that it's SD XL have anything to do with it?
@xSatouMatsuzakax SDXL has higher accuracy and details, SD1.5 also has good results.
@flyx3 Thank you! I haven't switched to SD XL because there's more content for 1.5, but maybe I will at some point. Then again, SD 3 is finally coming out, and I can't help but want to use that instead. I'm pretty torn.
I will use the MODEL and LoRA of the work.
This one was fun. Thanks for sharing.