Lora model customization: click here
Animation series: Click here
Version Description
V100: Can change clothes, hair color, eye color, support various main models, support 512-1280 resolution
The model is made based on ch modal。
The image source and training code come from the Internet, and the model is only used for exchange of scientific research interests. If there is any infringement, please contact to delete, thank you.
图源和训练代码来源于网络,模型仅作为科研兴趣交流使用,如有侵权请联系删除,谢谢!
画像ソースとトレーニング コードはインターネットから取得され、モデルは科学研究の利益の交換にのみ使用されます.侵害がある場合は、連絡して削除してください。
The model is made based on a100 processor, and it needs high configuration to generate graphics. It is recommended to use 3090+or 4090, and the gpu has a size of more than 20gb, and the initial resolution is not less than 1000, otherwise it cannot be used normally.
In many cases, it is very difficult to generate a long-range full-length photo of a character. This is the first lora model for a real full boby.
infor:Steps: 20, Sampler: DPM++ SDE Karras, CFG scale: 7, Seed: 3889164490, Face restoration: CodeFormer, Size: 768x960, Model hash: bbf07e3a1c, Model: basil mix, Denoising strength: 0.5, Clip skip: 2, Hires upscale: 2, Hires steps: 15, Hires upscaler: R-ESRGAN 4x+ Anime6B, AddNet Enabled: True, AddNet Module 1: LoRA, AddNet Model 1: bastly-3(60b5617a16b1), AddNet Weight A 1: 0.7, AddNet Weight B 1: 0.7
Existing problems
It takes a long time to draw a picture.
There are problems with toes and knees.