V2.0
Optimized the training set and improved skin tone
V1.0-1
This version is not much different from 1.0. I used a PS filter to enhance the red hues in the training set, giving the skin a more rosy tone in certain models
If you are using the noob's VP version, it may cause the skin to appear too red. Please lower the intensity to resolve this issue
V1.0
cfg scale:3~6
no prompts
This Lora can produce relatively good results with the models listed in the Suggested Resources below, and theoretically works best with the Illustrious model
This is a version with fairer skin, and the dataset has not been processed with filters
I strongly recommend using it in combination with the anime style I released earlier; the fusion of the two works quite well
描述:
训练词语:
名称: Xi410_type2_fix.safetensors
大小 (KB): 223123
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success