
SD1.5 version of the lora taking works by @Xdivine that they posted to discord, and combined with some of my work inspired by those images, trained up to make your anime images pop.
There is a heavy bias to females due to the dataset and nature of anime models, either reduce lora strength or weight the 1boy, man, male tokens
X/Y plots against MeinaMix, RealCartoon-Anime and TMND-Mix there for your comparison to see what it does. Made using sd forge, but only hires fix used for the plots, the front cover uses FreeU, SAG, and LatentModifier (details in PNG Info)
描述:
训练词语:
名称: 353055_training_data.zip
大小 (KB): 23803
类型: Training Data
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success
名称: Divine_RC_style.safetensors
大小 (KB): 36979
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success