
v2 should be more consistent and a bit more baked, recommended weight is 0.9
This LoRA tries to replicate the style used by Misaka12003, this model has a tendency to undress characters, due to how NSFW heavy the dataset was, so yes this lora works nicely with models like Hassaku wink wink. Requested and paid with some buzz by lycorisqian.
Recommended VAE are Counterfeit and OrangeMix or any NAI renamed, recommended Weight is 0.7 to 1.1, this was trained on NAI so it should be quite compatible.
All preview images have Hires and Adetailer, LoRA likes VectorScope but not as much as others.
Any and all reviews are very much appreciated so I can know what to improve for this LoRA or for other models, thank you~
I just opened a Discord server where you can go and bother me about which style I'm working rn, or suggest ones to try my hand at.
描述:
Initial release
训练词语:
名称: Misaka120003-Style.safetensors
大小 (KB): 147592
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success