
After testing, this LoRA seems very "strong". Weight 0.8 feels like weight 1, weight 1 feels like weight 1.2. I'd recommend running anywhere between weight 0.5 to 1.
Keyword is daijrenjisuzuka. Outfits can be prompted with (suzuka_outfit_dress, black_ribbon) and (suzuka_outfit_jacket, pink_ribbon). Note difference between smile and evil_smile.
Issues
-
Her dress outfit seems semi-baked into the model, especially at high weights. If you're trying different outfits from the standard ones, lower the weight.
-
Her fingernails are inconsistent.
Pic 1&4: MeinaUnreal
Pic 2: AnythingV5
Pic 3: AnyLoRA
描述:
Trained on 98 images.
训练词语: dairenjisuzuka,suzuka_outfit_dress,suzuka_outfit_jacket
名称: Dairenji_Suzuka.safetensors
大小 (KB): 36981
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success