
Feel free to play with the weight - it all depends on the model and prompts you are using. Generally, you can set the weight to 0.8-1 when generating your own character. If you are generating characters from the game (especially when using a character LoRA), however, you may want try a lower weight. Illustrious-based models can and will mimic the game's art style to some level when you include relevant character tags, and it is hard to predict outcomes when the effect is blended with this LoRA, so there is no "optimized weight", feel free to find one that is the best for your generation.
It is possible if you get poor-quality eyes. It can be easily fixed by Hires.fix and/or ADetailer.
描述:
训练词语:
名称: 1187845_training_data.zip
大小 (KB): 216993
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: PCR_Style-000019.safetensors
大小 (KB): 223129
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success