The LoRA model was built on 75 images taken from publicly available sources on the Internet. LoRA was trained on the FLUX Dev model Improved image quality, works well with other FLUX models.
Weight: 0.3 - 1.
Works well with weight: (0.4 - 0.6), experiment with weight
It depends on the model, but you can raise the Lora weight quite high.
High quality prompts reduce the negative effect.