Li Muwan - Xian Ni

lora uses the SD1.5 official model for training, recommended base model: Guoman SU 2.5D (homepage model)

The image above shows the comparison between the weights opened by this lora and the corresponding effects

Parameter recommendations:

1.clip:2

2.Sampling method recommendation:DPM++2M Karras

3. Sampling steps:25-30

4. When the generated full-body image or the area occupied by the character’s face is small,it is recommended to enableAdetailerplugin for facial and hand repairs to ensure normal appearance.

5. If you have any questions, you can discuss in the comment section. Welcome everyone to submit images!