I trained this LoRA using the "qwen_image_distill_full_fp8_e4m3fn" base model. It was tested together with "Qwen-Image-Lightning-8steps-V2.0.safetensors".For the Footgag LoRA, the strength may vary depending on the pose. I generally used it between 0.78 and 0.96. Sometimes it may be necessary to increase the power up to 1.2. For the Lightning LoRA, a 0.85 weight gave good results.The trigger word is "Footgagging".When depicting the act of putting the foot in the mouth, be sure to use "insert". Styles:animecartoonrealisticFantasy3D CGI Below are some example scenes. I tried to use as varied poses as possible.

qwen_fp8_footgag

I trained this LoRA using the "qwen_image_distill_full_fp8_e4m3fn" base model. It was tested together with "Qwen-Image-Lightning-8steps-V2.0.safetensors".For the Footgag LoRA, the strength may vary depending on the pose. I generally used it between 0.78 and 0.96. For the Lightning LoRA, a 0.85 weight gave good results.The trigger word is "Footgagging".When depicting the act of putting the foot in the mouth, be sure to use "insert".Below are some example scenes. I tried to use as varied poses as possible.

Files with SHA256: 556313e2929bc10203e7b4ebfbf796d5177f8c1eed95fdfda97e3fdd2b580869