qwen_distill_fp8_latex_mask

I trained this LoRA using the “qwen_image_distill_full_fp8_e4m3fn” base model.
I’m not sure if it works properly with other standard models. It was tested together with “Qwen-Image-Lightning-8steps-V2.0.safetensors”.

For the latex LoRA, I achieved the best results using a weight of 0.78.
For the Lightning LoRA, a weight of 0.85 worked well.

The main trigger phrase should describe the action of a woman inserting her foot (or feet) into the mask.
Make sure to use the verb “insert”, as it is essential for the model to understand the action correctly.

Example prompts can be found in the sample images below — feel free to analyze them with AI tools for reference.