If you use my models, please take the time to post some images to the pages so I can see how they are being used and make improvements in the future. Also feel free to reach out if you have ideas for future models!

This model celebrates the single above knee amputee.

It aims to be extremely reliable, generating passable photorealistic images at least 90% of the time, even with simple prompts. Expanding your prompts using an LLM can improve results further.

The main fault with this model is that it sometimes removes both legs. While I've been able to reduce the probability of this, it still happens with certain prompts more than it should. There were zero such examples in the training data, but Flux's understanding of anatomical symmetry is difficult to beat back.

The model was trained with a large regularization data set of high quality photographs, which can give it an artsy photographic feel. This also helps to improve on several aspects of Flux, including challenges with photorealism, skin texture, and overfitting certain facial features and body types.

Generating NSFW content was not a goal during model training, and NSFW images will exhibit typical Flux defects. Feel free to experiment with other LoRA's or base models if that is your goal.

All sample images use the simplest Flux ComfyUI workflows, with flux-dev as the base, and no other LoRAs loaded. 20 Steps, Euler, Guidance of 2.5-3.0, and lora weight=1.0. They all include the phrase "legless amputee" or "leg amputee". Use the phrase "35mm photo" to enhance photorealism.